Mercurial > emacs
annotate doc/lispintro/emacs-lisp-intro.texi @ 84201:1e032858dd3e
Move to ../doc/emacs/, misc/
author | Glenn Morris <rgm@gnu.org> |
---|---|
date | Thu, 06 Sep 2007 04:40:47 +0000 |
parents | f0912c72f201 |
children | 27b108c35388 |
rev | line source |
---|---|
83955 | 1 \input texinfo @c -*-texinfo-*- |
2 @comment %**start of header | |
83965
f0912c72f201
(setfilename): Go up one more level.
Glenn Morris <rgm@gnu.org>
parents:
83955
diff
changeset
|
3 @setfilename ../../info/eintr |
83955 | 4 @c setfilename emacs-lisp-intro.info |
5 @c sethtmlfilename emacs-lisp-intro.html | |
6 @settitle Programming in Emacs Lisp | |
7 @syncodeindex vr cp | |
8 @syncodeindex fn cp | |
9 @setchapternewpage odd | |
10 @finalout | |
11 | |
12 @c --------- | |
13 @c <<<< For hard copy printing, this file is now | |
14 @c set for smallbook, which works for all sizes | |
15 @c of paper, and with Postscript figures >>>> | |
16 @smallbook | |
17 @clear largebook | |
18 @set print-postscript-figures | |
19 @c set largebook | |
20 @c clear print-postscript-figures | |
21 @c --------- | |
22 | |
23 @comment %**end of header | |
24 | |
25 @set edition-number 3.07 | |
26 @set update-date 9 November 2006 | |
27 | |
28 @ignore | |
29 ## Summary of shell commands to create various output formats: | |
30 | |
31 pushd /usr/local/src/emacs/lispintro/ | |
32 ## pushd /u/intro/ | |
33 | |
34 ## Info output | |
35 makeinfo --paragraph-indent=0 --verbose emacs-lisp-intro.texi | |
36 | |
37 ## ;; (progn (when (bufferp (get-buffer "*info*")) (kill-buffer "*info*")) (info "/usr/local/src/emacs/info/eintr")) | |
38 | |
39 ## DVI output | |
40 texi2dvi emacs-lisp-intro.texi | |
41 | |
42 ## xdvi -margins 24pt -topmargin 4pt -offsets 24pt -geometry 760x1140 -s 5 -useTeXpages -mousemode 1 emacs-lisp-intro.dvi & | |
43 | |
44 ## HTML output | |
45 makeinfo --html --no-split --verbose emacs-lisp-intro.texi | |
46 | |
47 ## galeon emacs-lisp-intro.html | |
48 | |
49 ## Plain text output | |
50 makeinfo --fill-column=70 --no-split --paragraph-indent=0 \ | |
51 --verbose --no-headers --output=emacs-lisp-intro.txt emacs-lisp-intro.texi | |
52 | |
53 popd | |
54 | |
55 # as user `root' | |
56 # insert thumbdrive | |
57 mtusb # mount -v -t ext3 /dev/sda /mnt | |
58 cp -v /u/intro/emacs-lisp-intro.texi /mnt/backup/intro/emacs-lisp-intro.texi | |
59 umtusb # umount -v /mnt | |
60 # remove thumbdrive | |
61 | |
62 ## Other shell commands | |
63 | |
64 pushd /usr/local/src/emacs/lispintro/ | |
65 ## pushd /u/intro/ | |
66 | |
67 ## PDF | |
68 texi2dvi --pdf emacs-lisp-intro.texi | |
69 # xpdf emacs-lisp-intro.pdf & | |
70 | |
71 ## DocBook -- note file extension | |
72 makeinfo --docbook --no-split --paragraph-indent=0 \ | |
73 --verbose --output=emacs-lisp-intro.docbook emacs-lisp-intro.texi | |
74 | |
75 ## XML with a Texinfo DTD -- note file extension | |
76 makeinfo --xml --no-split --paragraph-indent=0 \ | |
77 --verbose --output=emacs-lisp-intro.texinfoxml emacs-lisp-intro.texi | |
78 | |
79 ## PostScript (needs DVI) | |
80 # gv emacs-lisp-intro.ps & | |
81 # Create DVI if we lack it | |
82 # texi2dvi emacs-lisp-intro.texi | |
83 dvips emacs-lisp-intro.dvi -o emacs-lisp-intro.ps | |
84 | |
85 ## RTF (needs HTML) | |
86 # Use OpenOffice to view RTF | |
87 # Create HTML if we lack it | |
88 # makeinfo --no-split --html emacs-lisp-intro.texi | |
89 /usr/local/src/html2rtf.pl emacs-lisp-intro.html | |
90 | |
91 ## LaTeX (needs RTF) | |
92 /usr/bin/rtf2latex emacs-lisp-intro.rtf | |
93 | |
94 popd | |
95 | |
96 @end ignore | |
97 | |
98 @c ================ Included Figures ================ | |
99 | |
100 @c Set print-postscript-figures if you print PostScript figures. | |
101 @c If you clear this, the ten figures will be printed as ASCII diagrams. | |
102 @c (This is not relevant to Info, since Info only handles ASCII.) | |
103 @c Your site may require editing changes to print PostScript; in this | |
104 @c case, search for `print-postscript-figures' and make appropriate changes. | |
105 | |
106 @c ================ How to Create an Info file ================ | |
107 | |
108 @c If you have `makeinfo' installed, run the following command | |
109 | |
110 @c makeinfo emacs-lisp-intro.texi | |
111 | |
112 @c or, if you want a single, large Info file, and no paragraph indents: | |
113 @c makeinfo --no-split --paragraph-indent=0 --verbose emacs-lisp-intro.texi | |
114 | |
115 @c After creating the Info file, edit your Info `dir' file, if the | |
116 @c `dircategory' section below does not enable your system to | |
117 @c install the manual automatically. | |
118 @c (The `dir' file is often in the `/usr/local/share/info/' directory.) | |
119 | |
120 @c ================ How to Create an HTML file ================ | |
121 | |
122 @c To convert to HTML format | |
123 @c makeinfo --html --no-split --verbose emacs-lisp-intro.texi | |
124 | |
125 @c ================ How to Print a Book in Various Sizes ================ | |
126 | |
127 @c This book can be printed in any of three different sizes. | |
128 @c In the above header, set @-commands appropriately. | |
129 | |
130 @c 7 by 9.25 inches: | |
131 @c @smallbook | |
132 @c @clear largebook | |
133 | |
134 @c 8.5 by 11 inches: | |
135 @c @c smallbook | |
136 @c @set largebook | |
137 | |
138 @c European A4 size paper: | |
139 @c @c smallbook | |
140 @c @afourpaper | |
141 @c @set largebook | |
142 | |
143 @c ================ How to Typeset and Print ================ | |
144 | |
145 @c If you do not include PostScript figures, run either of the | |
146 @c following command sequences, or similar commands suited to your | |
147 @c system: | |
148 | |
149 @c texi2dvi emacs-lisp-intro.texi | |
150 @c lpr -d emacs-lisp-intro.dvi | |
151 | |
152 @c or else: | |
153 | |
154 @c tex emacs-lisp-intro.texi | |
155 @c texindex emacs-lisp-intro.?? | |
156 @c tex emacs-lisp-intro.texi | |
157 @c lpr -d emacs-lisp-intro.dvi | |
158 | |
159 @c If you include the PostScript figures, and you have old software, | |
160 @c you may need to convert the .dvi file to a .ps file before | |
161 @c printing. Run either of the following command sequences, or one | |
162 @c similar: | |
163 @c | |
164 @c dvips -f < emacs-lisp-intro.dvi > emacs-lisp-intro.ps | |
165 @c | |
166 @c or else: | |
167 @c | |
168 @c postscript -p < emacs-lisp-intro.dvi > emacs-lisp-intro.ps | |
169 @c | |
170 | |
171 @c (Note: if you edit the book so as to change the length of the | |
172 @c table of contents, you may have to change the value of `pageno' below.) | |
173 | |
174 @c ================ End of Formatting Sections ================ | |
175 | |
176 @c For next or subsequent edition: | |
177 @c create function using with-output-to-temp-buffer | |
178 @c create a major mode, with keymaps | |
179 @c run an asynchronous process, like grep or diff | |
180 | |
181 @c For 8.5 by 11 inch format: do not use such a small amount of | |
182 @c whitespace between paragraphs as smallbook format | |
183 @ifset largebook | |
184 @tex | |
185 \global\parskip 6pt plus 1pt | |
186 @end tex | |
187 @end ifset | |
188 | |
189 @c For all sized formats: print within-book cross | |
190 @c reference with ``...'' rather than [...] | |
191 | |
192 @c This works with the texinfo.tex file, version 2003-05-04.08, | |
193 @c in the Texinfo version 4.6 of the 2003 Jun 13 distribution. | |
194 | |
195 @tex | |
196 \if \xrefprintnodename | |
197 \global\def\xrefprintnodename#1{\unskip, ``#1''} | |
198 \else | |
199 \global\def\xrefprintnodename#1{ ``#1''} | |
200 \fi | |
201 % \global\def\xrefprintnodename#1{, ``#1''} | |
202 @end tex | |
203 | |
204 @c ---------------------------------------------------- | |
205 | |
206 @dircategory Emacs | |
207 @direntry | |
208 * Emacs Lisp Intro: (eintr). | |
209 A simple introduction to Emacs Lisp programming. | |
210 @end direntry | |
211 | |
212 @copying | |
213 This is an @cite{Introduction to Programming in Emacs Lisp}, for | |
214 people who are not programmers. | |
215 @sp 1 | |
216 Edition @value{edition-number}, @value{update-date} | |
217 @sp 1 | |
218 Copyright @copyright{} 1990, 1991, 1992, 1993, 1994, 1995, 1997, 2001, | |
219 2002, 2003, 2004, 2005, 2006, 2007 Free Software Foundation, Inc. | |
220 @sp 1 | |
221 | |
222 @iftex | |
223 Published by the:@* | |
224 | |
225 GNU Press, @hfill @uref{http://www.gnupress.org}@* | |
226 a division of the @hfill General: @email{press@@gnu.org}@* | |
227 Free Software Foundation, Inc. @hfill Orders:@w{ } @email{sales@@gnu.org}@* | |
228 51 Franklin Street, Fifth Floor @hfill Tel: +1 (617) 542-5942@* | |
229 Boston, MA 02110-1301 USA @hfill Fax: +1 (617) 542-2652@* | |
230 @end iftex | |
231 | |
232 @ifnottex | |
233 Published by the: | |
234 | |
235 @example | |
236 GNU Press, Website: http://www.gnupress.org | |
237 a division of the General: press@@gnu.org | |
238 Free Software Foundation, Inc. Orders: sales@@gnu.org | |
239 51 Franklin Street, Fifth Floor Tel: +1 (617) 542-5942 | |
240 Boston, MA 02110-1301 USA Fax: +1 (617) 542-2652 | |
241 @end example | |
242 @end ifnottex | |
243 | |
244 @sp 1 | |
245 @c Printed copies are available for $30 each.@* | |
246 ISBN 1-882114-43-4 | |
247 | |
248 Permission is granted to copy, distribute and/or modify this document | |
249 under the terms of the GNU Free Documentation License, Version 1.2 or | |
250 any later version published by the Free Software Foundation; there | |
251 being no Invariant Section, with the Front-Cover Texts being ``A GNU | |
252 Manual'', and with the Back-Cover Texts as in (a) below. A copy of | |
253 the license is included in the section entitled ``GNU Free | |
254 Documentation License''. | |
255 | |
256 (a) The FSF's Back-Cover Text is: ``You have freedom to copy and | |
257 modify this GNU Manual, like GNU software. Copies published by the | |
258 Free Software Foundation raise funds for GNU development.'' | |
259 @end copying | |
260 | |
261 @c half title; two lines here, so do not use `shorttitlepage' | |
262 @tex | |
263 {\begingroup% | |
264 \hbox{}\vskip 1.5in \chaprm \centerline{An Introduction to}% | |
265 \endgroup}% | |
266 {\begingroup\hbox{}\vskip 0.25in \chaprm% | |
267 \centerline{Programming in Emacs Lisp}% | |
268 \endgroup\page\hbox{}\page} | |
269 @end tex | |
270 | |
271 @titlepage | |
272 @sp 6 | |
273 @center @titlefont{An Introduction to} | |
274 @sp 2 | |
275 @center @titlefont{Programming in Emacs Lisp} | |
276 @sp 2 | |
277 @center Revised Third Edition | |
278 @sp 4 | |
279 @center by Robert J. Chassell | |
280 | |
281 @page | |
282 @vskip 0pt plus 1filll | |
283 @insertcopying | |
284 @end titlepage | |
285 | |
286 @iftex | |
287 @headings off | |
288 @evenheading @thispage @| @| @thischapter | |
289 @oddheading @thissection @| @| @thispage | |
290 @end iftex | |
291 | |
292 @ifnothtml | |
293 @c Keep T.O.C. short by tightening up for largebook | |
294 @ifset largebook | |
295 @tex | |
296 \global\parskip 2pt plus 1pt | |
297 \global\advance\baselineskip by -1pt | |
298 @end tex | |
299 @end ifset | |
300 @end ifnothtml | |
301 | |
302 @shortcontents | |
303 @contents | |
304 | |
305 @ifnottex | |
306 @node Top, Preface, (dir), (dir) | |
307 @top An Introduction to Programming in Emacs Lisp | |
308 | |
309 @insertcopying | |
310 | |
311 This master menu first lists each chapter and index; then it lists | |
312 every node in every chapter. | |
313 @end ifnottex | |
314 | |
315 @c >>>> Set pageno appropriately <<<< | |
316 | |
317 @c The first page of the Preface is a roman numeral; it is the first | |
318 @c right handed page after the Table of Contents; hence the following | |
319 @c setting must be for an odd negative number. | |
320 | |
321 @iftex | |
322 @global@pageno = -11 | |
323 @end iftex | |
324 | |
325 @menu | |
326 * Preface:: What to look for. | |
327 * List Processing:: What is Lisp? | |
328 * Practicing Evaluation:: Running several programs. | |
329 * Writing Defuns:: How to write function definitions. | |
330 * Buffer Walk Through:: Exploring a few buffer-related functions. | |
331 * More Complex:: A few, even more complex functions. | |
332 * Narrowing & Widening:: Restricting your and Emacs attention to | |
333 a region. | |
334 * car cdr & cons:: Fundamental functions in Lisp. | |
335 * Cutting & Storing Text:: Removing text and saving it. | |
336 * List Implementation:: How lists are implemented in the computer. | |
337 * Yanking:: Pasting stored text. | |
338 * Loops & Recursion:: How to repeat a process. | |
339 * Regexp Search:: Regular expression searches. | |
340 * Counting Words:: A review of repetition and regexps. | |
341 * Words in a defun:: Counting words in a @code{defun}. | |
342 * Readying a Graph:: A prototype graph printing function. | |
343 * Emacs Initialization:: How to write a @file{.emacs} file. | |
344 * Debugging:: How to run the Emacs Lisp debuggers. | |
345 * Conclusion:: Now you have the basics. | |
346 * the-the:: An appendix: how to find reduplicated words. | |
347 * Kill Ring:: An appendix: how the kill ring works. | |
348 * Full Graph:: How to create a graph with labelled axes. | |
349 * Free Software and Free Manuals:: | |
350 * GNU Free Documentation License:: | |
351 * Index:: | |
352 * About the Author:: | |
353 | |
354 @detailmenu | |
355 --- The Detailed Node Listing --- | |
356 | |
357 Preface | |
358 | |
359 * Why:: Why learn Emacs Lisp? | |
360 * On Reading this Text:: Read, gain familiarity, pick up habits.... | |
361 * Who You Are:: For whom this is written. | |
362 * Lisp History:: | |
363 * Note for Novices:: You can read this as a novice. | |
364 * Thank You:: | |
365 | |
366 List Processing | |
367 | |
368 * Lisp Lists:: What are lists? | |
369 * Run a Program:: Any list in Lisp is a program ready to run. | |
370 * Making Errors:: Generating an error message. | |
371 * Names & Definitions:: Names of symbols and function definitions. | |
372 * Lisp Interpreter:: What the Lisp interpreter does. | |
373 * Evaluation:: Running a program. | |
374 * Variables:: Returning a value from a variable. | |
375 * Arguments:: Passing information to a function. | |
376 * set & setq:: Setting the value of a variable. | |
377 * Summary:: The major points. | |
378 * Error Message Exercises:: | |
379 | |
380 Lisp Lists | |
381 | |
382 * Numbers Lists:: List have numbers, other lists, in them. | |
383 * Lisp Atoms:: Elemental entities. | |
384 * Whitespace in Lists:: Formatting lists to be readable. | |
385 * Typing Lists:: How GNU Emacs helps you type lists. | |
386 | |
387 The Lisp Interpreter | |
388 | |
389 * Complications:: Variables, Special forms, Lists within. | |
390 * Byte Compiling:: Specially processing code for speed. | |
391 | |
392 Evaluation | |
393 | |
394 * How the Interpreter Acts:: Returns and Side Effects... | |
395 * Evaluating Inner Lists:: Lists within lists... | |
396 | |
397 Variables | |
398 | |
399 * fill-column Example:: | |
400 * Void Function:: The error message for a symbol | |
401 without a function. | |
402 * Void Variable:: The error message for a symbol without a value. | |
403 | |
404 Arguments | |
405 | |
406 * Data types:: Types of data passed to a function. | |
407 * Args as Variable or List:: An argument can be the value | |
408 of a variable or list. | |
409 * Variable Number of Arguments:: Some functions may take a | |
410 variable number of arguments. | |
411 * Wrong Type of Argument:: Passing an argument of the wrong type | |
412 to a function. | |
413 * message:: A useful function for sending messages. | |
414 | |
415 Setting the Value of a Variable | |
416 | |
417 * Using set:: Setting values. | |
418 * Using setq:: Setting a quoted value. | |
419 * Counting:: Using @code{setq} to count. | |
420 | |
421 Practicing Evaluation | |
422 | |
423 * How to Evaluate:: Typing editing commands or @kbd{C-x C-e} | |
424 causes evaluation. | |
425 * Buffer Names:: Buffers and files are different. | |
426 * Getting Buffers:: Getting a buffer itself, not merely its name. | |
427 * Switching Buffers:: How to change to another buffer. | |
428 * Buffer Size & Locations:: Where point is located and the size of | |
429 the buffer. | |
430 * Evaluation Exercise:: | |
431 | |
432 How To Write Function Definitions | |
433 | |
434 * Primitive Functions:: | |
435 * defun:: The @code{defun} special form. | |
436 * Install:: Install a function definition. | |
437 * Interactive:: Making a function interactive. | |
438 * Interactive Options:: Different options for @code{interactive}. | |
439 * Permanent Installation:: Installing code permanently. | |
440 * let:: Creating and initializing local variables. | |
441 * if:: What if? | |
442 * else:: If--then--else expressions. | |
443 * Truth & Falsehood:: What Lisp considers false and true. | |
444 * save-excursion:: Keeping track of point, mark, and buffer. | |
445 * Review:: | |
446 * defun Exercises:: | |
447 | |
448 Install a Function Definition | |
449 | |
450 * Effect of installation:: | |
451 * Change a defun:: How to change a function definition. | |
452 | |
453 Make a Function Interactive | |
454 | |
455 * Interactive multiply-by-seven:: An overview. | |
456 * multiply-by-seven in detail:: The interactive version. | |
457 | |
458 @code{let} | |
459 | |
460 * Prevent confusion:: | |
461 * Parts of let Expression:: | |
462 * Sample let Expression:: | |
463 * Uninitialized let Variables:: | |
464 | |
465 The @code{if} Special Form | |
466 | |
467 * if in more detail:: | |
468 * type-of-animal in detail:: An example of an @code{if} expression. | |
469 | |
470 Truth and Falsehood in Emacs Lisp | |
471 | |
472 * nil explained:: @code{nil} has two meanings. | |
473 | |
474 @code{save-excursion} | |
475 | |
476 * Point and mark:: A review of various locations. | |
477 * Template for save-excursion:: | |
478 | |
479 A Few Buffer--Related Functions | |
480 | |
481 * Finding More:: How to find more information. | |
482 * simplified-beginning-of-buffer:: Shows @code{goto-char}, | |
483 @code{point-min}, and @code{push-mark}. | |
484 * mark-whole-buffer:: Almost the same as @code{beginning-of-buffer}. | |
485 * append-to-buffer:: Uses @code{save-excursion} and | |
486 @code{insert-buffer-substring}. | |
487 * Buffer Related Review:: Review. | |
488 * Buffer Exercises:: | |
489 | |
490 The Definition of @code{mark-whole-buffer} | |
491 | |
492 * mark-whole-buffer overview:: | |
493 * Body of mark-whole-buffer:: Only three lines of code. | |
494 | |
495 The Definition of @code{append-to-buffer} | |
496 | |
497 * append-to-buffer overview:: | |
498 * append interactive:: A two part interactive expression. | |
499 * append-to-buffer body:: Incorporates a @code{let} expression. | |
500 * append save-excursion:: How the @code{save-excursion} works. | |
501 | |
502 A Few More Complex Functions | |
503 | |
504 * copy-to-buffer:: With @code{set-buffer}, @code{get-buffer-create}. | |
505 * insert-buffer:: Read-only, and with @code{or}. | |
506 * beginning-of-buffer:: Shows @code{goto-char}, | |
507 @code{point-min}, and @code{push-mark}. | |
508 * Second Buffer Related Review:: | |
509 * optional Exercise:: | |
510 | |
511 The Definition of @code{insert-buffer} | |
512 | |
513 * insert-buffer code:: | |
514 * insert-buffer interactive:: When you can read, but not write. | |
515 * insert-buffer body:: The body has an @code{or} and a @code{let}. | |
516 * if & or:: Using an @code{if} instead of an @code{or}. | |
517 * Insert or:: How the @code{or} expression works. | |
518 * Insert let:: Two @code{save-excursion} expressions. | |
519 * New insert-buffer:: | |
520 | |
521 The Interactive Expression in @code{insert-buffer} | |
522 | |
523 * Read-only buffer:: When a buffer cannot be modified. | |
524 * b for interactive:: An existing buffer or else its name. | |
525 | |
526 Complete Definition of @code{beginning-of-buffer} | |
527 | |
528 * Optional Arguments:: | |
529 * beginning-of-buffer opt arg:: Example with optional argument. | |
530 * beginning-of-buffer complete:: | |
531 | |
532 @code{beginning-of-buffer} with an Argument | |
533 | |
534 * Disentangle beginning-of-buffer:: | |
535 * Large buffer case:: | |
536 * Small buffer case:: | |
537 | |
538 Narrowing and Widening | |
539 | |
540 * Narrowing advantages:: The advantages of narrowing | |
541 * save-restriction:: The @code{save-restriction} special form. | |
542 * what-line:: The number of the line that point is on. | |
543 * narrow Exercise:: | |
544 | |
545 @code{car}, @code{cdr}, @code{cons}: Fundamental Functions | |
546 | |
547 * Strange Names:: An historical aside: why the strange names? | |
548 * car & cdr:: Functions for extracting part of a list. | |
549 * cons:: Constructing a list. | |
550 * nthcdr:: Calling @code{cdr} repeatedly. | |
551 * nth:: | |
552 * setcar:: Changing the first element of a list. | |
553 * setcdr:: Changing the rest of a list. | |
554 * cons Exercise:: | |
555 | |
556 @code{cons} | |
557 | |
558 * Build a list:: | |
559 * length:: How to find the length of a list. | |
560 | |
561 Cutting and Storing Text | |
562 | |
563 * Storing Text:: Text is stored in a list. | |
564 * zap-to-char:: Cutting out text up to a character. | |
565 * kill-region:: Cutting text out of a region. | |
566 * copy-region-as-kill:: A definition for copying text. | |
567 * Digression into C:: Minor note on C programming language macros. | |
568 * defvar:: How to give a variable an initial value. | |
569 * cons & search-fwd Review:: | |
570 * search Exercises:: | |
571 | |
572 @code{zap-to-char} | |
573 | |
574 * Complete zap-to-char:: The complete implementation. | |
575 * zap-to-char interactive:: A three part interactive expression. | |
576 * zap-to-char body:: A short overview. | |
577 * search-forward:: How to search for a string. | |
578 * progn:: The @code{progn} special form. | |
579 * Summing up zap-to-char:: Using @code{point} and @code{search-forward}. | |
580 | |
581 @code{kill-region} | |
582 | |
583 * Complete kill-region:: The function definition. | |
584 * condition-case:: Dealing with a problem. | |
585 * Lisp macro:: | |
586 | |
587 @code{copy-region-as-kill} | |
588 | |
589 * Complete copy-region-as-kill:: The complete function definition. | |
590 * copy-region-as-kill body:: The body of @code{copy-region-as-kill}. | |
591 | |
592 The Body of @code{copy-region-as-kill} | |
593 | |
594 * last-command & this-command:: | |
595 * kill-append function:: | |
596 * kill-new function:: | |
597 | |
598 Initializing a Variable with @code{defvar} | |
599 | |
600 * See variable current value:: | |
601 * defvar and asterisk:: | |
602 | |
603 How Lists are Implemented | |
604 | |
605 * Lists diagrammed:: | |
606 * Symbols as Chest:: Exploring a powerful metaphor. | |
607 * List Exercise:: | |
608 | |
609 Yanking Text Back | |
610 | |
611 * Kill Ring Overview:: | |
612 * kill-ring-yank-pointer:: The kill ring is a list. | |
613 * yank nthcdr Exercises:: The @code{kill-ring-yank-pointer} variable. | |
614 | |
615 Loops and Recursion | |
616 | |
617 * while:: Causing a stretch of code to repeat. | |
618 * dolist dotimes:: | |
619 * Recursion:: Causing a function to call itself. | |
620 * Looping exercise:: | |
621 | |
622 @code{while} | |
623 | |
624 * Looping with while:: Repeat so long as test returns true. | |
625 * Loop Example:: A @code{while} loop that uses a list. | |
626 * print-elements-of-list:: Uses @code{while}, @code{car}, @code{cdr}. | |
627 * Incrementing Loop:: A loop with an incrementing counter. | |
628 * Incrementing Loop Details:: | |
629 * Decrementing Loop:: A loop with a decrementing counter. | |
630 | |
631 Details of an Incrementing Loop | |
632 | |
633 * Incrementing Example:: Counting pebbles in a triangle. | |
634 * Inc Example parts:: The parts of the function definition. | |
635 * Inc Example altogether:: Putting the function definition together. | |
636 | |
637 Loop with a Decrementing Counter | |
638 | |
639 * Decrementing Example:: More pebbles on the beach. | |
640 * Dec Example parts:: The parts of the function definition. | |
641 * Dec Example altogether:: Putting the function definition together. | |
642 | |
643 Save your time: @code{dolist} and @code{dotimes} | |
644 | |
645 * dolist:: | |
646 * dotimes:: | |
647 | |
648 Recursion | |
649 | |
650 * Building Robots:: Same model, different serial number ... | |
651 * Recursive Definition Parts:: Walk until you stop ... | |
652 * Recursion with list:: Using a list as the test whether to recurse. | |
653 * Recursive triangle function:: | |
654 * Recursion with cond:: | |
655 * Recursive Patterns:: Often used templates. | |
656 * No Deferment:: Don't store up work ... | |
657 * No deferment solution:: | |
658 | |
659 Recursion in Place of a Counter | |
660 | |
661 * Recursive Example arg of 1 or 2:: | |
662 * Recursive Example arg of 3 or 4:: | |
663 | |
664 Recursive Patterns | |
665 | |
666 * Every:: | |
667 * Accumulate:: | |
668 * Keep:: | |
669 | |
670 Regular Expression Searches | |
671 | |
672 * sentence-end:: The regular expression for @code{sentence-end}. | |
673 * re-search-forward:: Very similar to @code{search-forward}. | |
674 * forward-sentence:: A straightforward example of regexp search. | |
675 * forward-paragraph:: A somewhat complex example. | |
676 * etags:: How to create your own @file{TAGS} table. | |
677 * Regexp Review:: | |
678 * re-search Exercises:: | |
679 | |
680 @code{forward-sentence} | |
681 | |
682 * Complete forward-sentence:: | |
683 * fwd-sentence while loops:: Two @code{while} loops. | |
684 * fwd-sentence re-search:: A regular expression search. | |
685 | |
686 @code{forward-paragraph}: a Goldmine of Functions | |
687 | |
688 * forward-paragraph in brief:: Key parts of the function definition. | |
689 * fwd-para let:: The @code{let*} expression. | |
690 * fwd-para while:: The forward motion @code{while} loop. | |
691 | |
692 Counting: Repetition and Regexps | |
693 | |
694 * Why Count Words:: | |
695 * count-words-region:: Use a regexp, but find a problem. | |
696 * recursive-count-words:: Start with case of no words in region. | |
697 * Counting Exercise:: | |
698 | |
699 The @code{count-words-region} Function | |
700 | |
701 * Design count-words-region:: The definition using a @code{while} loop. | |
702 * Whitespace Bug:: The Whitespace Bug in @code{count-words-region}. | |
703 | |
704 Counting Words in a @code{defun} | |
705 | |
706 * Divide and Conquer:: | |
707 * Words and Symbols:: What to count? | |
708 * Syntax:: What constitutes a word or symbol? | |
709 * count-words-in-defun:: Very like @code{count-words}. | |
710 * Several defuns:: Counting several defuns in a file. | |
711 * Find a File:: Do you want to look at a file? | |
712 * lengths-list-file:: A list of the lengths of many definitions. | |
713 * Several files:: Counting in definitions in different files. | |
714 * Several files recursively:: Recursively counting in different files. | |
715 * Prepare the data:: Prepare the data for display in a graph. | |
716 | |
717 Count Words in @code{defuns} in Different Files | |
718 | |
719 * lengths-list-many-files:: Return a list of the lengths of defuns. | |
720 * append:: Attach one list to another. | |
721 | |
722 Prepare the Data for Display in a Graph | |
723 | |
724 * Data for Display in Detail:: | |
725 * Sorting:: Sorting lists. | |
726 * Files List:: Making a list of files. | |
727 * Counting function definitions:: | |
728 | |
729 Readying a Graph | |
730 | |
731 * Columns of a graph:: | |
732 * graph-body-print:: How to print the body of a graph. | |
733 * recursive-graph-body-print:: | |
734 * Printed Axes:: | |
735 * Line Graph Exercise:: | |
736 | |
737 Your @file{.emacs} File | |
738 | |
739 * Default Configuration:: | |
740 * Site-wide Init:: You can write site-wide init files. | |
741 * defcustom:: Emacs will write code for you. | |
742 * Beginning a .emacs File:: How to write a @code{.emacs file}. | |
743 * Text and Auto-fill:: Automatically wrap lines. | |
744 * Mail Aliases:: Use abbreviations for email addresses. | |
745 * Indent Tabs Mode:: Don't use tabs with @TeX{} | |
746 * Keybindings:: Create some personal keybindings. | |
747 * Keymaps:: More about key binding. | |
748 * Loading Files:: Load (i.e., evaluate) files automatically. | |
749 * Autoload:: Make functions available. | |
750 * Simple Extension:: Define a function; bind it to a key. | |
751 * X11 Colors:: Colors in X. | |
752 * Miscellaneous:: | |
753 * Mode Line:: How to customize your mode line. | |
754 | |
755 Debugging | |
756 | |
757 * debug:: How to use the built-in debugger. | |
758 * debug-on-entry:: Start debugging when you call a function. | |
759 * debug-on-quit:: Start debugging when you quit with @kbd{C-g}. | |
760 * edebug:: How to use Edebug, a source level debugger. | |
761 * Debugging Exercises:: | |
762 | |
763 Handling the Kill Ring | |
764 | |
765 * What the Kill Ring Does:: | |
766 * current-kill:: | |
767 * yank:: Paste a copy of a clipped element. | |
768 * yank-pop:: Insert element pointed to. | |
769 * ring file:: | |
770 | |
771 The @code{current-kill} Function | |
772 | |
773 * Understanding current-kill:: | |
774 | |
775 @code{current-kill} in Outline | |
776 | |
777 * Body of current-kill:: | |
778 * Digression concerning error:: How to mislead humans, but not computers. | |
779 * Determining the Element:: | |
780 | |
781 A Graph with Labelled Axes | |
782 | |
783 * Labelled Example:: | |
784 * print-graph Varlist:: @code{let} expression in @code{print-graph}. | |
785 * print-Y-axis:: Print a label for the vertical axis. | |
786 * print-X-axis:: Print a horizontal label. | |
787 * Print Whole Graph:: The function to print a complete graph. | |
788 | |
789 The @code{print-Y-axis} Function | |
790 | |
791 * print-Y-axis in Detail:: | |
792 * Height of label:: What height for the Y axis? | |
793 * Compute a Remainder:: How to compute the remainder of a division. | |
794 * Y Axis Element:: Construct a line for the Y axis. | |
795 * Y-axis-column:: Generate a list of Y axis labels. | |
796 * print-Y-axis Penultimate:: A not quite final version. | |
797 | |
798 The @code{print-X-axis} Function | |
799 | |
800 * Similarities differences:: Much like @code{print-Y-axis}, but not exactly. | |
801 * X Axis Tic Marks:: Create tic marks for the horizontal axis. | |
802 | |
803 Printing the Whole Graph | |
804 | |
805 * The final version:: A few changes. | |
806 * Test print-graph:: Run a short test. | |
807 * Graphing words in defuns:: Executing the final code. | |
808 * lambda:: How to write an anonymous function. | |
809 * mapcar:: Apply a function to elements of a list. | |
810 * Another Bug:: Yet another bug @dots{} most insidious. | |
811 * Final printed graph:: The graph itself! | |
812 | |
813 @end detailmenu | |
814 @end menu | |
815 | |
816 @node Preface, List Processing, Top, Top | |
817 @comment node-name, next, previous, up | |
818 @unnumbered Preface | |
819 | |
820 Most of the GNU Emacs integrated environment is written in the programming | |
821 language called Emacs Lisp. The code written in this programming | |
822 language is the software---the sets of instructions---that tell the | |
823 computer what to do when you give it commands. Emacs is designed so | |
824 that you can write new code in Emacs Lisp and easily install it as an | |
825 extension to the editor. | |
826 | |
827 (GNU Emacs is sometimes called an ``extensible editor'', but it does | |
828 much more than provide editing capabilities. It is better to refer to | |
829 Emacs as an ``extensible computing environment''. However, that | |
830 phrase is quite a mouthful. It is easier to refer to Emacs simply as | |
831 an editor. Moreover, everything you do in Emacs---find the Mayan date | |
832 and phases of the moon, simplify polynomials, debug code, manage | |
833 files, read letters, write books---all these activities are kinds of | |
834 editing in the most general sense of the word.) | |
835 | |
836 @menu | |
837 * Why:: Why learn Emacs Lisp? | |
838 * On Reading this Text:: Read, gain familiarity, pick up habits.... | |
839 * Who You Are:: For whom this is written. | |
840 * Lisp History:: | |
841 * Note for Novices:: You can read this as a novice. | |
842 * Thank You:: | |
843 @end menu | |
844 | |
845 @node Why, On Reading this Text, Preface, Preface | |
846 @ifnottex | |
847 @unnumberedsec Why Study Emacs Lisp? | |
848 @end ifnottex | |
849 | |
850 Although Emacs Lisp is usually thought of in association only with Emacs, | |
851 it is a full computer programming language. You can use Emacs Lisp as | |
852 you would any other programming language. | |
853 | |
854 Perhaps you want to understand programming; perhaps you want to extend | |
855 Emacs; or perhaps you want to become a programmer. This introduction to | |
856 Emacs Lisp is designed to get you started: to guide you in learning the | |
857 fundamentals of programming, and more importantly, to show you how you | |
858 can teach yourself to go further. | |
859 | |
860 @node On Reading this Text, Who You Are, Why, Preface | |
861 @comment node-name, next, previous, up | |
862 @unnumberedsec On Reading this Text | |
863 | |
864 All through this document, you will see little sample programs you can | |
865 run inside of Emacs. If you read this document in Info inside of GNU | |
866 Emacs, you can run the programs as they appear. (This is easy to do and | |
867 is explained when the examples are presented.) Alternatively, you can | |
868 read this introduction as a printed book while sitting beside a computer | |
869 running Emacs. (This is what I like to do; I like printed books.) If | |
870 you don't have a running Emacs beside you, you can still read this book, | |
871 but in this case, it is best to treat it as a novel or as a travel guide | |
872 to a country not yet visited: interesting, but not the same as being | |
873 there. | |
874 | |
875 Much of this introduction is dedicated to walk-throughs or guided tours | |
876 of code used in GNU Emacs. These tours are designed for two purposes: | |
877 first, to give you familiarity with real, working code (code you use | |
878 every day); and, second, to give you familiarity with the way Emacs | |
879 works. It is interesting to see how a working environment is | |
880 implemented. | |
881 Also, I | |
882 hope that you will pick up the habit of browsing through source code. | |
883 You can learn from it and mine it for ideas. Having GNU Emacs is like | |
884 having a dragon's cave of treasures. | |
885 | |
886 In addition to learning about Emacs as an editor and Emacs Lisp as a | |
887 programming language, the examples and guided tours will give you an | |
888 opportunity to get acquainted with Emacs as a Lisp programming | |
889 environment. GNU Emacs supports programming and provides tools that | |
890 you will want to become comfortable using, such as @kbd{M-.} (the key | |
891 which invokes the @code{find-tag} command). You will also learn about | |
892 buffers and other objects that are part of the environment. | |
893 Learning about these features of Emacs is like learning new routes | |
894 around your home town. | |
895 | |
896 @ignore | |
897 In addition, I have written several programs as extended examples. | |
898 Although these are examples, the programs are real. I use them. | |
899 Other people use them. You may use them. Beyond the fragments of | |
900 programs used for illustrations, there is very little in here that is | |
901 `just for teaching purposes'; what you see is used. This is a great | |
902 advantage of Emacs Lisp: it is easy to learn to use it for work. | |
903 @end ignore | |
904 | |
905 Finally, I hope to convey some of the skills for using Emacs to | |
906 learn aspects of programming that you don't know. You can often use | |
907 Emacs to help you understand what puzzles you or to find out how to do | |
908 something new. This self-reliance is not only a pleasure, but an | |
909 advantage. | |
910 | |
911 @node Who You Are, Lisp History, On Reading this Text, Preface | |
912 @comment node-name, next, previous, up | |
913 @unnumberedsec For Whom This is Written | |
914 | |
915 This text is written as an elementary introduction for people who are | |
916 not programmers. If you are a programmer, you may not be satisfied with | |
917 this primer. The reason is that you may have become expert at reading | |
918 reference manuals and be put off by the way this text is organized. | |
919 | |
920 An expert programmer who reviewed this text said to me: | |
921 | |
922 @quotation | |
923 @i{I prefer to learn from reference manuals. I ``dive into'' each | |
924 paragraph, and ``come up for air'' between paragraphs.} | |
925 | |
926 @i{When I get to the end of a paragraph, I assume that that subject is | |
927 done, finished, that I know everything I need (with the | |
928 possible exception of the case when the next paragraph starts talking | |
929 about it in more detail). I expect that a well written reference manual | |
930 will not have a lot of redundancy, and that it will have excellent | |
931 pointers to the (one) place where the information I want is.} | |
932 @end quotation | |
933 | |
934 This introduction is not written for this person! | |
935 | |
936 Firstly, I try to say everything at least three times: first, to | |
937 introduce it; second, to show it in context; and third, to show it in a | |
938 different context, or to review it. | |
939 | |
940 Secondly, I hardly ever put all the information about a subject in one | |
941 place, much less in one paragraph. To my way of thinking, that imposes | |
942 too heavy a burden on the reader. Instead I try to explain only what | |
943 you need to know at the time. (Sometimes I include a little extra | |
944 information so you won't be surprised later when the additional | |
945 information is formally introduced.) | |
946 | |
947 When you read this text, you are not expected to learn everything the | |
948 first time. Frequently, you need only make, as it were, a `nodding | |
949 acquaintance' with some of the items mentioned. My hope is that I have | |
950 structured the text and given you enough hints that you will be alert to | |
951 what is important, and concentrate on it. | |
952 | |
953 You will need to ``dive into'' some paragraphs; there is no other way | |
954 to read them. But I have tried to keep down the number of such | |
955 paragraphs. This book is intended as an approachable hill, rather than | |
956 as a daunting mountain. | |
957 | |
958 This introduction to @cite{Programming in Emacs Lisp} has a companion | |
959 document, | |
960 @iftex | |
961 @cite{The GNU Emacs Lisp Reference Manual}. | |
962 @end iftex | |
963 @ifnottex | |
964 @ref{Top, , The GNU Emacs Lisp Reference Manual, elisp, The GNU | |
965 Emacs Lisp Reference Manual}. | |
966 @end ifnottex | |
967 The reference manual has more detail than this introduction. In the | |
968 reference manual, all the information about one topic is concentrated | |
969 in one place. You should turn to it if you are like the programmer | |
970 quoted above. And, of course, after you have read this | |
971 @cite{Introduction}, you will find the @cite{Reference Manual} useful | |
972 when you are writing your own programs. | |
973 | |
974 @node Lisp History, Note for Novices, Who You Are, Preface | |
975 @unnumberedsec Lisp History | |
976 @cindex Lisp history | |
977 | |
978 Lisp was first developed in the late 1950s at the Massachusetts | |
979 Institute of Technology for research in artificial intelligence. The | |
980 great power of the Lisp language makes it superior for other purposes as | |
981 well, such as writing editor commands and integrated environments. | |
982 | |
983 @cindex Maclisp | |
984 @cindex Common Lisp | |
985 GNU Emacs Lisp is largely inspired by Maclisp, which was written at MIT | |
986 in the 1960s. It is somewhat inspired by Common Lisp, which became a | |
987 standard in the 1980s. However, Emacs Lisp is much simpler than Common | |
988 Lisp. (The standard Emacs distribution contains an optional extensions | |
989 file, @file{cl.el}, that adds many Common Lisp features to Emacs Lisp.) | |
990 | |
991 @node Note for Novices, Thank You, Lisp History, Preface | |
992 @comment node-name, next, previous, up | |
993 @unnumberedsec A Note for Novices | |
994 | |
995 If you don't know GNU Emacs, you can still read this document | |
996 profitably. However, I recommend you learn Emacs, if only to learn to | |
997 move around your computer screen. You can teach yourself how to use | |
998 Emacs with the on-line tutorial. To use it, type @kbd{C-h t}. (This | |
999 means you press and release the @key{CTRL} key and the @kbd{h} at the | |
1000 same time, and then press and release @kbd{t}.) | |
1001 | |
1002 Also, I often refer to one of Emacs' standard commands by listing the | |
1003 keys which you press to invoke the command and then giving the name of | |
1004 the command in parentheses, like this: @kbd{M-C-\} | |
1005 (@code{indent-region}). What this means is that the | |
1006 @code{indent-region} command is customarily invoked by typing | |
1007 @kbd{M-C-\}. (You can, if you wish, change the keys that are typed to | |
1008 invoke the command; this is called @dfn{rebinding}. @xref{Keymaps, , | |
1009 Keymaps}.) The abbreviation @kbd{M-C-\} means that you type your | |
1010 @key{META} key, @key{CTRL} key and @key{\} key all at the same time. | |
1011 (On many modern keyboards the @key{META} key is labelled | |
1012 @key{ALT}.) | |
1013 Sometimes a combination like this is called a keychord, since it is | |
1014 similar to the way you play a chord on a piano. If your keyboard does | |
1015 not have a @key{META} key, the @key{ESC} key prefix is used in place | |
1016 of it. In this case, @kbd{M-C-\} means that you press and release your | |
1017 @key{ESC} key and then type the @key{CTRL} key and the @key{\} key at | |
1018 the same time. But usually @kbd{M-C-\} means press the @key{CTRL} key | |
1019 along with the key that is labelled @key{ALT} and, at the same time, | |
1020 press the @key{\} key. | |
1021 | |
1022 In addition to typing a lone keychord, you can prefix what you type | |
1023 with @kbd{C-u}, which is called the `universal argument'. The | |
1024 @kbd{C-u} keychord passes an argument to the subsequent command. | |
1025 Thus, to indent a region of plain text by 6 spaces, mark the region, | |
1026 and then type @w{@kbd{C-u 6 M-C-\}}. (If you do not specify a number, | |
1027 Emacs either passes the number 4 to the command or otherwise runs the | |
1028 command differently than it would otherwise.) @xref{Arguments, , | |
1029 Numeric Arguments, emacs, The GNU Emacs Manual}. | |
1030 | |
1031 If you are reading this in Info using GNU Emacs, you can read through | |
1032 this whole document just by pressing the space bar, @key{SPC}. | |
1033 (To learn about Info, type @kbd{C-h i} and then select Info.) | |
1034 | |
1035 A note on terminology: when I use the word Lisp alone, I often am | |
1036 referring to the various dialects of Lisp in general, but when I speak | |
1037 of Emacs Lisp, I am referring to GNU Emacs Lisp in particular. | |
1038 | |
1039 @node Thank You, , Note for Novices, Preface | |
1040 @comment node-name, next, previous, up | |
1041 @unnumberedsec Thank You | |
1042 | |
1043 My thanks to all who helped me with this book. My especial thanks to | |
1044 @r{Jim Blandy}, @r{Noah Friedman}, @w{Jim Kingdon}, @r{Roland | |
1045 McGrath}, @w{Frank Ritter}, @w{Randy Smith}, @w{Richard M.@: | |
1046 Stallman}, and @w{Melissa Weisshaus}. My thanks also go to both | |
1047 @w{Philip Johnson} and @w{David Stampe} for their patient | |
1048 encouragement. My mistakes are my own. | |
1049 | |
1050 @flushright | |
1051 Robert J. Chassell | |
1052 @end flushright | |
1053 | |
1054 @c ================ Beginning of main text ================ | |
1055 | |
1056 @c Start main text on right-hand (verso) page | |
1057 | |
1058 @tex | |
1059 \par\vfill\supereject | |
1060 \headings off | |
1061 \ifodd\pageno | |
1062 \par\vfill\supereject | |
1063 \else | |
1064 \par\vfill\supereject | |
1065 \page\hbox{}\page | |
1066 \par\vfill\supereject | |
1067 \fi | |
1068 @end tex | |
1069 | |
1070 @iftex | |
1071 @headings off | |
1072 @evenheading @thispage @| @| @thischapter | |
1073 @oddheading @thissection @| @| @thispage | |
1074 @global@pageno = 1 | |
1075 @end iftex | |
1076 | |
1077 @node List Processing, Practicing Evaluation, Preface, Top | |
1078 @comment node-name, next, previous, up | |
1079 @chapter List Processing | |
1080 | |
1081 To the untutored eye, Lisp is a strange programming language. In Lisp | |
1082 code there are parentheses everywhere. Some people even claim that | |
1083 the name stands for `Lots of Isolated Silly Parentheses'. But the | |
1084 claim is unwarranted. Lisp stands for LISt Processing, and the | |
1085 programming language handles @emph{lists} (and lists of lists) by | |
1086 putting them between parentheses. The parentheses mark the boundaries | |
1087 of the list. Sometimes a list is preceded by a single apostrophe or | |
1088 quotation mark, @samp{'}@footnote{The single apostrophe or quotation | |
1089 mark is an abbreviation for the function @code{quote}; you need not | |
1090 think about functions now; functions are defined in @ref{Making | |
1091 Errors, , Generate an Error Message}.} Lists are the basis of Lisp. | |
1092 | |
1093 @menu | |
1094 * Lisp Lists:: What are lists? | |
1095 * Run a Program:: Any list in Lisp is a program ready to run. | |
1096 * Making Errors:: Generating an error message. | |
1097 * Names & Definitions:: Names of symbols and function definitions. | |
1098 * Lisp Interpreter:: What the Lisp interpreter does. | |
1099 * Evaluation:: Running a program. | |
1100 * Variables:: Returning a value from a variable. | |
1101 * Arguments:: Passing information to a function. | |
1102 * set & setq:: Setting the value of a variable. | |
1103 * Summary:: The major points. | |
1104 * Error Message Exercises:: | |
1105 @end menu | |
1106 | |
1107 @node Lisp Lists, Run a Program, List Processing, List Processing | |
1108 @comment node-name, next, previous, up | |
1109 @section Lisp Lists | |
1110 @cindex Lisp Lists | |
1111 | |
1112 In Lisp, a list looks like this: @code{'(rose violet daisy buttercup)}. | |
1113 This list is preceded by a single apostrophe. It could just as well be | |
1114 written as follows, which looks more like the kind of list you are likely | |
1115 to be familiar with: | |
1116 | |
1117 @smallexample | |
1118 @group | |
1119 '(rose | |
1120 violet | |
1121 daisy | |
1122 buttercup) | |
1123 @end group | |
1124 @end smallexample | |
1125 | |
1126 @noindent | |
1127 The elements of this list are the names of the four different flowers, | |
1128 separated from each other by whitespace and surrounded by parentheses, | |
1129 like flowers in a field with a stone wall around them. | |
1130 @cindex Flowers in a field | |
1131 | |
1132 @menu | |
1133 * Numbers Lists:: List have numbers, other lists, in them. | |
1134 * Lisp Atoms:: Elemental entities. | |
1135 * Whitespace in Lists:: Formatting lists to be readable. | |
1136 * Typing Lists:: How GNU Emacs helps you type lists. | |
1137 @end menu | |
1138 | |
1139 @node Numbers Lists, Lisp Atoms, Lisp Lists, Lisp Lists | |
1140 @ifnottex | |
1141 @unnumberedsubsec Numbers, Lists inside of Lists | |
1142 @end ifnottex | |
1143 | |
1144 Lists can also have numbers in them, as in this list: @code{(+ 2 2)}. | |
1145 This list has a plus-sign, @samp{+}, followed by two @samp{2}s, each | |
1146 separated by whitespace. | |
1147 | |
1148 In Lisp, both data and programs are represented the same way; that is, | |
1149 they are both lists of words, numbers, or other lists, separated by | |
1150 whitespace and surrounded by parentheses. (Since a program looks like | |
1151 data, one program may easily serve as data for another; this is a very | |
1152 powerful feature of Lisp.) (Incidentally, these two parenthetical | |
1153 remarks are @emph{not} Lisp lists, because they contain @samp{;} and | |
1154 @samp{.} as punctuation marks.) | |
1155 | |
1156 @need 1200 | |
1157 Here is another list, this time with a list inside of it: | |
1158 | |
1159 @smallexample | |
1160 '(this list has (a list inside of it)) | |
1161 @end smallexample | |
1162 | |
1163 The components of this list are the words @samp{this}, @samp{list}, | |
1164 @samp{has}, and the list @samp{(a list inside of it)}. The interior | |
1165 list is made up of the words @samp{a}, @samp{list}, @samp{inside}, | |
1166 @samp{of}, @samp{it}. | |
1167 | |
1168 @node Lisp Atoms, Whitespace in Lists, Numbers Lists, Lisp Lists | |
1169 @comment node-name, next, previous, up | |
1170 @subsection Lisp Atoms | |
1171 @cindex Lisp Atoms | |
1172 | |
1173 In Lisp, what we have been calling words are called @dfn{atoms}. This | |
1174 term comes from the historical meaning of the word atom, which means | |
1175 `indivisible'. As far as Lisp is concerned, the words we have been | |
1176 using in the lists cannot be divided into any smaller parts and still | |
1177 mean the same thing as part of a program; likewise with numbers and | |
1178 single character symbols like @samp{+}. On the other hand, unlike an | |
1179 ancient atom, a list can be split into parts. (@xref{car cdr & cons, | |
1180 , @code{car} @code{cdr} & @code{cons} Fundamental Functions}.) | |
1181 | |
1182 In a list, atoms are separated from each other by whitespace. They can be | |
1183 right next to a parenthesis. | |
1184 | |
1185 @cindex @samp{empty list} defined | |
1186 Technically speaking, a list in Lisp consists of parentheses surrounding | |
1187 atoms separated by whitespace or surrounding other lists or surrounding | |
1188 both atoms and other lists. A list can have just one atom in it or | |
1189 have nothing in it at all. A list with nothing in it looks like this: | |
1190 @code{()}, and is called the @dfn{empty list}. Unlike anything else, an | |
1191 empty list is considered both an atom and a list at the same time. | |
1192 | |
1193 @cindex Symbolic expressions, introduced | |
1194 @cindex @samp{expression} defined | |
1195 @cindex @samp{form} defined | |
1196 The printed representation of both atoms and lists are called | |
1197 @dfn{symbolic expressions} or, more concisely, @dfn{s-expressions}. | |
1198 The word @dfn{expression} by itself can refer to either the printed | |
1199 representation, or to the atom or list as it is held internally in the | |
1200 computer. Often, people use the term @dfn{expression} | |
1201 indiscriminately. (Also, in many texts, the word @dfn{form} is used | |
1202 as a synonym for expression.) | |
1203 | |
1204 Incidentally, the atoms that make up our universe were named such when | |
1205 they were thought to be indivisible; but it has been found that physical | |
1206 atoms are not indivisible. Parts can split off an atom or it can | |
1207 fission into two parts of roughly equal size. Physical atoms were named | |
1208 prematurely, before their truer nature was found. In Lisp, certain | |
1209 kinds of atom, such as an array, can be separated into parts; but the | |
1210 mechanism for doing this is different from the mechanism for splitting a | |
1211 list. As far as list operations are concerned, the atoms of a list are | |
1212 unsplittable. | |
1213 | |
1214 As in English, the meanings of the component letters of a Lisp atom | |
1215 are different from the meaning the letters make as a word. For | |
1216 example, the word for the South American sloth, the @samp{ai}, is | |
1217 completely different from the two words, @samp{a}, and @samp{i}. | |
1218 | |
1219 There are many kinds of atom in nature but only a few in Lisp: for | |
1220 example, @dfn{numbers}, such as 37, 511, or 1729, and @dfn{symbols}, such | |
1221 as @samp{+}, @samp{foo}, or @samp{forward-line}. The words we have | |
1222 listed in the examples above are all symbols. In everyday Lisp | |
1223 conversation, the word ``atom'' is not often used, because programmers | |
1224 usually try to be more specific about what kind of atom they are dealing | |
1225 with. Lisp programming is mostly about symbols (and sometimes numbers) | |
1226 within lists. (Incidentally, the preceding three word parenthetical | |
1227 remark is a proper list in Lisp, since it consists of atoms, which in | |
1228 this case are symbols, separated by whitespace and enclosed by | |
1229 parentheses, without any non-Lisp punctuation.) | |
1230 | |
1231 @need 1250 | |
1232 In addition, text between double quotation marks---even sentences or | |
1233 paragraphs---is an atom. Here is an example: | |
1234 @cindex Text between double quotation marks | |
1235 | |
1236 @smallexample | |
1237 '(this list includes "text between quotation marks.") | |
1238 @end smallexample | |
1239 | |
1240 @cindex @samp{string} defined | |
1241 @noindent | |
1242 In Lisp, all of the quoted text including the punctuation mark and the | |
1243 blank spaces is a single atom. This kind of atom is called a | |
1244 @dfn{string} (for `string of characters') and is the sort of thing that | |
1245 is used for messages that a computer can print for a human to read. | |
1246 Strings are a different kind of atom than numbers or symbols and are | |
1247 used differently. | |
1248 | |
1249 @node Whitespace in Lists, Typing Lists, Lisp Atoms, Lisp Lists | |
1250 @comment node-name, next, previous, up | |
1251 @subsection Whitespace in Lists | |
1252 @cindex Whitespace in lists | |
1253 | |
1254 @need 1200 | |
1255 The amount of whitespace in a list does not matter. From the point of view | |
1256 of the Lisp language, | |
1257 | |
1258 @smallexample | |
1259 @group | |
1260 '(this list | |
1261 looks like this) | |
1262 @end group | |
1263 @end smallexample | |
1264 | |
1265 @need 800 | |
1266 @noindent | |
1267 is exactly the same as this: | |
1268 | |
1269 @smallexample | |
1270 '(this list looks like this) | |
1271 @end smallexample | |
1272 | |
1273 Both examples show what to Lisp is the same list, the list made up of | |
1274 the symbols @samp{this}, @samp{list}, @samp{looks}, @samp{like}, and | |
1275 @samp{this} in that order. | |
1276 | |
1277 Extra whitespace and newlines are designed to make a list more readable | |
1278 by humans. When Lisp reads the expression, it gets rid of all the extra | |
1279 whitespace (but it needs to have at least one space between atoms in | |
1280 order to tell them apart.) | |
1281 | |
1282 Odd as it seems, the examples we have seen cover almost all of what Lisp | |
1283 lists look like! Every other list in Lisp looks more or less like one | |
1284 of these examples, except that the list may be longer and more complex. | |
1285 In brief, a list is between parentheses, a string is between quotation | |
1286 marks, a symbol looks like a word, and a number looks like a number. | |
1287 (For certain situations, square brackets, dots and a few other special | |
1288 characters may be used; however, we will go quite far without them.) | |
1289 | |
1290 @node Typing Lists, , Whitespace in Lists, Lisp Lists | |
1291 @comment node-name, next, previous, up | |
1292 @subsection GNU Emacs Helps You Type Lists | |
1293 @cindex Help typing lists | |
1294 @cindex Formatting help | |
1295 | |
1296 When you type a Lisp expression in GNU Emacs using either Lisp | |
1297 Interaction mode or Emacs Lisp mode, you have available to you several | |
1298 commands to format the Lisp expression so it is easy to read. For | |
1299 example, pressing the @key{TAB} key automatically indents the line the | |
1300 cursor is on by the right amount. A command to properly indent the | |
1301 code in a region is customarily bound to @kbd{M-C-\}. Indentation is | |
1302 designed so that you can see which elements of a list belong to which | |
1303 list---elements of a sub-list are indented more than the elements of | |
1304 the enclosing list. | |
1305 | |
1306 In addition, when you type a closing parenthesis, Emacs momentarily | |
1307 jumps the cursor back to the matching opening parenthesis, so you can | |
1308 see which one it is. This is very useful, since every list you type | |
1309 in Lisp must have its closing parenthesis match its opening | |
1310 parenthesis. (@xref{Major Modes, , Major Modes, emacs, The GNU Emacs | |
1311 Manual}, for more information about Emacs' modes.) | |
1312 | |
1313 @node Run a Program, Making Errors, Lisp Lists, List Processing | |
1314 @comment node-name, next, previous, up | |
1315 @section Run a Program | |
1316 @cindex Run a program | |
1317 @cindex Program, running one | |
1318 | |
1319 @cindex @samp{evaluate} defined | |
1320 A list in Lisp---any list---is a program ready to run. If you run it | |
1321 (for which the Lisp jargon is @dfn{evaluate}), the computer will do one | |
1322 of three things: do nothing except return to you the list itself; send | |
1323 you an error message; or, treat the first symbol in the list as a | |
1324 command to do something. (Usually, of course, it is the last of these | |
1325 three things that you really want!) | |
1326 | |
1327 @c use code for the single apostrophe, not samp. | |
1328 The single apostrophe, @code{'}, that I put in front of some of the | |
1329 example lists in preceding sections is called a @dfn{quote}; when it | |
1330 precedes a list, it tells Lisp to do nothing with the list, other than | |
1331 take it as it is written. But if there is no quote preceding a list, | |
1332 the first item of the list is special: it is a command for the computer | |
1333 to obey. (In Lisp, these commands are called @emph{functions}.) The list | |
1334 @code{(+ 2 2)} shown above did not have a quote in front of it, so Lisp | |
1335 understands that the @code{+} is an instruction to do something with the | |
1336 rest of the list: add the numbers that follow. | |
1337 | |
1338 @need 1250 | |
1339 If you are reading this inside of GNU Emacs in Info, here is how you can | |
1340 evaluate such a list: place your cursor immediately after the right | |
1341 hand parenthesis of the following list and then type @kbd{C-x C-e}: | |
1342 | |
1343 @smallexample | |
1344 (+ 2 2) | |
1345 @end smallexample | |
1346 | |
1347 @c use code for the number four, not samp. | |
1348 @noindent | |
1349 You will see the number @code{4} appear in the echo area. (In the | |
1350 jargon, what you have just done is ``evaluate the list.'' The echo area | |
1351 is the line at the bottom of the screen that displays or ``echoes'' | |
1352 text.) Now try the same thing with a quoted list: place the cursor | |
1353 right after the following list and type @kbd{C-x C-e}: | |
1354 | |
1355 @smallexample | |
1356 '(this is a quoted list) | |
1357 @end smallexample | |
1358 | |
1359 @noindent | |
1360 You will see @code{(this is a quoted list)} appear in the echo area. | |
1361 | |
1362 @cindex Lisp interpreter, explained | |
1363 @cindex Interpreter, Lisp, explained | |
1364 In both cases, what you are doing is giving a command to the program | |
1365 inside of GNU Emacs called the @dfn{Lisp interpreter}---giving the | |
1366 interpreter a command to evaluate the expression. The name of the Lisp | |
1367 interpreter comes from the word for the task done by a human who comes | |
1368 up with the meaning of an expression---who ``interprets'' it. | |
1369 | |
1370 You can also evaluate an atom that is not part of a list---one that is | |
1371 not surrounded by parentheses; again, the Lisp interpreter translates | |
1372 from the humanly readable expression to the language of the computer. | |
1373 But before discussing this (@pxref{Variables}), we will discuss what the | |
1374 Lisp interpreter does when you make an error. | |
1375 | |
1376 @node Making Errors, Names & Definitions, Run a Program, List Processing | |
1377 @comment node-name, next, previous, up | |
1378 @section Generate an Error Message | |
1379 @cindex Generate an error message | |
1380 @cindex Error message generation | |
1381 | |
1382 Partly so you won't worry if you do it accidentally, we will now give | |
1383 a command to the Lisp interpreter that generates an error message. | |
1384 This is a harmless activity; and indeed, we will often try to generate | |
1385 error messages intentionally. Once you understand the jargon, error | |
1386 messages can be informative. Instead of being called ``error'' | |
1387 messages, they should be called ``help'' messages. They are like | |
1388 signposts to a traveller in a strange country; deciphering them can be | |
1389 hard, but once understood, they can point the way. | |
1390 | |
1391 The error message is generated by a built-in GNU Emacs debugger. We | |
1392 will `enter the debugger'. You get out of the debugger by typing @code{q}. | |
1393 | |
1394 What we will do is evaluate a list that is not quoted and does not | |
1395 have a meaningful command as its first element. Here is a list almost | |
1396 exactly the same as the one we just used, but without the single-quote | |
1397 in front of it. Position the cursor right after it and type @kbd{C-x | |
1398 C-e}: | |
1399 | |
1400 @smallexample | |
1401 (this is an unquoted list) | |
1402 @end smallexample | |
1403 | |
1404 @noindent | |
1405 What you see depends on which version of Emacs you are running. GNU | |
1406 Emacs version 22 provides more information than version 20 and before. | |
1407 First, the more recent result of generating an error; then the | |
1408 earlier, version 20 result. | |
1409 | |
1410 @need 1250 | |
1411 @noindent | |
1412 In GNU Emacs version 22, a @file{*Backtrace*} window will open up and | |
1413 you will see the following in it: | |
1414 | |
1415 @smallexample | |
1416 @group | |
1417 ---------- Buffer: *Backtrace* ---------- | |
1418 Debugger entered--Lisp error: (void-function this) | |
1419 (this is an unquoted list) | |
1420 eval((this is an unquoted list)) | |
1421 eval-last-sexp-1(nil) | |
1422 eval-last-sexp(nil) | |
1423 call-interactively(eval-last-sexp) | |
1424 ---------- Buffer: *Backtrace* ---------- | |
1425 @end group | |
1426 @end smallexample | |
1427 | |
1428 @need 1200 | |
1429 @noindent | |
1430 Your cursor will be in this window (you may have to wait a few seconds | |
1431 before it becomes visible). To quit the debugger and make the | |
1432 debugger window go away, type: | |
1433 | |
1434 @smallexample | |
1435 q | |
1436 @end smallexample | |
1437 | |
1438 @noindent | |
1439 Please type @kbd{q} right now, so you become confident that you can | |
1440 get out of the debugger. Then, type @kbd{C-x C-e} again to re-enter | |
1441 it. | |
1442 | |
1443 @cindex @samp{function} defined | |
1444 Based on what we already know, we can almost read this error message. | |
1445 | |
1446 You read the @file{*Backtrace*} buffer from the bottom up; it tells | |
1447 you what Emacs did. When you typed @kbd{C-x C-e}, you made an | |
1448 interactive call to the command @code{eval-last-sexp}. @code{eval} is | |
1449 an abbreviation for `evaluate' and @code{sexp} is an abbreviation for | |
1450 `symbolic expression'. The command means `evaluate last symbolic | |
1451 expression', which is the expression just before your cursor. | |
1452 | |
1453 Each line above tells you what the Lisp interpreter evaluated next. | |
1454 The most recent action is at the top. The buffer is called the | |
1455 @file{*Backtrace*} buffer because it enables you to track Emacs | |
1456 backwards. | |
1457 | |
1458 @need 800 | |
1459 At the top of the @file{*Backtrace*} buffer, you see the line: | |
1460 | |
1461 @smallexample | |
1462 Debugger entered--Lisp error: (void-function this) | |
1463 @end smallexample | |
1464 | |
1465 @noindent | |
1466 The Lisp interpreter tried to evaluate the first atom of the list, the | |
1467 word @samp{this}. It is this action that generated the error message | |
1468 @samp{void-function this}. | |
1469 | |
1470 The message contains the words @samp{void-function} and @samp{this}. | |
1471 | |
1472 @cindex @samp{function} defined | |
1473 The word @samp{function} was mentioned once before. It is a very | |
1474 important word. For our purposes, we can define it by saying that a | |
1475 @dfn{function} is a set of instructions to the computer that tell the | |
1476 computer to do something. | |
1477 | |
1478 Now we can begin to understand the error message: @samp{void-function | |
1479 this}. The function (that is, the word @samp{this}) does not have a | |
1480 definition of any set of instructions for the computer to carry out. | |
1481 | |
1482 The slightly odd word, @samp{void-function}, is designed to cover the | |
1483 way Emacs Lisp is implemented, which is that when a symbol does not | |
1484 have a function definition attached to it, the place that should | |
1485 contain the instructions is `void'. | |
1486 | |
1487 On the other hand, since we were able to add 2 plus 2 successfully, by | |
1488 evaluating @code{(+ 2 2)}, we can infer that the symbol @code{+} must | |
1489 have a set of instructions for the computer to obey and those | |
1490 instructions must be to add the numbers that follow the @code{+}. | |
1491 | |
1492 @need 1250 | |
1493 In GNU Emacs version 20, and in earlier versions, you will see only | |
1494 one line of error message; it will appear in the echo area and look | |
1495 like this: | |
1496 | |
1497 @smallexample | |
1498 Symbol's function definition is void:@: this | |
1499 @end smallexample | |
1500 | |
1501 @noindent | |
1502 (Also, your terminal may beep at you---some do, some don't; and others | |
1503 blink. This is just a device to get your attention.) The message goes | |
1504 away as soon as you type another key, even just to move the cursor. | |
1505 | |
1506 We know the meaning of the word @samp{Symbol}. It refers to the first | |
1507 atom of the list, the word @samp{this}. The word @samp{function} | |
1508 refers to the instructions that tell the computer what to do. | |
1509 (Technically, the symbol tells the computer where to find the | |
1510 instructions, but this is a complication we can ignore for the | |
1511 moment.) | |
1512 | |
1513 The error message can be understood: @samp{Symbol's function | |
1514 definition is void:@: this}. The symbol (that is, the word | |
1515 @samp{this}) lacks instructions for the computer to carry out. | |
1516 | |
1517 @node Names & Definitions, Lisp Interpreter, Making Errors, List Processing | |
1518 @comment node-name, next, previous, up | |
1519 @section Symbol Names and Function Definitions | |
1520 @cindex Symbol names | |
1521 | |
1522 We can articulate another characteristic of Lisp based on what we have | |
1523 discussed so far---an important characteristic: a symbol, like | |
1524 @code{+}, is not itself the set of instructions for the computer to | |
1525 carry out. Instead, the symbol is used, perhaps temporarily, as a way | |
1526 of locating the definition or set of instructions. What we see is the | |
1527 name through which the instructions can be found. Names of people | |
1528 work the same way. I can be referred to as @samp{Bob}; however, I am | |
1529 not the letters @samp{B}, @samp{o}, @samp{b} but am, or was, the | |
1530 consciousness consistently associated with a particular life-form. | |
1531 The name is not me, but it can be used to refer to me. | |
1532 | |
1533 In Lisp, one set of instructions can be attached to several names. | |
1534 For example, the computer instructions for adding numbers can be | |
1535 linked to the symbol @code{plus} as well as to the symbol @code{+} | |
1536 (and are in some dialects of Lisp). Among humans, I can be referred | |
1537 to as @samp{Robert} as well as @samp{Bob} and by other words as well. | |
1538 | |
1539 On the other hand, a symbol can have only one function definition | |
1540 attached to it at a time. Otherwise, the computer would be confused as | |
1541 to which definition to use. If this were the case among people, only | |
1542 one person in the world could be named @samp{Bob}. However, the function | |
1543 definition to which the name refers can be changed readily. | |
1544 (@xref{Install, , Install a Function Definition}.) | |
1545 | |
1546 Since Emacs Lisp is large, it is customary to name symbols in a way | |
1547 that identifies the part of Emacs to which the function belongs. | |
1548 Thus, all the names for functions that deal with Texinfo start with | |
1549 @samp{texinfo-} and those for functions that deal with reading mail | |
1550 start with @samp{rmail-}. | |
1551 | |
1552 @node Lisp Interpreter, Evaluation, Names & Definitions, List Processing | |
1553 @comment node-name, next, previous, up | |
1554 @section The Lisp Interpreter | |
1555 @cindex Lisp interpreter, what it does | |
1556 @cindex Interpreter, what it does | |
1557 | |
1558 Based on what we have seen, we can now start to figure out what the | |
1559 Lisp interpreter does when we command it to evaluate a list. | |
1560 First, it looks to see whether there is a quote before the list; if | |
1561 there is, the interpreter just gives us the list. On the other | |
1562 hand, if there is no quote, the interpreter looks at the first element | |
1563 in the list and sees whether it has a function definition. If it does, | |
1564 the interpreter carries out the instructions in the function definition. | |
1565 Otherwise, the interpreter prints an error message. | |
1566 | |
1567 This is how Lisp works. Simple. There are added complications which we | |
1568 will get to in a minute, but these are the fundamentals. Of course, to | |
1569 write Lisp programs, you need to know how to write function definitions | |
1570 and attach them to names, and how to do this without confusing either | |
1571 yourself or the computer. | |
1572 | |
1573 @menu | |
1574 * Complications:: Variables, Special forms, Lists within. | |
1575 * Byte Compiling:: Specially processing code for speed. | |
1576 @end menu | |
1577 | |
1578 @node Complications, Byte Compiling, Lisp Interpreter, Lisp Interpreter | |
1579 @ifnottex | |
1580 @unnumberedsubsec Complications | |
1581 @end ifnottex | |
1582 | |
1583 Now, for the first complication. In addition to lists, the Lisp | |
1584 interpreter can evaluate a symbol that is not quoted and does not have | |
1585 parentheses around it. The Lisp interpreter will attempt to determine | |
1586 the symbol's value as a @dfn{variable}. This situation is described | |
1587 in the section on variables. (@xref{Variables}.) | |
1588 | |
1589 @cindex Special form | |
1590 The second complication occurs because some functions are unusual and do | |
1591 not work in the usual manner. Those that don't are called @dfn{special | |
1592 forms}. They are used for special jobs, like defining a function, and | |
1593 there are not many of them. In the next few chapters, you will be | |
1594 introduced to several of the more important special forms. | |
1595 | |
1596 The third and final complication is this: if the function that the | |
1597 Lisp interpreter is looking at is not a special form, and if it is part | |
1598 of a list, the Lisp interpreter looks to see whether the list has a list | |
1599 inside of it. If there is an inner list, the Lisp interpreter first | |
1600 figures out what it should do with the inside list, and then it works on | |
1601 the outside list. If there is yet another list embedded inside the | |
1602 inner list, it works on that one first, and so on. It always works on | |
1603 the innermost list first. The interpreter works on the innermost list | |
1604 first, to evaluate the result of that list. The result may be | |
1605 used by the enclosing expression. | |
1606 | |
1607 Otherwise, the interpreter works left to right, from one expression to | |
1608 the next. | |
1609 | |
1610 @node Byte Compiling, , Complications, Lisp Interpreter | |
1611 @subsection Byte Compiling | |
1612 @cindex Byte compiling | |
1613 | |
1614 One other aspect of interpreting: the Lisp interpreter is able to | |
1615 interpret two kinds of entity: humanly readable code, on which we will | |
1616 focus exclusively, and specially processed code, called @dfn{byte | |
1617 compiled} code, which is not humanly readable. Byte compiled code | |
1618 runs faster than humanly readable code. | |
1619 | |
1620 You can transform humanly readable code into byte compiled code by | |
1621 running one of the compile commands such as @code{byte-compile-file}. | |
1622 Byte compiled code is usually stored in a file that ends with a | |
1623 @file{.elc} extension rather than a @file{.el} extension. You will | |
1624 see both kinds of file in the @file{emacs/lisp} directory; the files | |
1625 to read are those with @file{.el} extensions. | |
1626 | |
1627 As a practical matter, for most things you might do to customize or | |
1628 extend Emacs, you do not need to byte compile; and I will not discuss | |
1629 the topic here. @xref{Byte Compilation, , Byte Compilation, elisp, | |
1630 The GNU Emacs Lisp Reference Manual}, for a full description of byte | |
1631 compilation. | |
1632 | |
1633 @node Evaluation, Variables, Lisp Interpreter, List Processing | |
1634 @comment node-name, next, previous, up | |
1635 @section Evaluation | |
1636 @cindex Evaluation | |
1637 | |
1638 When the Lisp interpreter works on an expression, the term for the | |
1639 activity is called @dfn{evaluation}. We say that the interpreter | |
1640 `evaluates the expression'. I've used this term several times before. | |
1641 The word comes from its use in everyday language, `to ascertain the | |
1642 value or amount of; to appraise', according to @cite{Webster's New | |
1643 Collegiate Dictionary}. | |
1644 | |
1645 @menu | |
1646 * How the Interpreter Acts:: Returns and Side Effects... | |
1647 * Evaluating Inner Lists:: Lists within lists... | |
1648 @end menu | |
1649 | |
1650 @node How the Interpreter Acts, Evaluating Inner Lists, Evaluation, Evaluation | |
1651 @ifnottex | |
1652 @unnumberedsubsec How the Lisp Interpreter Acts | |
1653 @end ifnottex | |
1654 | |
1655 @cindex @samp{returned value} explained | |
1656 After evaluating an expression, the Lisp interpreter will most likely | |
1657 @dfn{return} the value that the computer produces by carrying out the | |
1658 instructions it found in the function definition, or perhaps it will | |
1659 give up on that function and produce an error message. (The interpreter | |
1660 may also find itself tossed, so to speak, to a different function or it | |
1661 may attempt to repeat continually what it is doing for ever and ever in | |
1662 what is called an `infinite loop'. These actions are less common; and | |
1663 we can ignore them.) Most frequently, the interpreter returns a value. | |
1664 | |
1665 @cindex @samp{side effect} defined | |
1666 At the same time the interpreter returns a value, it may do something | |
1667 else as well, such as move a cursor or copy a file; this other kind of | |
1668 action is called a @dfn{side effect}. Actions that we humans think are | |
1669 important, such as printing results, are often ``side effects'' to the | |
1670 Lisp interpreter. The jargon can sound peculiar, but it turns out that | |
1671 it is fairly easy to learn to use side effects. | |
1672 | |
1673 In summary, evaluating a symbolic expression most commonly causes the | |
1674 Lisp interpreter to return a value and perhaps carry out a side effect; | |
1675 or else produce an error. | |
1676 | |
1677 @node Evaluating Inner Lists, , How the Interpreter Acts, Evaluation | |
1678 @comment node-name, next, previous, up | |
1679 @subsection Evaluating Inner Lists | |
1680 @cindex Inner list evaluation | |
1681 @cindex Evaluating inner lists | |
1682 | |
1683 If evaluation applies to a list that is inside another list, the outer | |
1684 list may use the value returned by the first evaluation as information | |
1685 when the outer list is evaluated. This explains why inner expressions | |
1686 are evaluated first: the values they return are used by the outer | |
1687 expressions. | |
1688 | |
1689 @need 1250 | |
1690 We can investigate this process by evaluating another addition example. | |
1691 Place your cursor after the following expression and type @kbd{C-x C-e}: | |
1692 | |
1693 @smallexample | |
1694 (+ 2 (+ 3 3)) | |
1695 @end smallexample | |
1696 | |
1697 @noindent | |
1698 The number 8 will appear in the echo area. | |
1699 | |
1700 What happens is that the Lisp interpreter first evaluates the inner | |
1701 expression, @code{(+ 3 3)}, for which the value 6 is returned; then it | |
1702 evaluates the outer expression as if it were written @code{(+ 2 6)}, which | |
1703 returns the value 8. Since there are no more enclosing expressions to | |
1704 evaluate, the interpreter prints that value in the echo area. | |
1705 | |
1706 Now it is easy to understand the name of the command invoked by the | |
1707 keystrokes @kbd{C-x C-e}: the name is @code{eval-last-sexp}. The | |
1708 letters @code{sexp} are an abbreviation for `symbolic expression', and | |
1709 @code{eval} is an abbreviation for `evaluate'. The command means | |
1710 `evaluate last symbolic expression'. | |
1711 | |
1712 As an experiment, you can try evaluating the expression by putting the | |
1713 cursor at the beginning of the next line immediately following the | |
1714 expression, or inside the expression. | |
1715 | |
1716 @need 800 | |
1717 Here is another copy of the expression: | |
1718 | |
1719 @smallexample | |
1720 (+ 2 (+ 3 3)) | |
1721 @end smallexample | |
1722 | |
1723 @noindent | |
1724 If you place the cursor at the beginning of the blank line that | |
1725 immediately follows the expression and type @kbd{C-x C-e}, you will | |
1726 still get the value 8 printed in the echo area. Now try putting the | |
1727 cursor inside the expression. If you put it right after the next to | |
1728 last parenthesis (so it appears to sit on top of the last parenthesis), | |
1729 you will get a 6 printed in the echo area! This is because the command | |
1730 evaluates the expression @code{(+ 3 3)}. | |
1731 | |
1732 Now put the cursor immediately after a number. Type @kbd{C-x C-e} and | |
1733 you will get the number itself. In Lisp, if you evaluate a number, you | |
1734 get the number itself---this is how numbers differ from symbols. If you | |
1735 evaluate a list starting with a symbol like @code{+}, you will get a | |
1736 value returned that is the result of the computer carrying out the | |
1737 instructions in the function definition attached to that name. If a | |
1738 symbol by itself is evaluated, something different happens, as we will | |
1739 see in the next section. | |
1740 | |
1741 @node Variables, Arguments, Evaluation, List Processing | |
1742 @comment node-name, next, previous, up | |
1743 @section Variables | |
1744 @cindex Variables | |
1745 | |
1746 In Emacs Lisp, a symbol can have a value attached to it just as it can | |
1747 have a function definition attached to it. The two are different. | |
1748 The function definition is a set of instructions that a computer will | |
1749 obey. A value, on the other hand, is something, such as number or a | |
1750 name, that can vary (which is why such a symbol is called a variable). | |
1751 The value of a symbol can be any expression in Lisp, such as a symbol, | |
1752 number, list, or string. A symbol that has a value is often called a | |
1753 @dfn{variable}. | |
1754 | |
1755 A symbol can have both a function definition and a value attached to | |
1756 it at the same time. Or it can have just one or the other. | |
1757 The two are separate. This is somewhat similar | |
1758 to the way the name Cambridge can refer to the city in Massachusetts | |
1759 and have some information attached to the name as well, such as | |
1760 ``great programming center''. | |
1761 | |
1762 @ignore | |
1763 (Incidentally, in Emacs Lisp, a symbol can have two | |
1764 other things attached to it, too: a property list and a documentation | |
1765 string; these are discussed later.) | |
1766 @end ignore | |
1767 | |
1768 Another way to think about this is to imagine a symbol as being a chest | |
1769 of drawers. The function definition is put in one drawer, the value in | |
1770 another, and so on. What is put in the drawer holding the value can be | |
1771 changed without affecting the contents of the drawer holding the | |
1772 function definition, and vice-verse. | |
1773 | |
1774 @menu | |
1775 * fill-column Example:: | |
1776 * Void Function:: The error message for a symbol | |
1777 without a function. | |
1778 * Void Variable:: The error message for a symbol without a value. | |
1779 @end menu | |
1780 | |
1781 @node fill-column Example, Void Function, Variables, Variables | |
1782 @ifnottex | |
1783 @unnumberedsubsec @code{fill-column}, an Example Variable | |
1784 @end ifnottex | |
1785 | |
1786 @findex fill-column, @r{an example variable} | |
1787 @cindex Example variable, @code{fill-column} | |
1788 @cindex Variable, example of, @code{fill-column} | |
1789 The variable @code{fill-column} illustrates a symbol with a value | |
1790 attached to it: in every GNU Emacs buffer, this symbol is set to some | |
1791 value, usually 72 or 70, but sometimes to some other value. To find the | |
1792 value of this symbol, evaluate it by itself. If you are reading this in | |
1793 Info inside of GNU Emacs, you can do this by putting the cursor after | |
1794 the symbol and typing @kbd{C-x C-e}: | |
1795 | |
1796 @smallexample | |
1797 fill-column | |
1798 @end smallexample | |
1799 | |
1800 @noindent | |
1801 After I typed @kbd{C-x C-e}, Emacs printed the number 72 in my echo | |
1802 area. This is the value for which @code{fill-column} is set for me as I | |
1803 write this. It may be different for you in your Info buffer. Notice | |
1804 that the value returned as a variable is printed in exactly the same way | |
1805 as the value returned by a function carrying out its instructions. From | |
1806 the point of view of the Lisp interpreter, a value returned is a value | |
1807 returned. What kind of expression it came from ceases to matter once | |
1808 the value is known. | |
1809 | |
1810 A symbol can have any value attached to it or, to use the jargon, we can | |
1811 @dfn{bind} the variable to a value: to a number, such as 72; to a | |
1812 string, @code{"such as this"}; to a list, such as @code{(spruce pine | |
1813 oak)}; we can even bind a variable to a function definition. | |
1814 | |
1815 A symbol can be bound to a value in several ways. @xref{set & setq, , | |
1816 Setting the Value of a Variable}, for information about one way to do | |
1817 this. | |
1818 | |
1819 @node Void Function, Void Variable, fill-column Example, Variables | |
1820 @comment node-name, next, previous, up | |
1821 @subsection Error Message for a Symbol Without a Function | |
1822 @cindex Symbol without function error | |
1823 @cindex Error for symbol without function | |
1824 | |
1825 When we evaluated @code{fill-column} to find its value as a variable, | |
1826 we did not place parentheses around the word. This is because we did | |
1827 not intend to use it as a function name. | |
1828 | |
1829 If @code{fill-column} were the first or only element of a list, the | |
1830 Lisp interpreter would attempt to find the function definition | |
1831 attached to it. But @code{fill-column} has no function definition. | |
1832 Try evaluating this: | |
1833 | |
1834 @smallexample | |
1835 (fill-column) | |
1836 @end smallexample | |
1837 | |
1838 @need 1250 | |
1839 @noindent | |
1840 In GNU Emacs version 22, you will create a @file{*Backtrace*} buffer | |
1841 that says: | |
1842 | |
1843 @smallexample | |
1844 @group | |
1845 ---------- Buffer: *Backtrace* ---------- | |
1846 Debugger entered--Lisp error: (void-function fill-column) | |
1847 (fill-column) | |
1848 eval((fill-column)) | |
1849 eval-last-sexp-1(nil) | |
1850 eval-last-sexp(nil) | |
1851 call-interactively(eval-last-sexp) | |
1852 ---------- Buffer: *Backtrace* ---------- | |
1853 @end group | |
1854 @end smallexample | |
1855 | |
1856 @noindent | |
1857 (Remember, to quit the debugger and make the debugger window go away, | |
1858 type @kbd{q} in the @file{*Backtrace*} buffer.) | |
1859 | |
1860 @ignore | |
1861 @need 800 | |
1862 In GNU Emacs 20 and before, you will produce an error message that says: | |
1863 | |
1864 @smallexample | |
1865 Symbol's function definition is void:@: fill-column | |
1866 @end smallexample | |
1867 | |
1868 @noindent | |
1869 (The message will go away as soon as you move the cursor or type | |
1870 another key.) | |
1871 @end ignore | |
1872 | |
1873 @node Void Variable, , Void Function, Variables | |
1874 @comment node-name, next, previous, up | |
1875 @subsection Error Message for a Symbol Without a Value | |
1876 @cindex Symbol without value error | |
1877 @cindex Error for symbol without value | |
1878 | |
1879 If you attempt to evaluate a symbol that does not have a value bound to | |
1880 it, you will receive an error message. You can see this by | |
1881 experimenting with our 2 plus 2 addition. In the following expression, | |
1882 put your cursor right after the @code{+}, before the first number 2, | |
1883 type @kbd{C-x C-e}: | |
1884 | |
1885 @smallexample | |
1886 (+ 2 2) | |
1887 @end smallexample | |
1888 | |
1889 @need 1500 | |
1890 @noindent | |
1891 In GNU Emacs 22, you will create a @file{*Backtrace*} buffer that | |
1892 says: | |
1893 | |
1894 @smallexample | |
1895 @group | |
1896 ---------- Buffer: *Backtrace* ---------- | |
1897 Debugger entered--Lisp error: (void-variable +) | |
1898 eval(+) | |
1899 eval-last-sexp-1(nil) | |
1900 eval-last-sexp(nil) | |
1901 call-interactively(eval-last-sexp) | |
1902 ---------- Buffer: *Backtrace* ---------- | |
1903 @end group | |
1904 @end smallexample | |
1905 | |
1906 @noindent | |
1907 (As with the other times we entered the debugger, you can quit by | |
1908 typing @kbd{q} in the @file{*Backtrace*} buffer.) | |
1909 | |
1910 This backtrace is different from the very first error message we saw, | |
1911 which said, @samp{Debugger entered--Lisp error: (void-function this)}. | |
1912 In this case, the function does not have a value as a variable; while | |
1913 in the other error message, the function (the word `this') did not | |
1914 have a definition. | |
1915 | |
1916 In this experiment with the @code{+}, what we did was cause the Lisp | |
1917 interpreter to evaluate the @code{+} and look for the value of the | |
1918 variable instead of the function definition. We did this by placing the | |
1919 cursor right after the symbol rather than after the parenthesis of the | |
1920 enclosing list as we did before. As a consequence, the Lisp interpreter | |
1921 evaluated the preceding s-expression, which in this case was the | |
1922 @code{+} by itself. | |
1923 | |
1924 Since @code{+} does not have a value bound to it, just the function | |
1925 definition, the error message reported that the symbol's value as a | |
1926 variable was void. | |
1927 | |
1928 @ignore | |
1929 @need 800 | |
1930 In GNU Emacs version 20 and before, your error message will say: | |
1931 | |
1932 @example | |
1933 Symbol's value as variable is void:@: + | |
1934 @end example | |
1935 | |
1936 @noindent | |
1937 The meaning is the same as in GNU Emacs 22. | |
1938 @end ignore | |
1939 | |
1940 @node Arguments, set & setq, Variables, List Processing | |
1941 @comment node-name, next, previous, up | |
1942 @section Arguments | |
1943 @cindex Arguments | |
1944 @cindex Passing information to functions | |
1945 | |
1946 To see how information is passed to functions, let's look again at | |
1947 our old standby, the addition of two plus two. In Lisp, this is written | |
1948 as follows: | |
1949 | |
1950 @smallexample | |
1951 (+ 2 2) | |
1952 @end smallexample | |
1953 | |
1954 If you evaluate this expression, the number 4 will appear in your echo | |
1955 area. What the Lisp interpreter does is add the numbers that follow | |
1956 the @code{+}. | |
1957 | |
1958 @cindex @samp{argument} defined | |
1959 The numbers added by @code{+} are called the @dfn{arguments} of the | |
1960 function @code{+}. These numbers are the information that is given to | |
1961 or @dfn{passed} to the function. | |
1962 | |
1963 The word `argument' comes from the way it is used in mathematics and | |
1964 does not refer to a disputation between two people; instead it refers to | |
1965 the information presented to the function, in this case, to the | |
1966 @code{+}. In Lisp, the arguments to a function are the atoms or lists | |
1967 that follow the function. The values returned by the evaluation of | |
1968 these atoms or lists are passed to the function. Different functions | |
1969 require different numbers of arguments; some functions require none at | |
1970 all.@footnote{It is curious to track the path by which the word `argument' | |
1971 came to have two different meanings, one in mathematics and the other in | |
1972 everyday English. According to the @cite{Oxford English Dictionary}, | |
1973 the word derives from the Latin for @samp{to make clear, prove}; thus it | |
1974 came to mean, by one thread of derivation, `the evidence offered as | |
1975 proof', which is to say, `the information offered', which led to its | |
1976 meaning in Lisp. But in the other thread of derivation, it came to mean | |
1977 `to assert in a manner against which others may make counter | |
1978 assertions', which led to the meaning of the word as a disputation. | |
1979 (Note here that the English word has two different definitions attached | |
1980 to it at the same time. By contrast, in Emacs Lisp, a symbol cannot | |
1981 have two different function definitions at the same time.)} | |
1982 | |
1983 @menu | |
1984 * Data types:: Types of data passed to a function. | |
1985 * Args as Variable or List:: An argument can be the value | |
1986 of a variable or list. | |
1987 * Variable Number of Arguments:: Some functions may take a | |
1988 variable number of arguments. | |
1989 * Wrong Type of Argument:: Passing an argument of the wrong type | |
1990 to a function. | |
1991 * message:: A useful function for sending messages. | |
1992 @end menu | |
1993 | |
1994 @node Data types, Args as Variable or List, Arguments, Arguments | |
1995 @comment node-name, next, previous, up | |
1996 @subsection Arguments' Data Types | |
1997 @cindex Data types | |
1998 @cindex Types of data | |
1999 @cindex Arguments' data types | |
2000 | |
2001 The type of data that should be passed to a function depends on what | |
2002 kind of information it uses. The arguments to a function such as | |
2003 @code{+} must have values that are numbers, since @code{+} adds numbers. | |
2004 Other functions use different kinds of data for their arguments. | |
2005 | |
2006 @need 1250 | |
2007 @findex concat | |
2008 For example, the @code{concat} function links together or unites two or | |
2009 more strings of text to produce a string. The arguments are strings. | |
2010 Concatenating the two character strings @code{abc}, @code{def} produces | |
2011 the single string @code{abcdef}. This can be seen by evaluating the | |
2012 following: | |
2013 | |
2014 @smallexample | |
2015 (concat "abc" "def") | |
2016 @end smallexample | |
2017 | |
2018 @noindent | |
2019 The value produced by evaluating this expression is @code{"abcdef"}. | |
2020 | |
2021 A function such as @code{substring} uses both a string and numbers as | |
2022 arguments. The function returns a part of the string, a substring of | |
2023 the first argument. This function takes three arguments. Its first | |
2024 argument is the string of characters, the second and third arguments are | |
2025 numbers that indicate the beginning and end of the substring. The | |
2026 numbers are a count of the number of characters (including spaces and | |
2027 punctuations) from the beginning of the string. | |
2028 | |
2029 @need 800 | |
2030 For example, if you evaluate the following: | |
2031 | |
2032 @smallexample | |
2033 (substring "The quick brown fox jumped." 16 19) | |
2034 @end smallexample | |
2035 | |
2036 @noindent | |
2037 you will see @code{"fox"} appear in the echo area. The arguments are the | |
2038 string and the two numbers. | |
2039 | |
2040 Note that the string passed to @code{substring} is a single atom even | |
2041 though it is made up of several words separated by spaces. Lisp counts | |
2042 everything between the two quotation marks as part of the string, | |
2043 including the spaces. You can think of the @code{substring} function as | |
2044 a kind of `atom smasher' since it takes an otherwise indivisible atom | |
2045 and extracts a part. However, @code{substring} is only able to extract | |
2046 a substring from an argument that is a string, not from another type of | |
2047 atom such as a number or symbol. | |
2048 | |
2049 @node Args as Variable or List, Variable Number of Arguments, Data types, Arguments | |
2050 @comment node-name, next, previous, up | |
2051 @subsection An Argument as the Value of a Variable or List | |
2052 | |
2053 An argument can be a symbol that returns a value when it is evaluated. | |
2054 For example, when the symbol @code{fill-column} by itself is evaluated, | |
2055 it returns a number. This number can be used in an addition. | |
2056 | |
2057 @need 1250 | |
2058 Position the cursor after the following expression and type @kbd{C-x | |
2059 C-e}: | |
2060 | |
2061 @smallexample | |
2062 (+ 2 fill-column) | |
2063 @end smallexample | |
2064 | |
2065 @noindent | |
2066 The value will be a number two more than what you get by evaluating | |
2067 @code{fill-column} alone. For me, this is 74, because my value of | |
2068 @code{fill-column} is 72. | |
2069 | |
2070 As we have just seen, an argument can be a symbol that returns a value | |
2071 when evaluated. In addition, an argument can be a list that returns a | |
2072 value when it is evaluated. For example, in the following expression, | |
2073 the arguments to the function @code{concat} are the strings | |
2074 @w{@code{"The "}} and @w{@code{" red foxes."}} and the list | |
2075 @code{(number-to-string (+ 2 fill-column))}. | |
2076 | |
2077 @c For GNU Emacs 22, need number-to-string | |
2078 @smallexample | |
2079 (concat "The " (number-to-string (+ 2 fill-column)) " red foxes.") | |
2080 @end smallexample | |
2081 | |
2082 @noindent | |
2083 If you evaluate this expression---and if, as with my Emacs, | |
2084 @code{fill-column} evaluates to 72---@code{"The 74 red foxes."} will | |
2085 appear in the echo area. (Note that you must put spaces after the | |
2086 word @samp{The} and before the word @samp{red} so they will appear in | |
2087 the final string. The function @code{number-to-string} converts the | |
2088 integer that the addition function returns to a string. | |
2089 @code{number-to-string} is also known as @code{int-to-string}.) | |
2090 | |
2091 @node Variable Number of Arguments, Wrong Type of Argument, Args as Variable or List, Arguments | |
2092 @comment node-name, next, previous, up | |
2093 @subsection Variable Number of Arguments | |
2094 @cindex Variable number of arguments | |
2095 @cindex Arguments, variable number of | |
2096 | |
2097 Some functions, such as @code{concat}, @code{+} or @code{*}, take any | |
2098 number of arguments. (The @code{*} is the symbol for multiplication.) | |
2099 This can be seen by evaluating each of the following expressions in | |
2100 the usual way. What you will see in the echo area is printed in this | |
2101 text after @samp{@result{}}, which you may read as `evaluates to'. | |
2102 | |
2103 @need 1250 | |
2104 In the first set, the functions have no arguments: | |
2105 | |
2106 @smallexample | |
2107 @group | |
2108 (+) @result{} 0 | |
2109 | |
2110 (*) @result{} 1 | |
2111 @end group | |
2112 @end smallexample | |
2113 | |
2114 @need 1250 | |
2115 In this set, the functions have one argument each: | |
2116 | |
2117 @smallexample | |
2118 @group | |
2119 (+ 3) @result{} 3 | |
2120 | |
2121 (* 3) @result{} 3 | |
2122 @end group | |
2123 @end smallexample | |
2124 | |
2125 @need 1250 | |
2126 In this set, the functions have three arguments each: | |
2127 | |
2128 @smallexample | |
2129 @group | |
2130 (+ 3 4 5) @result{} 12 | |
2131 | |
2132 (* 3 4 5) @result{} 60 | |
2133 @end group | |
2134 @end smallexample | |
2135 | |
2136 @node Wrong Type of Argument, message, Variable Number of Arguments, Arguments | |
2137 @comment node-name, next, previous, up | |
2138 @subsection Using the Wrong Type Object as an Argument | |
2139 @cindex Wrong type of argument | |
2140 @cindex Argument, wrong type of | |
2141 | |
2142 When a function is passed an argument of the wrong type, the Lisp | |
2143 interpreter produces an error message. For example, the @code{+} | |
2144 function expects the values of its arguments to be numbers. As an | |
2145 experiment we can pass it the quoted symbol @code{hello} instead of a | |
2146 number. Position the cursor after the following expression and type | |
2147 @kbd{C-x C-e}: | |
2148 | |
2149 @smallexample | |
2150 (+ 2 'hello) | |
2151 @end smallexample | |
2152 | |
2153 @noindent | |
2154 When you do this you will generate an error message. What has happened | |
2155 is that @code{+} has tried to add the 2 to the value returned by | |
2156 @code{'hello}, but the value returned by @code{'hello} is the symbol | |
2157 @code{hello}, not a number. Only numbers can be added. So @code{+} | |
2158 could not carry out its addition. | |
2159 | |
2160 @need 1250 | |
2161 In GNU Emacs version 22, you will create and enter a | |
2162 @file{*Backtrace*} buffer that says: | |
2163 | |
2164 @noindent | |
2165 @smallexample | |
2166 @group | |
2167 ---------- Buffer: *Backtrace* ---------- | |
2168 Debugger entered--Lisp error: | |
2169 (wrong-type-argument number-or-marker-p hello) | |
2170 +(2 hello) | |
2171 eval((+ 2 (quote hello))) | |
2172 eval-last-sexp-1(nil) | |
2173 eval-last-sexp(nil) | |
2174 call-interactively(eval-last-sexp) | |
2175 ---------- Buffer: *Backtrace* ---------- | |
2176 @end group | |
2177 @end smallexample | |
2178 | |
2179 @need 1250 | |
2180 As usual, the error message tries to be helpful and makes sense after you | |
2181 learn how to read it.@footnote{@code{(quote hello)} is an expansion of | |
2182 the abbreviation @code{'hello}.} | |
2183 | |
2184 The first part of the error message is straightforward; it says | |
2185 @samp{wrong type argument}. Next comes the mysterious jargon word | |
2186 @w{@samp{number-or-marker-p}}. This word is trying to tell you what | |
2187 kind of argument the @code{+} expected. | |
2188 | |
2189 The symbol @code{number-or-marker-p} says that the Lisp interpreter is | |
2190 trying to determine whether the information presented it (the value of | |
2191 the argument) is a number or a marker (a special object representing a | |
2192 buffer position). What it does is test to see whether the @code{+} is | |
2193 being given numbers to add. It also tests to see whether the | |
2194 argument is something called a marker, which is a specific feature of | |
2195 Emacs Lisp. (In Emacs, locations in a buffer are recorded as markers. | |
2196 When the mark is set with the @kbd{C-@@} or @kbd{C-@key{SPC}} command, | |
2197 its position is kept as a marker. The mark can be considered a | |
2198 number---the number of characters the location is from the beginning | |
2199 of the buffer.) In Emacs Lisp, @code{+} can be used to add the | |
2200 numeric value of marker positions as numbers. | |
2201 | |
2202 The @samp{p} of @code{number-or-marker-p} is the embodiment of a | |
2203 practice started in the early days of Lisp programming. The @samp{p} | |
2204 stands for `predicate'. In the jargon used by the early Lisp | |
2205 researchers, a predicate refers to a function to determine whether some | |
2206 property is true or false. So the @samp{p} tells us that | |
2207 @code{number-or-marker-p} is the name of a function that determines | |
2208 whether it is true or false that the argument supplied is a number or | |
2209 a marker. Other Lisp symbols that end in @samp{p} include @code{zerop}, | |
2210 a function that tests whether its argument has the value of zero, and | |
2211 @code{listp}, a function that tests whether its argument is a list. | |
2212 | |
2213 Finally, the last part of the error message is the symbol @code{hello}. | |
2214 This is the value of the argument that was passed to @code{+}. If the | |
2215 addition had been passed the correct type of object, the value passed | |
2216 would have been a number, such as 37, rather than a symbol like | |
2217 @code{hello}. But then you would not have got the error message. | |
2218 | |
2219 @ignore | |
2220 @need 1250 | |
2221 In GNU Emacs version 20 and before, the echo area displays an error | |
2222 message that says: | |
2223 | |
2224 @smallexample | |
2225 Wrong type argument:@: number-or-marker-p, hello | |
2226 @end smallexample | |
2227 | |
2228 This says, in different words, the same as the top line of the | |
2229 @file{*Backtrace*} buffer. | |
2230 @end ignore | |
2231 | |
2232 @node message, , Wrong Type of Argument, Arguments | |
2233 @comment node-name, next, previous, up | |
2234 @subsection The @code{message} Function | |
2235 @findex message | |
2236 | |
2237 Like @code{+}, the @code{message} function takes a variable number of | |
2238 arguments. It is used to send messages to the user and is so useful | |
2239 that we will describe it here. | |
2240 | |
2241 @need 1250 | |
2242 A message is printed in the echo area. For example, you can print a | |
2243 message in your echo area by evaluating the following list: | |
2244 | |
2245 @smallexample | |
2246 (message "This message appears in the echo area!") | |
2247 @end smallexample | |
2248 | |
2249 The whole string between double quotation marks is a single argument | |
2250 and is printed @i{in toto}. (Note that in this example, the message | |
2251 itself will appear in the echo area within double quotes; that is | |
2252 because you see the value returned by the @code{message} function. In | |
2253 most uses of @code{message} in programs that you write, the text will | |
2254 be printed in the echo area as a side-effect, without the quotes. | |
2255 @xref{multiply-by-seven in detail, , @code{multiply-by-seven} in | |
2256 detail}, for an example of this.) | |
2257 | |
2258 However, if there is a @samp{%s} in the quoted string of characters, the | |
2259 @code{message} function does not print the @samp{%s} as such, but looks | |
2260 to the argument that follows the string. It evaluates the second | |
2261 argument and prints the value at the location in the string where the | |
2262 @samp{%s} is. | |
2263 | |
2264 @need 1250 | |
2265 You can see this by positioning the cursor after the following | |
2266 expression and typing @kbd{C-x C-e}: | |
2267 | |
2268 @smallexample | |
2269 (message "The name of this buffer is: %s." (buffer-name)) | |
2270 @end smallexample | |
2271 | |
2272 @noindent | |
2273 In Info, @code{"The name of this buffer is: *info*."} will appear in the | |
2274 echo area. The function @code{buffer-name} returns the name of the | |
2275 buffer as a string, which the @code{message} function inserts in place | |
2276 of @code{%s}. | |
2277 | |
2278 To print a value as an integer, use @samp{%d} in the same way as | |
2279 @samp{%s}. For example, to print a message in the echo area that | |
2280 states the value of the @code{fill-column}, evaluate the following: | |
2281 | |
2282 @smallexample | |
2283 (message "The value of fill-column is %d." fill-column) | |
2284 @end smallexample | |
2285 | |
2286 @noindent | |
2287 On my system, when I evaluate this list, @code{"The value of | |
2288 fill-column is 72."} appears in my echo area@footnote{Actually, you | |
2289 can use @code{%s} to print a number. It is non-specific. @code{%d} | |
2290 prints only the part of a number left of a decimal point, and not | |
2291 anything that is not a number.}. | |
2292 | |
2293 If there is more than one @samp{%s} in the quoted string, the value of | |
2294 the first argument following the quoted string is printed at the | |
2295 location of the first @samp{%s} and the value of the second argument is | |
2296 printed at the location of the second @samp{%s}, and so on. | |
2297 | |
2298 @need 1250 | |
2299 For example, if you evaluate the following, | |
2300 | |
2301 @smallexample | |
2302 @group | |
2303 (message "There are %d %s in the office!" | |
2304 (- fill-column 14) "pink elephants") | |
2305 @end group | |
2306 @end smallexample | |
2307 | |
2308 @noindent | |
2309 a rather whimsical message will appear in your echo area. On my system | |
2310 it says, @code{"There are 58 pink elephants in the office!"}. | |
2311 | |
2312 The expression @code{(- fill-column 14)} is evaluated and the resulting | |
2313 number is inserted in place of the @samp{%d}; and the string in double | |
2314 quotes, @code{"pink elephants"}, is treated as a single argument and | |
2315 inserted in place of the @samp{%s}. (That is to say, a string between | |
2316 double quotes evaluates to itself, like a number.) | |
2317 | |
2318 Finally, here is a somewhat complex example that not only illustrates | |
2319 the computation of a number, but also shows how you can use an | |
2320 expression within an expression to generate the text that is substituted | |
2321 for @samp{%s}: | |
2322 | |
2323 @smallexample | |
2324 @group | |
2325 (message "He saw %d %s" | |
2326 (- fill-column 32) | |
2327 (concat "red " | |
2328 (substring | |
2329 "The quick brown foxes jumped." 16 21) | |
2330 " leaping.")) | |
2331 @end group | |
2332 @end smallexample | |
2333 | |
2334 In this example, @code{message} has three arguments: the string, | |
2335 @code{"He saw %d %s"}, the expression, @code{(- fill-column 32)}, and | |
2336 the expression beginning with the function @code{concat}. The value | |
2337 resulting from the evaluation of @code{(- fill-column 32)} is inserted | |
2338 in place of the @samp{%d}; and the value returned by the expression | |
2339 beginning with @code{concat} is inserted in place of the @samp{%s}. | |
2340 | |
2341 When your fill column is 70 and you evaluate the expression, the | |
2342 message @code{"He saw 38 red foxes leaping."} appears in your echo | |
2343 area. | |
2344 | |
2345 @node set & setq, Summary, Arguments, List Processing | |
2346 @comment node-name, next, previous, up | |
2347 @section Setting the Value of a Variable | |
2348 @cindex Variable, setting value | |
2349 @cindex Setting value of variable | |
2350 | |
2351 @cindex @samp{bind} defined | |
2352 There are several ways by which a variable can be given a value. One of | |
2353 the ways is to use either the function @code{set} or the function | |
2354 @code{setq}. Another way is to use @code{let} (@pxref{let}). (The | |
2355 jargon for this process is to @dfn{bind} a variable to a value.) | |
2356 | |
2357 The following sections not only describe how @code{set} and @code{setq} | |
2358 work but also illustrate how arguments are passed. | |
2359 | |
2360 @menu | |
2361 * Using set:: Setting values. | |
2362 * Using setq:: Setting a quoted value. | |
2363 * Counting:: Using @code{setq} to count. | |
2364 @end menu | |
2365 | |
2366 @node Using set, Using setq, set & setq, set & setq | |
2367 @comment node-name, next, previous, up | |
2368 @subsection Using @code{set} | |
2369 @findex set | |
2370 | |
2371 To set the value of the symbol @code{flowers} to the list @code{'(rose | |
2372 violet daisy buttercup)}, evaluate the following expression by | |
2373 positioning the cursor after the expression and typing @kbd{C-x C-e}. | |
2374 | |
2375 @smallexample | |
2376 (set 'flowers '(rose violet daisy buttercup)) | |
2377 @end smallexample | |
2378 | |
2379 @noindent | |
2380 The list @code{(rose violet daisy buttercup)} will appear in the echo | |
2381 area. This is what is @emph{returned} by the @code{set} function. As a | |
2382 side effect, the symbol @code{flowers} is bound to the list; that is, | |
2383 the symbol @code{flowers}, which can be viewed as a variable, is given | |
2384 the list as its value. (This process, by the way, illustrates how a | |
2385 side effect to the Lisp interpreter, setting the value, can be the | |
2386 primary effect that we humans are interested in. This is because every | |
2387 Lisp function must return a value if it does not get an error, but it | |
2388 will only have a side effect if it is designed to have one.) | |
2389 | |
2390 After evaluating the @code{set} expression, you can evaluate the symbol | |
2391 @code{flowers} and it will return the value you just set. Here is the | |
2392 symbol. Place your cursor after it and type @kbd{C-x C-e}. | |
2393 | |
2394 @smallexample | |
2395 flowers | |
2396 @end smallexample | |
2397 | |
2398 @noindent | |
2399 When you evaluate @code{flowers}, the list | |
2400 @code{(rose violet daisy buttercup)} appears in the echo area. | |
2401 | |
2402 Incidentally, if you evaluate @code{'flowers}, the variable with a quote | |
2403 in front of it, what you will see in the echo area is the symbol itself, | |
2404 @code{flowers}. Here is the quoted symbol, so you can try this: | |
2405 | |
2406 @smallexample | |
2407 'flowers | |
2408 @end smallexample | |
2409 | |
2410 Note also, that when you use @code{set}, you need to quote both | |
2411 arguments to @code{set}, unless you want them evaluated. Since we do | |
2412 not want either argument evaluated, neither the variable | |
2413 @code{flowers} nor the list @code{(rose violet daisy buttercup)}, both | |
2414 are quoted. (When you use @code{set} without quoting its first | |
2415 argument, the first argument is evaluated before anything else is | |
2416 done. If you did this and @code{flowers} did not have a value | |
2417 already, you would get an error message that the @samp{Symbol's value | |
2418 as variable is void}; on the other hand, if @code{flowers} did return | |
2419 a value after it was evaluated, the @code{set} would attempt to set | |
2420 the value that was returned. There are situations where this is the | |
2421 right thing for the function to do; but such situations are rare.) | |
2422 | |
2423 @node Using setq, Counting, Using set, set & setq | |
2424 @comment node-name, next, previous, up | |
2425 @subsection Using @code{setq} | |
2426 @findex setq | |
2427 | |
2428 As a practical matter, you almost always quote the first argument to | |
2429 @code{set}. The combination of @code{set} and a quoted first argument | |
2430 is so common that it has its own name: the special form @code{setq}. | |
2431 This special form is just like @code{set} except that the first argument | |
2432 is quoted automatically, so you don't need to type the quote mark | |
2433 yourself. Also, as an added convenience, @code{setq} permits you to set | |
2434 several different variables to different values, all in one expression. | |
2435 | |
2436 To set the value of the variable @code{carnivores} to the list | |
2437 @code{'(lion tiger leopard)} using @code{setq}, the following expression | |
2438 is used: | |
2439 | |
2440 @smallexample | |
2441 (setq carnivores '(lion tiger leopard)) | |
2442 @end smallexample | |
2443 | |
2444 @noindent | |
2445 This is exactly the same as using @code{set} except the first argument | |
2446 is automatically quoted by @code{setq}. (The @samp{q} in @code{setq} | |
2447 means @code{quote}.) | |
2448 | |
2449 @need 1250 | |
2450 With @code{set}, the expression would look like this: | |
2451 | |
2452 @smallexample | |
2453 (set 'carnivores '(lion tiger leopard)) | |
2454 @end smallexample | |
2455 | |
2456 Also, @code{setq} can be used to assign different values to | |
2457 different variables. The first argument is bound to the value | |
2458 of the second argument, the third argument is bound to the value of the | |
2459 fourth argument, and so on. For example, you could use the following to | |
2460 assign a list of trees to the symbol @code{trees} and a list of herbivores | |
2461 to the symbol @code{herbivores}: | |
2462 | |
2463 @smallexample | |
2464 @group | |
2465 (setq trees '(pine fir oak maple) | |
2466 herbivores '(gazelle antelope zebra)) | |
2467 @end group | |
2468 @end smallexample | |
2469 | |
2470 @noindent | |
2471 (The expression could just as well have been on one line, but it might | |
2472 not have fit on a page; and humans find it easier to read nicely | |
2473 formatted lists.) | |
2474 | |
2475 Although I have been using the term `assign', there is another way of | |
2476 thinking about the workings of @code{set} and @code{setq}; and that is to | |
2477 say that @code{set} and @code{setq} make the symbol @emph{point} to the | |
2478 list. This latter way of thinking is very common and in forthcoming | |
2479 chapters we shall come upon at least one symbol that has `pointer' as | |
2480 part of its name. The name is chosen because the symbol has a value, | |
2481 specifically a list, attached to it; or, expressed another way, | |
2482 the symbol is set to ``point'' to the list. | |
2483 | |
2484 @node Counting, , Using setq, set & setq | |
2485 @comment node-name, next, previous, up | |
2486 @subsection Counting | |
2487 @cindex Counting | |
2488 | |
2489 Here is an example that shows how to use @code{setq} in a counter. You | |
2490 might use this to count how many times a part of your program repeats | |
2491 itself. First set a variable to zero; then add one to the number each | |
2492 time the program repeats itself. To do this, you need a variable that | |
2493 serves as a counter, and two expressions: an initial @code{setq} | |
2494 expression that sets the counter variable to zero; and a second | |
2495 @code{setq} expression that increments the counter each time it is | |
2496 evaluated. | |
2497 | |
2498 @smallexample | |
2499 @group | |
2500 (setq counter 0) ; @r{Let's call this the initializer.} | |
2501 | |
2502 (setq counter (+ counter 1)) ; @r{This is the incrementer.} | |
2503 | |
2504 counter ; @r{This is the counter.} | |
2505 @end group | |
2506 @end smallexample | |
2507 | |
2508 @noindent | |
2509 (The text following the @samp{;} are comments. @xref{Change a | |
2510 defun, , Change a Function Definition}.) | |
2511 | |
2512 If you evaluate the first of these expressions, the initializer, | |
2513 @code{(setq counter 0)}, and then evaluate the third expression, | |
2514 @code{counter}, the number @code{0} will appear in the echo area. If | |
2515 you then evaluate the second expression, the incrementer, @code{(setq | |
2516 counter (+ counter 1))}, the counter will get the value 1. So if you | |
2517 again evaluate @code{counter}, the number @code{1} will appear in the | |
2518 echo area. Each time you evaluate the second expression, the value of | |
2519 the counter will be incremented. | |
2520 | |
2521 When you evaluate the incrementer, @code{(setq counter (+ counter 1))}, | |
2522 the Lisp interpreter first evaluates the innermost list; this is the | |
2523 addition. In order to evaluate this list, it must evaluate the variable | |
2524 @code{counter} and the number @code{1}. When it evaluates the variable | |
2525 @code{counter}, it receives its current value. It passes this value and | |
2526 the number @code{1} to the @code{+} which adds them together. The sum | |
2527 is then returned as the value of the inner list and passed to the | |
2528 @code{setq} which sets the variable @code{counter} to this new value. | |
2529 Thus, the value of the variable, @code{counter}, is changed. | |
2530 | |
2531 @node Summary, Error Message Exercises, set & setq, List Processing | |
2532 @comment node-name, next, previous, up | |
2533 @section Summary | |
2534 | |
2535 Learning Lisp is like climbing a hill in which the first part is the | |
2536 steepest. You have now climbed the most difficult part; what remains | |
2537 becomes easier as you progress onwards. | |
2538 | |
2539 @need 1000 | |
2540 In summary, | |
2541 | |
2542 @itemize @bullet | |
2543 | |
2544 @item | |
2545 Lisp programs are made up of expressions, which are lists or single atoms. | |
2546 | |
2547 @item | |
2548 Lists are made up of zero or more atoms or inner lists, separated by whitespace and | |
2549 surrounded by parentheses. A list can be empty. | |
2550 | |
2551 @item | |
2552 Atoms are multi-character symbols, like @code{forward-paragraph}, single | |
2553 character symbols like @code{+}, strings of characters between double | |
2554 quotation marks, or numbers. | |
2555 | |
2556 @item | |
2557 A number evaluates to itself. | |
2558 | |
2559 @item | |
2560 A string between double quotes also evaluates to itself. | |
2561 | |
2562 @item | |
2563 When you evaluate a symbol by itself, its value is returned. | |
2564 | |
2565 @item | |
2566 When you evaluate a list, the Lisp interpreter looks at the first symbol | |
2567 in the list and then at the function definition bound to that symbol. | |
2568 Then the instructions in the function definition are carried out. | |
2569 | |
2570 @item | |
2571 A single quotation mark, | |
2572 @ifinfo | |
2573 ' | |
2574 @end ifinfo | |
2575 @ifnotinfo | |
2576 @code{'} | |
2577 @end ifnotinfo | |
2578 , tells the Lisp interpreter that it should | |
2579 return the following expression as written, and not evaluate it as it | |
2580 would if the quote were not there. | |
2581 | |
2582 @item | |
2583 Arguments are the information passed to a function. The arguments to a | |
2584 function are computed by evaluating the rest of the elements of the list | |
2585 of which the function is the first element. | |
2586 | |
2587 @item | |
2588 A function always returns a value when it is evaluated (unless it gets | |
2589 an error); in addition, it may also carry out some action called a | |
2590 ``side effect''. In many cases, a function's primary purpose is to | |
2591 create a side effect. | |
2592 @end itemize | |
2593 | |
2594 @node Error Message Exercises, , Summary, List Processing | |
2595 @comment node-name, next, previous, up | |
2596 @section Exercises | |
2597 | |
2598 A few simple exercises: | |
2599 | |
2600 @itemize @bullet | |
2601 @item | |
2602 Generate an error message by evaluating an appropriate symbol that is | |
2603 not within parentheses. | |
2604 | |
2605 @item | |
2606 Generate an error message by evaluating an appropriate symbol that is | |
2607 between parentheses. | |
2608 | |
2609 @item | |
2610 Create a counter that increments by two rather than one. | |
2611 | |
2612 @item | |
2613 Write an expression that prints a message in the echo area when | |
2614 evaluated. | |
2615 @end itemize | |
2616 | |
2617 @node Practicing Evaluation, Writing Defuns, List Processing, Top | |
2618 @comment node-name, next, previous, up | |
2619 @chapter Practicing Evaluation | |
2620 @cindex Practicing evaluation | |
2621 @cindex Evaluation practice | |
2622 | |
2623 Before learning how to write a function definition in Emacs Lisp, it is | |
2624 useful to spend a little time evaluating various expressions that have | |
2625 already been written. These expressions will be lists with the | |
2626 functions as their first (and often only) element. Since some of the | |
2627 functions associated with buffers are both simple and interesting, we | |
2628 will start with those. In this section, we will evaluate a few of | |
2629 these. In another section, we will study the code of several other | |
2630 buffer-related functions, to see how they were written. | |
2631 | |
2632 @menu | |
2633 * How to Evaluate:: Typing editing commands or @kbd{C-x C-e} | |
2634 causes evaluation. | |
2635 * Buffer Names:: Buffers and files are different. | |
2636 * Getting Buffers:: Getting a buffer itself, not merely its name. | |
2637 * Switching Buffers:: How to change to another buffer. | |
2638 * Buffer Size & Locations:: Where point is located and the size of | |
2639 the buffer. | |
2640 * Evaluation Exercise:: | |
2641 @end menu | |
2642 | |
2643 @node How to Evaluate, Buffer Names, Practicing Evaluation, Practicing Evaluation | |
2644 @ifnottex | |
2645 @unnumberedsec How to Evaluate | |
2646 @end ifnottex | |
2647 | |
2648 @i{Whenever you give an editing command} to Emacs Lisp, such as the | |
2649 command to move the cursor or to scroll the screen, @i{you are evaluating | |
2650 an expression,} the first element of which is a function. @i{This is | |
2651 how Emacs works.} | |
2652 | |
2653 @cindex @samp{interactive function} defined | |
2654 @cindex @samp{command} defined | |
2655 When you type keys, you cause the Lisp interpreter to evaluate an | |
2656 expression and that is how you get your results. Even typing plain text | |
2657 involves evaluating an Emacs Lisp function, in this case, one that uses | |
2658 @code{self-insert-command}, which simply inserts the character you | |
2659 typed. The functions you evaluate by typing keystrokes are called | |
2660 @dfn{interactive} functions, or @dfn{commands}; how you make a function | |
2661 interactive will be illustrated in the chapter on how to write function | |
2662 definitions. @xref{Interactive, , Making a Function Interactive}. | |
2663 | |
2664 In addition to typing keyboard commands, we have seen a second way to | |
2665 evaluate an expression: by positioning the cursor after a list and | |
2666 typing @kbd{C-x C-e}. This is what we will do in the rest of this | |
2667 section. There are other ways to evaluate an expression as well; these | |
2668 will be described as we come to them. | |
2669 | |
2670 Besides being used for practicing evaluation, the functions shown in the | |
2671 next few sections are important in their own right. A study of these | |
2672 functions makes clear the distinction between buffers and files, how to | |
2673 switch to a buffer, and how to determine a location within it. | |
2674 | |
2675 @node Buffer Names, Getting Buffers, How to Evaluate, Practicing Evaluation | |
2676 @comment node-name, next, previous, up | |
2677 @section Buffer Names | |
2678 @findex buffer-name | |
2679 @findex buffer-file-name | |
2680 | |
2681 The two functions, @code{buffer-name} and @code{buffer-file-name}, show | |
2682 the difference between a file and a buffer. When you evaluate the | |
2683 following expression, @code{(buffer-name)}, the name of the buffer | |
2684 appears in the echo area. When you evaluate @code{(buffer-file-name)}, | |
2685 the name of the file to which the buffer refers appears in the echo | |
2686 area. Usually, the name returned by @code{(buffer-name)} is the same as | |
2687 the name of the file to which it refers, and the name returned by | |
2688 @code{(buffer-file-name)} is the full path-name of the file. | |
2689 | |
2690 A file and a buffer are two different entities. A file is information | |
2691 recorded permanently in the computer (unless you delete it). A buffer, | |
2692 on the other hand, is information inside of Emacs that will vanish at | |
2693 the end of the editing session (or when you kill the buffer). Usually, | |
2694 a buffer contains information that you have copied from a file; we say | |
2695 the buffer is @dfn{visiting} that file. This copy is what you work on | |
2696 and modify. Changes to the buffer do not change the file, until you | |
2697 save the buffer. When you save the buffer, the buffer is copied to the file | |
2698 and is thus saved permanently. | |
2699 | |
2700 @need 1250 | |
2701 If you are reading this in Info inside of GNU Emacs, you can evaluate | |
2702 each of the following expressions by positioning the cursor after it and | |
2703 typing @kbd{C-x C-e}. | |
2704 | |
2705 @example | |
2706 @group | |
2707 (buffer-name) | |
2708 | |
2709 (buffer-file-name) | |
2710 @end group | |
2711 @end example | |
2712 | |
2713 @noindent | |
2714 When I do this in Info, the value returned by evaluating | |
2715 @code{(buffer-name)} is @file{"*info*"}, and the value returned by | |
2716 evaluating @code{(buffer-file-name)} is @file{nil}. | |
2717 | |
2718 On the other hand, while I am writing this Introduction, the value | |
2719 returned by evaluating @code{(buffer-name)} is | |
2720 @file{"introduction.texinfo"}, and the value returned by evaluating | |
2721 @code{(buffer-file-name)} is | |
2722 @file{"/gnu/work/intro/introduction.texinfo"}. | |
2723 | |
2724 @cindex @code{nil}, history of word | |
2725 The former is the name of the buffer and the latter is the name of the | |
2726 file. In Info, the buffer name is @file{"*info*"}. Info does not | |
2727 point to any file, so the result of evaluating | |
2728 @code{(buffer-file-name)} is @file{nil}. The symbol @code{nil} is | |
2729 from the Latin word for `nothing'; in this case, it means that the | |
2730 buffer is not associated with any file. (In Lisp, @code{nil} is also | |
2731 used to mean `false' and is a synonym for the empty list, @code{()}.) | |
2732 | |
2733 When I am writing, the name of my buffer is | |
2734 @file{"introduction.texinfo"}. The name of the file to which it | |
2735 points is @file{"/gnu/work/intro/introduction.texinfo"}. | |
2736 | |
2737 (In the expressions, the parentheses tell the Lisp interpreter to | |
2738 treat @w{@code{buffer-name}} and @w{@code{buffer-file-name}} as | |
2739 functions; without the parentheses, the interpreter would attempt to | |
2740 evaluate the symbols as variables. @xref{Variables}.) | |
2741 | |
2742 In spite of the distinction between files and buffers, you will often | |
2743 find that people refer to a file when they mean a buffer and vice-verse. | |
2744 Indeed, most people say, ``I am editing a file,'' rather than saying, | |
2745 ``I am editing a buffer which I will soon save to a file.'' It is | |
2746 almost always clear from context what people mean. When dealing with | |
2747 computer programs, however, it is important to keep the distinction in mind, | |
2748 since the computer is not as smart as a person. | |
2749 | |
2750 @cindex Buffer, history of word | |
2751 The word `buffer', by the way, comes from the meaning of the word as a | |
2752 cushion that deadens the force of a collision. In early computers, a | |
2753 buffer cushioned the interaction between files and the computer's | |
2754 central processing unit. The drums or tapes that held a file and the | |
2755 central processing unit were pieces of equipment that were very | |
2756 different from each other, working at their own speeds, in spurts. The | |
2757 buffer made it possible for them to work together effectively. | |
2758 Eventually, the buffer grew from being an intermediary, a temporary | |
2759 holding place, to being the place where work is done. This | |
2760 transformation is rather like that of a small seaport that grew into a | |
2761 great city: once it was merely the place where cargo was warehoused | |
2762 temporarily before being loaded onto ships; then it became a business | |
2763 and cultural center in its own right. | |
2764 | |
2765 Not all buffers are associated with files. For example, a | |
2766 @file{*scratch*} buffer does not visit any file. Similarly, a | |
2767 @file{*Help*} buffer is not associated with any file. | |
2768 | |
2769 In the old days, when you lacked a @file{~/.emacs} file and started an | |
2770 Emacs session by typing the command @code{emacs} alone, without naming | |
2771 any files, Emacs started with the @file{*scratch*} buffer visible. | |
2772 Nowadays, you will see a splash screen. You can follow one of the | |
2773 commands suggested on the splash screen, visit a file, or press the | |
2774 spacebar to reach the @file{*scratch*} buffer. | |
2775 | |
2776 If you switch to the @file{*scratch*} buffer, type | |
2777 @code{(buffer-name)}, position the cursor after it, and then type | |
2778 @kbd{C-x C-e} to evaluate the expression. The name @code{"*scratch*"} | |
2779 will be returned and will appear in the echo area. @code{"*scratch*"} | |
2780 is the name of the buffer. When you type @code{(buffer-file-name)} in | |
2781 the @file{*scratch*} buffer and evaluate that, @code{nil} will appear | |
2782 in the echo area, just as it does when you evaluate | |
2783 @code{(buffer-file-name)} in Info. | |
2784 | |
2785 Incidentally, if you are in the @file{*scratch*} buffer and want the | |
2786 value returned by an expression to appear in the @file{*scratch*} | |
2787 buffer itself rather than in the echo area, type @kbd{C-u C-x C-e} | |
2788 instead of @kbd{C-x C-e}. This causes the value returned to appear | |
2789 after the expression. The buffer will look like this: | |
2790 | |
2791 @smallexample | |
2792 (buffer-name)"*scratch*" | |
2793 @end smallexample | |
2794 | |
2795 @noindent | |
2796 You cannot do this in Info since Info is read-only and it will not allow | |
2797 you to change the contents of the buffer. But you can do this in any | |
2798 buffer you can edit; and when you write code or documentation (such as | |
2799 this book), this feature is very useful. | |
2800 | |
2801 @node Getting Buffers, Switching Buffers, Buffer Names, Practicing Evaluation | |
2802 @comment node-name, next, previous, up | |
2803 @section Getting Buffers | |
2804 @findex current-buffer | |
2805 @findex other-buffer | |
2806 @cindex Getting a buffer | |
2807 | |
2808 The @code{buffer-name} function returns the @emph{name} of the buffer; | |
2809 to get the buffer @emph{itself}, a different function is needed: the | |
2810 @code{current-buffer} function. If you use this function in code, what | |
2811 you get is the buffer itself. | |
2812 | |
2813 A name and the object or entity to which the name refers are different | |
2814 from each other. You are not your name. You are a person to whom | |
2815 others refer by name. If you ask to speak to George and someone hands you | |
2816 a card with the letters @samp{G}, @samp{e}, @samp{o}, @samp{r}, | |
2817 @samp{g}, and @samp{e} written on it, you might be amused, but you would | |
2818 not be satisfied. You do not want to speak to the name, but to the | |
2819 person to whom the name refers. A buffer is similar: the name of the | |
2820 scratch buffer is @file{*scratch*}, but the name is not the buffer. To | |
2821 get a buffer itself, you need to use a function such as | |
2822 @code{current-buffer}. | |
2823 | |
2824 However, there is a slight complication: if you evaluate | |
2825 @code{current-buffer} in an expression on its own, as we will do here, | |
2826 what you see is a printed representation of the name of the buffer | |
2827 without the contents of the buffer. Emacs works this way for two | |
2828 reasons: the buffer may be thousands of lines long---too long to be | |
2829 conveniently displayed; and, another buffer may have the same contents | |
2830 but a different name, and it is important to distinguish between them. | |
2831 | |
2832 @need 800 | |
2833 Here is an expression containing the function: | |
2834 | |
2835 @smallexample | |
2836 (current-buffer) | |
2837 @end smallexample | |
2838 | |
2839 @noindent | |
2840 If you evaluate this expression in Info in Emacs in the usual way, | |
2841 @file{#<buffer *info*>} will appear in the echo area. The special | |
2842 format indicates that the buffer itself is being returned, rather than | |
2843 just its name. | |
2844 | |
2845 Incidentally, while you can type a number or symbol into a program, you | |
2846 cannot do that with the printed representation of a buffer: the only way | |
2847 to get a buffer itself is with a function such as @code{current-buffer}. | |
2848 | |
2849 A related function is @code{other-buffer}. This returns the most | |
2850 recently selected buffer other than the one you are in currently, not | |
2851 a printed representation of its name. If you have recently switched | |
2852 back and forth from the @file{*scratch*} buffer, @code{other-buffer} | |
2853 will return that buffer. | |
2854 | |
2855 @need 800 | |
2856 You can see this by evaluating the expression: | |
2857 | |
2858 @smallexample | |
2859 (other-buffer) | |
2860 @end smallexample | |
2861 | |
2862 @noindent | |
2863 You should see @file{#<buffer *scratch*>} appear in the echo area, or | |
2864 the name of whatever other buffer you switched back from most | |
2865 recently@footnote{Actually, by default, if the buffer from which you | |
2866 just switched is visible to you in another window, @code{other-buffer} | |
2867 will choose the most recent buffer that you cannot see; this is a | |
2868 subtlety that I often forget.}. | |
2869 | |
2870 @node Switching Buffers, Buffer Size & Locations, Getting Buffers, Practicing Evaluation | |
2871 @comment node-name, next, previous, up | |
2872 @section Switching Buffers | |
2873 @findex switch-to-buffer | |
2874 @findex set-buffer | |
2875 @cindex Switching to a buffer | |
2876 | |
2877 The @code{other-buffer} function actually provides a buffer when it is | |
2878 used as an argument to a function that requires one. We can see this | |
2879 by using @code{other-buffer} and @code{switch-to-buffer} to switch to a | |
2880 different buffer. | |
2881 | |
2882 But first, a brief introduction to the @code{switch-to-buffer} | |
2883 function. When you switched back and forth from Info to the | |
2884 @file{*scratch*} buffer to evaluate @code{(buffer-name)}, you most | |
2885 likely typed @kbd{C-x b} and then typed @file{*scratch*}@footnote{Or | |
2886 rather, to save typing, you probably only typed @kbd{RET} if the | |
2887 default buffer was @file{*scratch*}, or if it was different, then you | |
2888 typed just part of the name, such as @code{*sc}, pressed your | |
2889 @kbd{TAB} key to cause it to expand to the full name, and then typed | |
2890 your @kbd{RET} key.} when prompted in the minibuffer for the name of | |
2891 the buffer to which you wanted to switch. The keystrokes, @kbd{C-x | |
2892 b}, cause the Lisp interpreter to evaluate the interactive function | |
2893 @code{switch-to-buffer}. As we said before, this is how Emacs works: | |
2894 different keystrokes call or run different functions. For example, | |
2895 @kbd{C-f} calls @code{forward-char}, @kbd{M-e} calls | |
2896 @code{forward-sentence}, and so on. | |
2897 | |
2898 By writing @code{switch-to-buffer} in an expression, and giving it a | |
2899 buffer to switch to, we can switch buffers just the way @kbd{C-x b} | |
2900 does. | |
2901 | |
2902 @need 1000 | |
2903 Here is the Lisp expression: | |
2904 | |
2905 @smallexample | |
2906 (switch-to-buffer (other-buffer)) | |
2907 @end smallexample | |
2908 | |
2909 @noindent | |
2910 The symbol @code{switch-to-buffer} is the first element of the list, | |
2911 so the Lisp interpreter will treat it as a function and carry out the | |
2912 instructions that are attached to it. But before doing that, the | |
2913 interpreter will note that @code{other-buffer} is inside parentheses | |
2914 and work on that symbol first. @code{other-buffer} is the first (and | |
2915 in this case, the only) element of this list, so the Lisp interpreter | |
2916 calls or runs the function. It returns another buffer. Next, the | |
2917 interpreter runs @code{switch-to-buffer}, passing to it, as an | |
2918 argument, the other buffer, which is what Emacs will switch to. If | |
2919 you are reading this in Info, try this now. Evaluate the expression. | |
2920 (To get back, type @kbd{C-x b @key{RET}}.)@footnote{Remember, this | |
2921 expression will move you to your most recent other buffer that you | |
2922 cannot see. If you really want to go to your most recently selected | |
2923 buffer, even if you can still see it, you need to evaluate the | |
2924 following more complex expression: | |
2925 | |
2926 @smallexample | |
2927 (switch-to-buffer (other-buffer (current-buffer) t)) | |
2928 @end smallexample | |
2929 | |
2930 @c noindent | |
2931 In this case, the first argument to @code{other-buffer} tells it which | |
2932 buffer to skip---the current one---and the second argument tells | |
2933 @code{other-buffer} it is OK to switch to a visible buffer. | |
2934 In regular use, @code{switch-to-buffer} takes you to an invisible | |
2935 window since you would most likely use @kbd{C-x o} (@code{other-window}) | |
2936 to go to another visible buffer.} | |
2937 | |
2938 In the programming examples in later sections of this document, you will | |
2939 see the function @code{set-buffer} more often than | |
2940 @code{switch-to-buffer}. This is because of a difference between | |
2941 computer programs and humans: humans have eyes and expect to see the | |
2942 buffer on which they are working on their computer terminals. This is | |
2943 so obvious, it almost goes without saying. However, programs do not | |
2944 have eyes. When a computer program works on a buffer, that buffer does | |
2945 not need to be visible on the screen. | |
2946 | |
2947 @code{switch-to-buffer} is designed for humans and does two different | |
2948 things: it switches the buffer to which Emacs' attention is directed; and | |
2949 it switches the buffer displayed in the window to the new buffer. | |
2950 @code{set-buffer}, on the other hand, does only one thing: it switches | |
2951 the attention of the computer program to a different buffer. The buffer | |
2952 on the screen remains unchanged (of course, normally nothing happens | |
2953 there until the command finishes running). | |
2954 | |
2955 @cindex @samp{call} defined | |
2956 Also, we have just introduced another jargon term, the word @dfn{call}. | |
2957 When you evaluate a list in which the first symbol is a function, you | |
2958 are calling that function. The use of the term comes from the notion of | |
2959 the function as an entity that can do something for you if you `call' | |
2960 it---just as a plumber is an entity who can fix a leak if you call him | |
2961 or her. | |
2962 | |
2963 @node Buffer Size & Locations, Evaluation Exercise, Switching Buffers, Practicing Evaluation | |
2964 @comment node-name, next, previous, up | |
2965 @section Buffer Size and the Location of Point | |
2966 @cindex Size of buffer | |
2967 @cindex Buffer size | |
2968 @cindex Point location | |
2969 @cindex Location of point | |
2970 | |
2971 Finally, let's look at several rather simple functions, | |
2972 @code{buffer-size}, @code{point}, @code{point-min}, and | |
2973 @code{point-max}. These give information about the size of a buffer and | |
2974 the location of point within it. | |
2975 | |
2976 The function @code{buffer-size} tells you the size of the current | |
2977 buffer; that is, the function returns a count of the number of | |
2978 characters in the buffer. | |
2979 | |
2980 @smallexample | |
2981 (buffer-size) | |
2982 @end smallexample | |
2983 | |
2984 @noindent | |
2985 You can evaluate this in the usual way, by positioning the | |
2986 cursor after the expression and typing @kbd{C-x C-e}. | |
2987 | |
2988 @cindex @samp{point} defined | |
2989 In Emacs, the current position of the cursor is called @dfn{point}. | |
2990 The expression @code{(point)} returns a number that tells you where the | |
2991 cursor is located as a count of the number of characters from the | |
2992 beginning of the buffer up to point. | |
2993 | |
2994 @need 1250 | |
2995 You can see the character count for point in this buffer by evaluating | |
2996 the following expression in the usual way: | |
2997 | |
2998 @smallexample | |
2999 (point) | |
3000 @end smallexample | |
3001 | |
3002 @noindent | |
3003 As I write this, the value of @code{point} is 65724. The @code{point} | |
3004 function is frequently used in some of the examples later in this | |
3005 book. | |
3006 | |
3007 @need 1250 | |
3008 The value of point depends, of course, on its location within the | |
3009 buffer. If you evaluate point in this spot, the number will be larger: | |
3010 | |
3011 @smallexample | |
3012 (point) | |
3013 @end smallexample | |
3014 | |
3015 @noindent | |
3016 For me, the value of point in this location is 66043, which means that | |
3017 there are 319 characters (including spaces) between the two | |
3018 expressions. (Doubtless, you will see different numbers, since I will | |
3019 have edited this since I first evaluated point.) | |
3020 | |
3021 @cindex @samp{narrowing} defined | |
3022 The function @code{point-min} is somewhat similar to @code{point}, but | |
3023 it returns the value of the minimum permissible value of point in the | |
3024 current buffer. This is the number 1 unless @dfn{narrowing} is in | |
3025 effect. (Narrowing is a mechanism whereby you can restrict yourself, | |
3026 or a program, to operations on just a part of a buffer. | |
3027 @xref{Narrowing & Widening, , Narrowing and Widening}.) Likewise, the | |
3028 function @code{point-max} returns the value of the maximum permissible | |
3029 value of point in the current buffer. | |
3030 | |
3031 @node Evaluation Exercise, , Buffer Size & Locations, Practicing Evaluation | |
3032 @section Exercise | |
3033 | |
3034 Find a file with which you are working and move towards its middle. | |
3035 Find its buffer name, file name, length, and your position in the file. | |
3036 | |
3037 @node Writing Defuns, Buffer Walk Through, Practicing Evaluation, Top | |
3038 @comment node-name, next, previous, up | |
3039 @chapter How To Write Function Definitions | |
3040 @cindex Definition writing | |
3041 @cindex Function definition writing | |
3042 @cindex Writing a function definition | |
3043 | |
3044 When the Lisp interpreter evaluates a list, it looks to see whether the | |
3045 first symbol on the list has a function definition attached to it; or, | |
3046 put another way, whether the symbol points to a function definition. If | |
3047 it does, the computer carries out the instructions in the definition. A | |
3048 symbol that has a function definition is called, simply, a function | |
3049 (although, properly speaking, the definition is the function and the | |
3050 symbol refers to it.) | |
3051 | |
3052 @menu | |
3053 * Primitive Functions:: | |
3054 * defun:: The @code{defun} special form. | |
3055 * Install:: Install a function definition. | |
3056 * Interactive:: Making a function interactive. | |
3057 * Interactive Options:: Different options for @code{interactive}. | |
3058 * Permanent Installation:: Installing code permanently. | |
3059 * let:: Creating and initializing local variables. | |
3060 * if:: What if? | |
3061 * else:: If--then--else expressions. | |
3062 * Truth & Falsehood:: What Lisp considers false and true. | |
3063 * save-excursion:: Keeping track of point, mark, and buffer. | |
3064 * Review:: | |
3065 * defun Exercises:: | |
3066 @end menu | |
3067 | |
3068 @node Primitive Functions, defun, Writing Defuns, Writing Defuns | |
3069 @ifnottex | |
3070 @unnumberedsec An Aside about Primitive Functions | |
3071 @end ifnottex | |
3072 @cindex Primitive functions | |
3073 @cindex Functions, primitive | |
3074 | |
3075 @cindex C language primitives | |
3076 @cindex Primitives written in C | |
3077 All functions are defined in terms of other functions, except for a few | |
3078 @dfn{primitive} functions that are written in the C programming | |
3079 language. When you write functions' definitions, you will write them in | |
3080 Emacs Lisp and use other functions as your building blocks. Some of the | |
3081 functions you will use will themselves be written in Emacs Lisp (perhaps | |
3082 by you) and some will be primitives written in C. The primitive | |
3083 functions are used exactly like those written in Emacs Lisp and behave | |
3084 like them. They are written in C so we can easily run GNU Emacs on any | |
3085 computer that has sufficient power and can run C. | |
3086 | |
3087 Let me re-emphasize this: when you write code in Emacs Lisp, you do not | |
3088 distinguish between the use of functions written in C and the use of | |
3089 functions written in Emacs Lisp. The difference is irrelevant. I | |
3090 mention the distinction only because it is interesting to know. Indeed, | |
3091 unless you investigate, you won't know whether an already-written | |
3092 function is written in Emacs Lisp or C. | |
3093 | |
3094 @node defun, Install, Primitive Functions, Writing Defuns | |
3095 @comment node-name, next, previous, up | |
3096 @section The @code{defun} Special Form | |
3097 @findex defun | |
3098 @cindex Special form of @code{defun} | |
3099 | |
3100 @cindex @samp{function definition} defined | |
3101 In Lisp, a symbol such as @code{mark-whole-buffer} has code attached to | |
3102 it that tells the computer what to do when the function is called. | |
3103 This code is called the @dfn{function definition} and is created by | |
3104 evaluating a Lisp expression that starts with the symbol @code{defun} | |
3105 (which is an abbreviation for @emph{define function}). Because | |
3106 @code{defun} does not evaluate its arguments in the usual way, it is | |
3107 called a @dfn{special form}. | |
3108 | |
3109 In subsequent sections, we will look at function definitions from the | |
3110 Emacs source code, such as @code{mark-whole-buffer}. In this section, | |
3111 we will describe a simple function definition so you can see how it | |
3112 looks. This function definition uses arithmetic because it makes for a | |
3113 simple example. Some people dislike examples using arithmetic; however, | |
3114 if you are such a person, do not despair. Hardly any of the code we | |
3115 will study in the remainder of this introduction involves arithmetic or | |
3116 mathematics. The examples mostly involve text in one way or another. | |
3117 | |
3118 A function definition has up to five parts following the word | |
3119 @code{defun}: | |
3120 | |
3121 @enumerate | |
3122 @item | |
3123 The name of the symbol to which the function definition should be | |
3124 attached. | |
3125 | |
3126 @item | |
3127 A list of the arguments that will be passed to the function. If no | |
3128 arguments will be passed to the function, this is an empty list, | |
3129 @code{()}. | |
3130 | |
3131 @item | |
3132 Documentation describing the function. (Technically optional, but | |
3133 strongly recommended.) | |
3134 | |
3135 @item | |
3136 Optionally, an expression to make the function interactive so you can | |
3137 use it by typing @kbd{M-x} and then the name of the function; or by | |
3138 typing an appropriate key or keychord. | |
3139 | |
3140 @cindex @samp{body} defined | |
3141 @item | |
3142 The code that instructs the computer what to do: the @dfn{body} of the | |
3143 function definition. | |
3144 @end enumerate | |
3145 | |
3146 It is helpful to think of the five parts of a function definition as | |
3147 being organized in a template, with slots for each part: | |
3148 | |
3149 @smallexample | |
3150 @group | |
3151 (defun @var{function-name} (@var{arguments}@dots{}) | |
3152 "@var{optional-documentation}@dots{}" | |
3153 (interactive @var{argument-passing-info}) ; @r{optional} | |
3154 @var{body}@dots{}) | |
3155 @end group | |
3156 @end smallexample | |
3157 | |
3158 As an example, here is the code for a function that multiplies its | |
3159 argument by 7. (This example is not interactive. @xref{Interactive, | |
3160 , Making a Function Interactive}, for that information.) | |
3161 | |
3162 @smallexample | |
3163 @group | |
3164 (defun multiply-by-seven (number) | |
3165 "Multiply NUMBER by seven." | |
3166 (* 7 number)) | |
3167 @end group | |
3168 @end smallexample | |
3169 | |
3170 This definition begins with a parenthesis and the symbol @code{defun}, | |
3171 followed by the name of the function. | |
3172 | |
3173 @cindex @samp{argument list} defined | |
3174 The name of the function is followed by a list that contains the | |
3175 arguments that will be passed to the function. This list is called | |
3176 the @dfn{argument list}. In this example, the list has only one | |
3177 element, the symbol, @code{number}. When the function is used, the | |
3178 symbol will be bound to the value that is used as the argument to the | |
3179 function. | |
3180 | |
3181 Instead of choosing the word @code{number} for the name of the argument, | |
3182 I could have picked any other name. For example, I could have chosen | |
3183 the word @code{multiplicand}. I picked the word `number' because it | |
3184 tells what kind of value is intended for this slot; but I could just as | |
3185 well have chosen the word `multiplicand' to indicate the role that the | |
3186 value placed in this slot will play in the workings of the function. I | |
3187 could have called it @code{foogle}, but that would have been a bad | |
3188 choice because it would not tell humans what it means. The choice of | |
3189 name is up to the programmer and should be chosen to make the meaning of | |
3190 the function clear. | |
3191 | |
3192 Indeed, you can choose any name you wish for a symbol in an argument | |
3193 list, even the name of a symbol used in some other function: the name | |
3194 you use in an argument list is private to that particular definition. | |
3195 In that definition, the name refers to a different entity than any use | |
3196 of the same name outside the function definition. Suppose you have a | |
3197 nick-name `Shorty' in your family; when your family members refer to | |
3198 `Shorty', they mean you. But outside your family, in a movie, for | |
3199 example, the name `Shorty' refers to someone else. Because a name in an | |
3200 argument list is private to the function definition, you can change the | |
3201 value of such a symbol inside the body of a function without changing | |
3202 its value outside the function. The effect is similar to that produced | |
3203 by a @code{let} expression. (@xref{let, , @code{let}}.) | |
3204 | |
3205 @ignore | |
3206 Note also that we discuss the word `number' in two different ways: as a | |
3207 symbol that appears in the code, and as the name of something that will | |
3208 be replaced by a something else during the evaluation of the function. | |
3209 In the first case, @code{number} is a symbol, not a number; it happens | |
3210 that within the function, it is a variable who value is the number in | |
3211 question, but our primary interest in it is as a symbol. On the other | |
3212 hand, when we are talking about the function, our interest is that we | |
3213 will substitute a number for the word @var{number}. To keep this | |
3214 distinction clear, we use different typography for the two | |
3215 circumstances. When we talk about this function, or about how it works, | |
3216 we refer to this number by writing @var{number}. In the function | |
3217 itself, we refer to it by writing @code{number}. | |
3218 @end ignore | |
3219 | |
3220 The argument list is followed by the documentation string that | |
3221 describes the function. This is what you see when you type | |
3222 @w{@kbd{C-h f}} and the name of a function. Incidentally, when you | |
3223 write a documentation string like this, you should make the first line | |
3224 a complete sentence since some commands, such as @code{apropos}, print | |
3225 only the first line of a multi-line documentation string. Also, you | |
3226 should not indent the second line of a documentation string, if you | |
3227 have one, because that looks odd when you use @kbd{C-h f} | |
3228 (@code{describe-function}). The documentation string is optional, but | |
3229 it is so useful, it should be included in almost every function you | |
3230 write. | |
3231 | |
3232 @findex * @r{(multiplication)} | |
3233 The third line of the example consists of the body of the function | |
3234 definition. (Most functions' definitions, of course, are longer than | |
3235 this.) In this function, the body is the list, @code{(* 7 number)}, which | |
3236 says to multiply the value of @var{number} by 7. (In Emacs Lisp, | |
3237 @code{*} is the function for multiplication, just as @code{+} is the | |
3238 function for addition.) | |
3239 | |
3240 When you use the @code{multiply-by-seven} function, the argument | |
3241 @code{number} evaluates to the actual number you want used. Here is an | |
3242 example that shows how @code{multiply-by-seven} is used; but don't try | |
3243 to evaluate this yet! | |
3244 | |
3245 @smallexample | |
3246 (multiply-by-seven 3) | |
3247 @end smallexample | |
3248 | |
3249 @noindent | |
3250 The symbol @code{number}, specified in the function definition in the | |
3251 next section, is given or ``bound to'' the value 3 in the actual use of | |
3252 the function. Note that although @code{number} was inside parentheses | |
3253 in the function definition, the argument passed to the | |
3254 @code{multiply-by-seven} function is not in parentheses. The | |
3255 parentheses are written in the function definition so the computer can | |
3256 figure out where the argument list ends and the rest of the function | |
3257 definition begins. | |
3258 | |
3259 If you evaluate this example, you are likely to get an error message. | |
3260 (Go ahead, try it!) This is because we have written the function | |
3261 definition, but not yet told the computer about the definition---we have | |
3262 not yet installed (or `loaded') the function definition in Emacs. | |
3263 Installing a function is the process that tells the Lisp interpreter the | |
3264 definition of the function. Installation is described in the next | |
3265 section. | |
3266 | |
3267 @node Install, Interactive, defun, Writing Defuns | |
3268 @comment node-name, next, previous, up | |
3269 @section Install a Function Definition | |
3270 @cindex Install a Function Definition | |
3271 @cindex Definition installation | |
3272 @cindex Function definition installation | |
3273 | |
3274 If you are reading this inside of Info in Emacs, you can try out the | |
3275 @code{multiply-by-seven} function by first evaluating the function | |
3276 definition and then evaluating @code{(multiply-by-seven 3)}. A copy of | |
3277 the function definition follows. Place the cursor after the last | |
3278 parenthesis of the function definition and type @kbd{C-x C-e}. When you | |
3279 do this, @code{multiply-by-seven} will appear in the echo area. (What | |
3280 this means is that when a function definition is evaluated, the value it | |
3281 returns is the name of the defined function.) At the same time, this | |
3282 action installs the function definition. | |
3283 | |
3284 @smallexample | |
3285 @group | |
3286 (defun multiply-by-seven (number) | |
3287 "Multiply NUMBER by seven." | |
3288 (* 7 number)) | |
3289 @end group | |
3290 @end smallexample | |
3291 | |
3292 @noindent | |
3293 By evaluating this @code{defun}, you have just installed | |
3294 @code{multiply-by-seven} in Emacs. The function is now just as much a | |
3295 part of Emacs as @code{forward-word} or any other editing function you | |
3296 use. (@code{multiply-by-seven} will stay installed until you quit | |
3297 Emacs. To reload code automatically whenever you start Emacs, see | |
3298 @ref{Permanent Installation, , Installing Code Permanently}.) | |
3299 | |
3300 @menu | |
3301 * Effect of installation:: | |
3302 * Change a defun:: How to change a function definition. | |
3303 @end menu | |
3304 | |
3305 @node Effect of installation, Change a defun, Install, Install | |
3306 @ifnottex | |
3307 @unnumberedsubsec The effect of installation | |
3308 @end ifnottex | |
3309 | |
3310 You can see the effect of installing @code{multiply-by-seven} by | |
3311 evaluating the following sample. Place the cursor after the following | |
3312 expression and type @kbd{C-x C-e}. The number 21 will appear in the | |
3313 echo area. | |
3314 | |
3315 @smallexample | |
3316 (multiply-by-seven 3) | |
3317 @end smallexample | |
3318 | |
3319 If you wish, you can read the documentation for the function by typing | |
3320 @kbd{C-h f} (@code{describe-function}) and then the name of the | |
3321 function, @code{multiply-by-seven}. When you do this, a | |
3322 @file{*Help*} window will appear on your screen that says: | |
3323 | |
3324 @smallexample | |
3325 @group | |
3326 multiply-by-seven is a Lisp function. | |
3327 (multiply-by-seven NUMBER) | |
3328 | |
3329 Multiply NUMBER by seven. | |
3330 @end group | |
3331 @end smallexample | |
3332 | |
3333 @noindent | |
3334 (To return to a single window on your screen, type @kbd{C-x 1}.) | |
3335 | |
3336 @node Change a defun, , Effect of installation, Install | |
3337 @comment node-name, next, previous, up | |
3338 @subsection Change a Function Definition | |
3339 @cindex Changing a function definition | |
3340 @cindex Function definition, how to change | |
3341 @cindex Definition, how to change | |
3342 | |
3343 If you want to change the code in @code{multiply-by-seven}, just rewrite | |
3344 it. To install the new version in place of the old one, evaluate the | |
3345 function definition again. This is how you modify code in Emacs. It is | |
3346 very simple. | |
3347 | |
3348 As an example, you can change the @code{multiply-by-seven} function to | |
3349 add the number to itself seven times instead of multiplying the number | |
3350 by seven. It produces the same answer, but by a different path. At | |
3351 the same time, we will add a comment to the code; a comment is text | |
3352 that the Lisp interpreter ignores, but that a human reader may find | |
3353 useful or enlightening. The comment is that this is the ``second | |
3354 version''. | |
3355 | |
3356 @smallexample | |
3357 @group | |
3358 (defun multiply-by-seven (number) ; @r{Second version.} | |
3359 "Multiply NUMBER by seven." | |
3360 (+ number number number number number number number)) | |
3361 @end group | |
3362 @end smallexample | |
3363 | |
3364 @cindex Comments in Lisp code | |
3365 The comment follows a semicolon, @samp{;}. In Lisp, everything on a | |
3366 line that follows a semicolon is a comment. The end of the line is the | |
3367 end of the comment. To stretch a comment over two or more lines, begin | |
3368 each line with a semicolon. | |
3369 | |
3370 @xref{Beginning a .emacs File, , Beginning a @file{.emacs} | |
3371 File}, and @ref{Comments, , Comments, elisp, The GNU Emacs Lisp | |
3372 Reference Manual}, for more about comments. | |
3373 | |
3374 You can install this version of the @code{multiply-by-seven} function by | |
3375 evaluating it in the same way you evaluated the first function: place | |
3376 the cursor after the last parenthesis and type @kbd{C-x C-e}. | |
3377 | |
3378 In summary, this is how you write code in Emacs Lisp: you write a | |
3379 function; install it; test it; and then make fixes or enhancements and | |
3380 install it again. | |
3381 | |
3382 @node Interactive, Interactive Options, Install, Writing Defuns | |
3383 @comment node-name, next, previous, up | |
3384 @section Make a Function Interactive | |
3385 @cindex Interactive functions | |
3386 @findex interactive | |
3387 | |
3388 You make a function interactive by placing a list that begins with | |
3389 the special form @code{interactive} immediately after the | |
3390 documentation. A user can invoke an interactive function by typing | |
3391 @kbd{M-x} and then the name of the function; or by typing the keys to | |
3392 which it is bound, for example, by typing @kbd{C-n} for | |
3393 @code{next-line} or @kbd{C-x h} for @code{mark-whole-buffer}. | |
3394 | |
3395 Interestingly, when you call an interactive function interactively, | |
3396 the value returned is not automatically displayed in the echo area. | |
3397 This is because you often call an interactive function for its side | |
3398 effects, such as moving forward by a word or line, and not for the | |
3399 value returned. If the returned value were displayed in the echo area | |
3400 each time you typed a key, it would be very distracting. | |
3401 | |
3402 @menu | |
3403 * Interactive multiply-by-seven:: An overview. | |
3404 * multiply-by-seven in detail:: The interactive version. | |
3405 @end menu | |
3406 | |
3407 @node Interactive multiply-by-seven, multiply-by-seven in detail, Interactive, Interactive | |
3408 @ifnottex | |
3409 @unnumberedsubsec An Interactive @code{multiply-by-seven}, An Overview | |
3410 @end ifnottex | |
3411 | |
3412 Both the use of the special form @code{interactive} and one way to | |
3413 display a value in the echo area can be illustrated by creating an | |
3414 interactive version of @code{multiply-by-seven}. | |
3415 | |
3416 @need 1250 | |
3417 Here is the code: | |
3418 | |
3419 @smallexample | |
3420 @group | |
3421 (defun multiply-by-seven (number) ; @r{Interactive version.} | |
3422 "Multiply NUMBER by seven." | |
3423 (interactive "p") | |
3424 (message "The result is %d" (* 7 number))) | |
3425 @end group | |
3426 @end smallexample | |
3427 | |
3428 @noindent | |
3429 You can install this code by placing your cursor after it and typing | |
3430 @kbd{C-x C-e}. The name of the function will appear in your echo area. | |
3431 Then, you can use this code by typing @kbd{C-u} and a number and then | |
3432 typing @kbd{M-x multiply-by-seven} and pressing @key{RET}. The phrase | |
3433 @samp{The result is @dots{}} followed by the product will appear in the | |
3434 echo area. | |
3435 | |
3436 Speaking more generally, you invoke a function like this in either of two | |
3437 ways: | |
3438 | |
3439 @enumerate | |
3440 @item | |
3441 By typing a prefix argument that contains the number to be passed, and | |
3442 then typing @kbd{M-x} and the name of the function, as with | |
3443 @kbd{C-u 3 M-x forward-sentence}; or, | |
3444 | |
3445 @item | |
3446 By typing whatever key or keychord the function is bound to, as with | |
3447 @kbd{C-u 3 M-e}. | |
3448 @end enumerate | |
3449 | |
3450 @noindent | |
3451 Both the examples just mentioned work identically to move point forward | |
3452 three sentences. (Since @code{multiply-by-seven} is not bound to a key, | |
3453 it could not be used as an example of key binding.) | |
3454 | |
3455 (@xref{Keybindings, , Some Keybindings}, to learn how to bind a command | |
3456 to a key.) | |
3457 | |
3458 A prefix argument is passed to an interactive function by typing the | |
3459 @key{META} key followed by a number, for example, @kbd{M-3 M-e}, or by | |
3460 typing @kbd{C-u} and then a number, for example, @kbd{C-u 3 M-e} (if you | |
3461 type @kbd{C-u} without a number, it defaults to 4). | |
3462 | |
3463 @node multiply-by-seven in detail, , Interactive multiply-by-seven, Interactive | |
3464 @comment node-name, next, previous, up | |
3465 @subsection An Interactive @code{multiply-by-seven} | |
3466 | |
3467 Let's look at the use of the special form @code{interactive} and then at | |
3468 the function @code{message} in the interactive version of | |
3469 @code{multiply-by-seven}. You will recall that the function definition | |
3470 looks like this: | |
3471 | |
3472 @smallexample | |
3473 @group | |
3474 (defun multiply-by-seven (number) ; @r{Interactive version.} | |
3475 "Multiply NUMBER by seven." | |
3476 (interactive "p") | |
3477 (message "The result is %d" (* 7 number))) | |
3478 @end group | |
3479 @end smallexample | |
3480 | |
3481 In this function, the expression, @code{(interactive "p")}, is a list of | |
3482 two elements. The @code{"p"} tells Emacs to pass the prefix argument to | |
3483 the function and use its value for the argument of the function. | |
3484 | |
3485 @need 1000 | |
3486 The argument will be a number. This means that the symbol | |
3487 @code{number} will be bound to a number in the line: | |
3488 | |
3489 @smallexample | |
3490 (message "The result is %d" (* 7 number)) | |
3491 @end smallexample | |
3492 | |
3493 @need 1250 | |
3494 @noindent | |
3495 For example, if your prefix argument is 5, the Lisp interpreter will | |
3496 evaluate the line as if it were: | |
3497 | |
3498 @smallexample | |
3499 (message "The result is %d" (* 7 5)) | |
3500 @end smallexample | |
3501 | |
3502 @noindent | |
3503 (If you are reading this in GNU Emacs, you can evaluate this expression | |
3504 yourself.) First, the interpreter will evaluate the inner list, which | |
3505 is @code{(* 7 5)}. This returns a value of 35. Next, it | |
3506 will evaluate the outer list, passing the values of the second and | |
3507 subsequent elements of the list to the function @code{message}. | |
3508 | |
3509 As we have seen, @code{message} is an Emacs Lisp function especially | |
3510 designed for sending a one line message to a user. (@xref{message, , | |
3511 The @code{message} function}.) In summary, the @code{message} | |
3512 function prints its first argument in the echo area as is, except for | |
3513 occurrences of @samp{%d} or @samp{%s} (and various other %-sequences | |
3514 which we have not mentioned). When it sees a control sequence, the | |
3515 function looks to the second or subsequent arguments and prints the | |
3516 value of the argument in the location in the string where the control | |
3517 sequence is located. | |
3518 | |
3519 In the interactive @code{multiply-by-seven} function, the control string | |
3520 is @samp{%d}, which requires a number, and the value returned by | |
3521 evaluating @code{(* 7 5)} is the number 35. Consequently, the number 35 | |
3522 is printed in place of the @samp{%d} and the message is @samp{The result | |
3523 is 35}. | |
3524 | |
3525 (Note that when you call the function @code{multiply-by-seven}, the | |
3526 message is printed without quotes, but when you call @code{message}, the | |
3527 text is printed in double quotes. This is because the value returned by | |
3528 @code{message} is what appears in the echo area when you evaluate an | |
3529 expression whose first element is @code{message}; but when embedded in a | |
3530 function, @code{message} prints the text as a side effect without | |
3531 quotes.) | |
3532 | |
3533 @node Interactive Options, Permanent Installation, Interactive, Writing Defuns | |
3534 @comment node-name, next, previous, up | |
3535 @section Different Options for @code{interactive} | |
3536 @cindex Options for @code{interactive} | |
3537 @cindex Interactive options | |
3538 | |
3539 In the example, @code{multiply-by-seven} used @code{"p"} as the | |
3540 argument to @code{interactive}. This argument told Emacs to interpret | |
3541 your typing either @kbd{C-u} followed by a number or @key{META} | |
3542 followed by a number as a command to pass that number to the function | |
3543 as its argument. Emacs has more than twenty characters predefined for | |
3544 use with @code{interactive}. In almost every case, one of these | |
3545 options will enable you to pass the right information interactively to | |
3546 a function. (@xref{Interactive Codes, , Code Characters for | |
3547 @code{interactive}, elisp, The GNU Emacs Lisp Reference Manual}.) | |
3548 | |
3549 @need 1250 | |
3550 Consider the function @code{zap-to-char}. Its interactive expression | |
3551 is | |
3552 | |
3553 @smallexample | |
3554 (interactive "p\ncZap to char: ") | |
3555 @end smallexample | |
3556 | |
3557 The first part of the argument to @code{interactive} is @samp{p}, with | |
3558 which you are already familiar. This argument tells Emacs to | |
3559 interpret a `prefix', as a number to be passed to the function. You | |
3560 can specify a prefix either by typing @kbd{C-u} followed by a number | |
3561 or by typing @key{META} followed by a number. The prefix is the | |
3562 number of specified characters. Thus, if your prefix is three and the | |
3563 specified character is @samp{x}, then you will delete all the text up | |
3564 to and including the third next @samp{x}. If you do not set a prefix, | |
3565 then you delete all the text up to and including the specified | |
3566 character, but no more. | |
3567 | |
3568 The @samp{c} tells the function the name of the character to which to delete. | |
3569 | |
3570 More formally, a function with two or more arguments can have | |
3571 information passed to each argument by adding parts to the string that | |
3572 follows @code{interactive}. When you do this, the information is | |
3573 passed to each argument in the same order it is specified in the | |
3574 @code{interactive} list. In the string, each part is separated from | |
3575 the next part by a @samp{\n}, which is a newline. For example, you | |
3576 can follow @samp{p} with a @samp{\n} and an @samp{cZap to char:@: }. | |
3577 This causes Emacs to pass the value of the prefix argument (if there | |
3578 is one) and the character. | |
3579 | |
3580 In this case, the function definition looks like the following, where | |
3581 @code{arg} and @code{char} are the symbols to which @code{interactive} | |
3582 binds the prefix argument and the specified character: | |
3583 | |
3584 @smallexample | |
3585 @group | |
3586 (defun @var{name-of-function} (arg char) | |
3587 "@var{documentation}@dots{}" | |
3588 (interactive "p\ncZap to char: ") | |
3589 @var{body-of-function}@dots{}) | |
3590 @end group | |
3591 @end smallexample | |
3592 | |
3593 @noindent | |
3594 (The space after the colon in the prompt makes it look better when you | |
3595 are prompted. @xref{copy-to-buffer, , The Definition of | |
3596 @code{copy-to-buffer}}, for an example.) | |
3597 | |
3598 When a function does not take arguments, @code{interactive} does not | |
3599 require any. Such a function contains the simple expression | |
3600 @code{(interactive)}. The @code{mark-whole-buffer} function is like | |
3601 this. | |
3602 | |
3603 Alternatively, if the special letter-codes are not right for your | |
3604 application, you can pass your own arguments to @code{interactive} as | |
3605 a list. | |
3606 | |
3607 @xref{append-to-buffer, , The Definition of @code{append-to-buffer}}, | |
3608 for an example. @xref{Using Interactive, , Using @code{Interactive}, | |
3609 elisp, The GNU Emacs Lisp Reference Manual}, for a more complete | |
3610 explanation about this technique. | |
3611 | |
3612 @node Permanent Installation, let, Interactive Options, Writing Defuns | |
3613 @comment node-name, next, previous, up | |
3614 @section Install Code Permanently | |
3615 @cindex Install code permanently | |
3616 @cindex Permanent code installation | |
3617 @cindex Code installation | |
3618 | |
3619 When you install a function definition by evaluating it, it will stay | |
3620 installed until you quit Emacs. The next time you start a new session | |
3621 of Emacs, the function will not be installed unless you evaluate the | |
3622 function definition again. | |
3623 | |
3624 At some point, you may want to have code installed automatically | |
3625 whenever you start a new session of Emacs. There are several ways of | |
3626 doing this: | |
3627 | |
3628 @itemize @bullet | |
3629 @item | |
3630 If you have code that is just for yourself, you can put the code for the | |
3631 function definition in your @file{.emacs} initialization file. When you | |
3632 start Emacs, your @file{.emacs} file is automatically evaluated and all | |
3633 the function definitions within it are installed. | |
3634 @xref{Emacs Initialization, , Your @file{.emacs} File}. | |
3635 | |
3636 @item | |
3637 Alternatively, you can put the function definitions that you want | |
3638 installed in one or more files of their own and use the @code{load} | |
3639 function to cause Emacs to evaluate and thereby install each of the | |
3640 functions in the files. | |
3641 @xref{Loading Files, , Loading Files}. | |
3642 | |
3643 @item | |
3644 Thirdly, if you have code that your whole site will use, it is usual | |
3645 to put it in a file called @file{site-init.el} that is loaded when | |
3646 Emacs is built. This makes the code available to everyone who uses | |
3647 your machine. (See the @file{INSTALL} file that is part of the Emacs | |
3648 distribution.) | |
3649 @end itemize | |
3650 | |
3651 Finally, if you have code that everyone who uses Emacs may want, you | |
3652 can post it on a computer network or send a copy to the Free Software | |
3653 Foundation. (When you do this, please license the code and its | |
3654 documentation under a license that permits other people to run, copy, | |
3655 study, modify, and redistribute the code and which protects you from | |
3656 having your work taken from you.) If you send a copy of your code to | |
3657 the Free Software Foundation, and properly protect yourself and | |
3658 others, it may be included in the next release of Emacs. In large | |
3659 part, this is how Emacs has grown over the past years, by donations. | |
3660 | |
3661 @node let, if, Permanent Installation, Writing Defuns | |
3662 @comment node-name, next, previous, up | |
3663 @section @code{let} | |
3664 @findex let | |
3665 | |
3666 The @code{let} expression is a special form in Lisp that you will need | |
3667 to use in most function definitions. | |
3668 | |
3669 @code{let} is used to attach or bind a symbol to a value in such a way | |
3670 that the Lisp interpreter will not confuse the variable with a | |
3671 variable of the same name that is not part of the function. | |
3672 | |
3673 To understand why the @code{let} special form is necessary, consider | |
3674 the situation in which you own a home that you generally refer to as | |
3675 `the house', as in the sentence, ``The house needs painting.'' If you | |
3676 are visiting a friend and your host refers to `the house', he is | |
3677 likely to be referring to @emph{his} house, not yours, that is, to a | |
3678 different house. | |
3679 | |
3680 If your friend is referring to his house and you think he is referring | |
3681 to your house, you may be in for some confusion. The same thing could | |
3682 happen in Lisp if a variable that is used inside of one function has | |
3683 the same name as a variable that is used inside of another function, | |
3684 and the two are not intended to refer to the same value. The | |
3685 @code{let} special form prevents this kind of confusion. | |
3686 | |
3687 @menu | |
3688 * Prevent confusion:: | |
3689 * Parts of let Expression:: | |
3690 * Sample let Expression:: | |
3691 * Uninitialized let Variables:: | |
3692 @end menu | |
3693 | |
3694 @node Prevent confusion, Parts of let Expression, let, let | |
3695 @ifnottex | |
3696 @unnumberedsubsec @code{let} Prevents Confusion | |
3697 @end ifnottex | |
3698 | |
3699 @cindex @samp{local variable} defined | |
3700 @cindex @samp{variable, local}, defined | |
3701 The @code{let} special form prevents confusion. @code{let} creates a | |
3702 name for a @dfn{local variable} that overshadows any use of the same | |
3703 name outside the @code{let} expression. This is like understanding | |
3704 that whenever your host refers to `the house', he means his house, not | |
3705 yours. (Symbols used in argument lists work the same way. | |
3706 @xref{defun, , The @code{defun} Special Form}.) | |
3707 | |
3708 Local variables created by a @code{let} expression retain their value | |
3709 @emph{only} within the @code{let} expression itself (and within | |
3710 expressions called within the @code{let} expression); the local | |
3711 variables have no effect outside the @code{let} expression. | |
3712 | |
3713 Another way to think about @code{let} is that it is like a @code{setq} | |
3714 that is temporary and local. The values set by @code{let} are | |
3715 automatically undone when the @code{let} is finished. The setting | |
3716 only affects expressions that are inside the bounds of the @code{let} | |
3717 expression. In computer science jargon, we would say ``the binding of | |
3718 a symbol is visible only in functions called in the @code{let} form; | |
3719 in Emacs Lisp, scoping is dynamic, not lexical.'' | |
3720 | |
3721 @code{let} can create more than one variable at once. Also, | |
3722 @code{let} gives each variable it creates an initial value, either a | |
3723 value specified by you, or @code{nil}. (In the jargon, this is called | |
3724 `binding the variable to the value'.) After @code{let} has created | |
3725 and bound the variables, it executes the code in the body of the | |
3726 @code{let}, and returns the value of the last expression in the body, | |
3727 as the value of the whole @code{let} expression. (`Execute' is a jargon | |
3728 term that means to evaluate a list; it comes from the use of the word | |
3729 meaning `to give practical effect to' (@cite{Oxford English | |
3730 Dictionary}). Since you evaluate an expression to perform an action, | |
3731 `execute' has evolved as a synonym to `evaluate'.) | |
3732 | |
3733 @node Parts of let Expression, Sample let Expression, Prevent confusion, let | |
3734 @comment node-name, next, previous, up | |
3735 @subsection The Parts of a @code{let} Expression | |
3736 @cindex @code{let} expression, parts of | |
3737 @cindex Parts of @code{let} expression | |
3738 | |
3739 @cindex @samp{varlist} defined | |
3740 A @code{let} expression is a list of three parts. The first part is | |
3741 the symbol @code{let}. The second part is a list, called a | |
3742 @dfn{varlist}, each element of which is either a symbol by itself or a | |
3743 two-element list, the first element of which is a symbol. The third | |
3744 part of the @code{let} expression is the body of the @code{let}. The | |
3745 body usually consists of one or more lists. | |
3746 | |
3747 @need 800 | |
3748 A template for a @code{let} expression looks like this: | |
3749 | |
3750 @smallexample | |
3751 (let @var{varlist} @var{body}@dots{}) | |
3752 @end smallexample | |
3753 | |
3754 @noindent | |
3755 The symbols in the varlist are the variables that are given initial | |
3756 values by the @code{let} special form. Symbols by themselves are given | |
3757 the initial value of @code{nil}; and each symbol that is the first | |
3758 element of a two-element list is bound to the value that is returned | |
3759 when the Lisp interpreter evaluates the second element. | |
3760 | |
3761 Thus, a varlist might look like this: @code{(thread (needles 3))}. In | |
3762 this case, in a @code{let} expression, Emacs binds the symbol | |
3763 @code{thread} to an initial value of @code{nil}, and binds the symbol | |
3764 @code{needles} to an initial value of 3. | |
3765 | |
3766 When you write a @code{let} expression, what you do is put the | |
3767 appropriate expressions in the slots of the @code{let} expression | |
3768 template. | |
3769 | |
3770 If the varlist is composed of two-element lists, as is often the case, | |
3771 the template for the @code{let} expression looks like this: | |
3772 | |
3773 @smallexample | |
3774 @group | |
3775 (let ((@var{variable} @var{value}) | |
3776 (@var{variable} @var{value}) | |
3777 @dots{}) | |
3778 @var{body}@dots{}) | |
3779 @end group | |
3780 @end smallexample | |
3781 | |
3782 @node Sample let Expression, Uninitialized let Variables, Parts of let Expression, let | |
3783 @comment node-name, next, previous, up | |
3784 @subsection Sample @code{let} Expression | |
3785 @cindex Sample @code{let} expression | |
3786 @cindex @code{let} expression sample | |
3787 | |
3788 The following expression creates and gives initial values | |
3789 to the two variables @code{zebra} and @code{tiger}. The body of the | |
3790 @code{let} expression is a list which calls the @code{message} function. | |
3791 | |
3792 @smallexample | |
3793 @group | |
3794 (let ((zebra 'stripes) | |
3795 (tiger 'fierce)) | |
3796 (message "One kind of animal has %s and another is %s." | |
3797 zebra tiger)) | |
3798 @end group | |
3799 @end smallexample | |
3800 | |
3801 Here, the varlist is @code{((zebra 'stripes) (tiger 'fierce))}. | |
3802 | |
3803 The two variables are @code{zebra} and @code{tiger}. Each variable is | |
3804 the first element of a two-element list and each value is the second | |
3805 element of its two-element list. In the varlist, Emacs binds the | |
3806 variable @code{zebra} to the value @code{stripes}@footnote{According | |
3807 to Jared Diamond in @cite{Guns, Germs, and Steel}, ``@dots{} zebras | |
3808 become impossibly dangerous as they grow older'' but the claim here is | |
3809 that they do not become fierce like a tiger. (1997, W. W. Norton and | |
3810 Co., ISBN 0-393-03894-2, page 171)}, and binds the | |
3811 variable @code{tiger} to the value @code{fierce}. In this example, | |
3812 both values are symbols preceded by a quote. The values could just as | |
3813 well have been another list or a string. The body of the @code{let} | |
3814 follows after the list holding the variables. In this example, the | |
3815 body is a list that uses the @code{message} function to print a string | |
3816 in the echo area. | |
3817 | |
3818 @need 1500 | |
3819 You may evaluate the example in the usual fashion, by placing the | |
3820 cursor after the last parenthesis and typing @kbd{C-x C-e}. When you do | |
3821 this, the following will appear in the echo area: | |
3822 | |
3823 @smallexample | |
3824 "One kind of animal has stripes and another is fierce." | |
3825 @end smallexample | |
3826 | |
3827 As we have seen before, the @code{message} function prints its first | |
3828 argument, except for @samp{%s}. In this example, the value of the variable | |
3829 @code{zebra} is printed at the location of the first @samp{%s} and the | |
3830 value of the variable @code{tiger} is printed at the location of the | |
3831 second @samp{%s}. | |
3832 | |
3833 @node Uninitialized let Variables, , Sample let Expression, let | |
3834 @comment node-name, next, previous, up | |
3835 @subsection Uninitialized Variables in a @code{let} Statement | |
3836 @cindex Uninitialized @code{let} variables | |
3837 @cindex @code{let} variables uninitialized | |
3838 | |
3839 If you do not bind the variables in a @code{let} statement to specific | |
3840 initial values, they will automatically be bound to an initial value of | |
3841 @code{nil}, as in the following expression: | |
3842 | |
3843 @smallexample | |
3844 @group | |
3845 (let ((birch 3) | |
3846 pine | |
3847 fir | |
3848 (oak 'some)) | |
3849 (message | |
3850 "Here are %d variables with %s, %s, and %s value." | |
3851 birch pine fir oak)) | |
3852 @end group | |
3853 @end smallexample | |
3854 | |
3855 @noindent | |
3856 Here, the varlist is @code{((birch 3) pine fir (oak 'some))}. | |
3857 | |
3858 @need 1250 | |
3859 If you evaluate this expression in the usual way, the following will | |
3860 appear in your echo area: | |
3861 | |
3862 @smallexample | |
3863 "Here are 3 variables with nil, nil, and some value." | |
3864 @end smallexample | |
3865 | |
3866 @noindent | |
3867 In this example, Emacs binds the symbol @code{birch} to the number 3, | |
3868 binds the symbols @code{pine} and @code{fir} to @code{nil}, and binds | |
3869 the symbol @code{oak} to the value @code{some}. | |
3870 | |
3871 Note that in the first part of the @code{let}, the variables @code{pine} | |
3872 and @code{fir} stand alone as atoms that are not surrounded by | |
3873 parentheses; this is because they are being bound to @code{nil}, the | |
3874 empty list. But @code{oak} is bound to @code{some} and so is a part of | |
3875 the list @code{(oak 'some)}. Similarly, @code{birch} is bound to the | |
3876 number 3 and so is in a list with that number. (Since a number | |
3877 evaluates to itself, the number does not need to be quoted. Also, the | |
3878 number is printed in the message using a @samp{%d} rather than a | |
3879 @samp{%s}.) The four variables as a group are put into a list to | |
3880 delimit them from the body of the @code{let}. | |
3881 | |
3882 @node if, else, let, Writing Defuns | |
3883 @comment node-name, next, previous, up | |
3884 @section The @code{if} Special Form | |
3885 @findex if | |
3886 @cindex Conditional with @code{if} | |
3887 | |
3888 A third special form, in addition to @code{defun} and @code{let}, is the | |
3889 conditional @code{if}. This form is used to instruct the computer to | |
3890 make decisions. You can write function definitions without using | |
3891 @code{if}, but it is used often enough, and is important enough, to be | |
3892 included here. It is used, for example, in the code for the | |
3893 function @code{beginning-of-buffer}. | |
3894 | |
3895 The basic idea behind an @code{if}, is that ``@emph{if} a test is true, | |
3896 @emph{then} an expression is evaluated.'' If the test is not true, the | |
3897 expression is not evaluated. For example, you might make a decision | |
3898 such as, ``if it is warm and sunny, then go to the beach!'' | |
3899 | |
3900 @menu | |
3901 * if in more detail:: | |
3902 * type-of-animal in detail:: An example of an @code{if} expression. | |
3903 @end menu | |
3904 | |
3905 @node if in more detail, type-of-animal in detail, if, if | |
3906 @ifnottex | |
3907 @unnumberedsubsec @code{if} in more detail | |
3908 @end ifnottex | |
3909 | |
3910 @cindex @samp{if-part} defined | |
3911 @cindex @samp{then-part} defined | |
3912 An @code{if} expression written in Lisp does not use the word `then'; | |
3913 the test and the action are the second and third elements of the list | |
3914 whose first element is @code{if}. Nonetheless, the test part of an | |
3915 @code{if} expression is often called the @dfn{if-part} and the second | |
3916 argument is often called the @dfn{then-part}. | |
3917 | |
3918 Also, when an @code{if} expression is written, the true-or-false-test | |
3919 is usually written on the same line as the symbol @code{if}, but the | |
3920 action to carry out if the test is true, the ``then-part'', is written | |
3921 on the second and subsequent lines. This makes the @code{if} | |
3922 expression easier to read. | |
3923 | |
3924 @smallexample | |
3925 @group | |
3926 (if @var{true-or-false-test} | |
3927 @var{action-to-carry-out-if-test-is-true}) | |
3928 @end group | |
3929 @end smallexample | |
3930 | |
3931 @noindent | |
3932 The true-or-false-test will be an expression that | |
3933 is evaluated by the Lisp interpreter. | |
3934 | |
3935 Here is an example that you can evaluate in the usual manner. The test | |
3936 is whether the number 5 is greater than the number 4. Since it is, the | |
3937 message @samp{5 is greater than 4!} will be printed. | |
3938 | |
3939 @smallexample | |
3940 @group | |
3941 (if (> 5 4) ; @r{if-part} | |
3942 (message "5 is greater than 4!")) ; @r{then-part} | |
3943 @end group | |
3944 @end smallexample | |
3945 | |
3946 @noindent | |
3947 (The function @code{>} tests whether its first argument is greater than | |
3948 its second argument and returns true if it is.) | |
3949 @findex > (greater than) | |
3950 | |
3951 Of course, in actual use, the test in an @code{if} expression will not | |
3952 be fixed for all time as it is by the expression @code{(> 5 4)}. | |
3953 Instead, at least one of the variables used in the test will be bound to | |
3954 a value that is not known ahead of time. (If the value were known ahead | |
3955 of time, we would not need to run the test!) | |
3956 | |
3957 For example, the value may be bound to an argument of a function | |
3958 definition. In the following function definition, the character of the | |
3959 animal is a value that is passed to the function. If the value bound to | |
3960 @code{characteristic} is @code{fierce}, then the message, @samp{It's a | |
3961 tiger!} will be printed; otherwise, @code{nil} will be returned. | |
3962 | |
3963 @smallexample | |
3964 @group | |
3965 (defun type-of-animal (characteristic) | |
3966 "Print message in echo area depending on CHARACTERISTIC. | |
3967 If the CHARACTERISTIC is the symbol `fierce', | |
3968 then warn of a tiger." | |
3969 (if (equal characteristic 'fierce) | |
3970 (message "It's a tiger!"))) | |
3971 @end group | |
3972 @end smallexample | |
3973 | |
3974 @need 1500 | |
3975 @noindent | |
3976 If you are reading this inside of GNU Emacs, you can evaluate the | |
3977 function definition in the usual way to install it in Emacs, and then you | |
3978 can evaluate the following two expressions to see the results: | |
3979 | |
3980 @smallexample | |
3981 @group | |
3982 (type-of-animal 'fierce) | |
3983 | |
3984 (type-of-animal 'zebra) | |
3985 | |
3986 @end group | |
3987 @end smallexample | |
3988 | |
3989 @c Following sentences rewritten to prevent overfull hbox. | |
3990 @noindent | |
3991 When you evaluate @code{(type-of-animal 'fierce)}, you will see the | |
3992 following message printed in the echo area: @code{"It's a tiger!"}; and | |
3993 when you evaluate @code{(type-of-animal 'zebra)} you will see @code{nil} | |
3994 printed in the echo area. | |
3995 | |
3996 @node type-of-animal in detail, , if in more detail, if | |
3997 @comment node-name, next, previous, up | |
3998 @subsection The @code{type-of-animal} Function in Detail | |
3999 | |
4000 Let's look at the @code{type-of-animal} function in detail. | |
4001 | |
4002 The function definition for @code{type-of-animal} was written by filling | |
4003 the slots of two templates, one for a function definition as a whole, and | |
4004 a second for an @code{if} expression. | |
4005 | |
4006 @need 1250 | |
4007 The template for every function that is not interactive is: | |
4008 | |
4009 @smallexample | |
4010 @group | |
4011 (defun @var{name-of-function} (@var{argument-list}) | |
4012 "@var{documentation}@dots{}" | |
4013 @var{body}@dots{}) | |
4014 @end group | |
4015 @end smallexample | |
4016 | |
4017 @need 800 | |
4018 The parts of the function that match this template look like this: | |
4019 | |
4020 @smallexample | |
4021 @group | |
4022 (defun type-of-animal (characteristic) | |
4023 "Print message in echo area depending on CHARACTERISTIC. | |
4024 If the CHARACTERISTIC is the symbol `fierce', | |
4025 then warn of a tiger." | |
4026 @var{body: the} @code{if} @var{expression}) | |
4027 @end group | |
4028 @end smallexample | |
4029 | |
4030 The name of function is @code{type-of-animal}; it is passed the value | |
4031 of one argument. The argument list is followed by a multi-line | |
4032 documentation string. The documentation string is included in the | |
4033 example because it is a good habit to write documentation string for | |
4034 every function definition. The body of the function definition | |
4035 consists of the @code{if} expression. | |
4036 | |
4037 @need 800 | |
4038 The template for an @code{if} expression looks like this: | |
4039 | |
4040 @smallexample | |
4041 @group | |
4042 (if @var{true-or-false-test} | |
4043 @var{action-to-carry-out-if-the-test-returns-true}) | |
4044 @end group | |
4045 @end smallexample | |
4046 | |
4047 @need 1250 | |
4048 In the @code{type-of-animal} function, the code for the @code{if} | |
4049 looks like this: | |
4050 | |
4051 @smallexample | |
4052 @group | |
4053 (if (equal characteristic 'fierce) | |
4054 (message "It's a tiger!"))) | |
4055 @end group | |
4056 @end smallexample | |
4057 | |
4058 @need 800 | |
4059 Here, the true-or-false-test is the expression: | |
4060 | |
4061 @smallexample | |
4062 (equal characteristic 'fierce) | |
4063 @end smallexample | |
4064 | |
4065 @noindent | |
4066 In Lisp, @code{equal} is a function that determines whether its first | |
4067 argument is equal to its second argument. The second argument is the | |
4068 quoted symbol @code{'fierce} and the first argument is the value of the | |
4069 symbol @code{characteristic}---in other words, the argument passed to | |
4070 this function. | |
4071 | |
4072 In the first exercise of @code{type-of-animal}, the argument | |
4073 @code{fierce} is passed to @code{type-of-animal}. Since @code{fierce} | |
4074 is equal to @code{fierce}, the expression, @code{(equal characteristic | |
4075 'fierce)}, returns a value of true. When this happens, the @code{if} | |
4076 evaluates the second argument or then-part of the @code{if}: | |
4077 @code{(message "It's tiger!")}. | |
4078 | |
4079 On the other hand, in the second exercise of @code{type-of-animal}, the | |
4080 argument @code{zebra} is passed to @code{type-of-animal}. @code{zebra} | |
4081 is not equal to @code{fierce}, so the then-part is not evaluated and | |
4082 @code{nil} is returned by the @code{if} expression. | |
4083 | |
4084 @node else, Truth & Falsehood, if, Writing Defuns | |
4085 @comment node-name, next, previous, up | |
4086 @section If--then--else Expressions | |
4087 @cindex Else | |
4088 | |
4089 An @code{if} expression may have an optional third argument, called | |
4090 the @dfn{else-part}, for the case when the true-or-false-test returns | |
4091 false. When this happens, the second argument or then-part of the | |
4092 overall @code{if} expression is @emph{not} evaluated, but the third or | |
4093 else-part @emph{is} evaluated. You might think of this as the cloudy | |
4094 day alternative for the decision ``if it is warm and sunny, then go to | |
4095 the beach, else read a book!''. | |
4096 | |
4097 The word ``else'' is not written in the Lisp code; the else-part of an | |
4098 @code{if} expression comes after the then-part. In the written Lisp, the | |
4099 else-part is usually written to start on a line of its own and is | |
4100 indented less than the then-part: | |
4101 | |
4102 @smallexample | |
4103 @group | |
4104 (if @var{true-or-false-test} | |
4105 @var{action-to-carry-out-if-the-test-returns-true} | |
4106 @var{action-to-carry-out-if-the-test-returns-false}) | |
4107 @end group | |
4108 @end smallexample | |
4109 | |
4110 For example, the following @code{if} expression prints the message @samp{4 | |
4111 is not greater than 5!} when you evaluate it in the usual way: | |
4112 | |
4113 @smallexample | |
4114 @group | |
4115 (if (> 4 5) ; @r{if-part} | |
4116 (message "4 falsely greater than 5!") ; @r{then-part} | |
4117 (message "4 is not greater than 5!")) ; @r{else-part} | |
4118 @end group | |
4119 @end smallexample | |
4120 | |
4121 @noindent | |
4122 Note that the different levels of indentation make it easy to | |
4123 distinguish the then-part from the else-part. (GNU Emacs has several | |
4124 commands that automatically indent @code{if} expressions correctly. | |
4125 @xref{Typing Lists, , GNU Emacs Helps You Type Lists}.) | |
4126 | |
4127 We can extend the @code{type-of-animal} function to include an | |
4128 else-part by simply incorporating an additional part to the @code{if} | |
4129 expression. | |
4130 | |
4131 @need 1500 | |
4132 You can see the consequences of doing this if you evaluate the following | |
4133 version of the @code{type-of-animal} function definition to install it | |
4134 and then evaluate the two subsequent expressions to pass different | |
4135 arguments to the function. | |
4136 | |
4137 @smallexample | |
4138 @group | |
4139 (defun type-of-animal (characteristic) ; @r{Second version.} | |
4140 "Print message in echo area depending on CHARACTERISTIC. | |
4141 If the CHARACTERISTIC is the symbol `fierce', | |
4142 then warn of a tiger; | |
4143 else say it's not fierce." | |
4144 (if (equal characteristic 'fierce) | |
4145 (message "It's a tiger!") | |
4146 (message "It's not fierce!"))) | |
4147 @end group | |
4148 @end smallexample | |
4149 @sp 1 | |
4150 | |
4151 @smallexample | |
4152 @group | |
4153 (type-of-animal 'fierce) | |
4154 | |
4155 (type-of-animal 'zebra) | |
4156 | |
4157 @end group | |
4158 @end smallexample | |
4159 | |
4160 @c Following sentence rewritten to prevent overfull hbox. | |
4161 @noindent | |
4162 When you evaluate @code{(type-of-animal 'fierce)}, you will see the | |
4163 following message printed in the echo area: @code{"It's a tiger!"}; but | |
4164 when you evaluate @code{(type-of-animal 'zebra)}, you will see | |
4165 @code{"It's not fierce!"}. | |
4166 | |
4167 (Of course, if the @var{characteristic} were @code{ferocious}, the | |
4168 message @code{"It's not fierce!"} would be printed; and it would be | |
4169 misleading! When you write code, you need to take into account the | |
4170 possibility that some such argument will be tested by the @code{if} | |
4171 and write your program accordingly.) | |
4172 | |
4173 @node Truth & Falsehood, save-excursion, else, Writing Defuns | |
4174 @comment node-name, next, previous, up | |
4175 @section Truth and Falsehood in Emacs Lisp | |
4176 @cindex Truth and falsehood in Emacs Lisp | |
4177 @cindex Falsehood and truth in Emacs Lisp | |
4178 @findex nil | |
4179 | |
4180 There is an important aspect to the truth test in an @code{if} | |
4181 expression. So far, we have spoken of `true' and `false' as values of | |
4182 predicates as if they were new kinds of Emacs Lisp objects. In fact, | |
4183 `false' is just our old friend @code{nil}. Anything else---anything | |
4184 at all---is `true'. | |
4185 | |
4186 The expression that tests for truth is interpreted as @dfn{true} | |
4187 if the result of evaluating it is a value that is not @code{nil}. In | |
4188 other words, the result of the test is considered true if the value | |
4189 returned is a number such as 47, a string such as @code{"hello"}, or a | |
4190 symbol (other than @code{nil}) such as @code{flowers}, or a list (so | |
4191 long as it is not empty), or even a buffer! | |
4192 | |
4193 @menu | |
4194 * nil explained:: @code{nil} has two meanings. | |
4195 @end menu | |
4196 | |
4197 @node nil explained, , Truth & Falsehood, Truth & Falsehood | |
4198 @ifnottex | |
4199 @unnumberedsubsec An explanation of @code{nil} | |
4200 @end ifnottex | |
4201 | |
4202 Before illustrating a test for truth, we need an explanation of @code{nil}. | |
4203 | |
4204 In Emacs Lisp, the symbol @code{nil} has two meanings. First, it means the | |
4205 empty list. Second, it means false and is the value returned when a | |
4206 true-or-false-test tests false. @code{nil} can be written as an empty | |
4207 list, @code{()}, or as @code{nil}. As far as the Lisp interpreter is | |
4208 concerned, @code{()} and @code{nil} are the same. Humans, however, tend | |
4209 to use @code{nil} for false and @code{()} for the empty list. | |
4210 | |
4211 In Emacs Lisp, any value that is not @code{nil}---is not the empty | |
4212 list---is considered true. This means that if an evaluation returns | |
4213 something that is not an empty list, an @code{if} expression will test | |
4214 true. For example, if a number is put in the slot for the test, it | |
4215 will be evaluated and will return itself, since that is what numbers | |
4216 do when evaluated. In this conditional, the @code{if} expression will | |
4217 test true. The expression tests false only when @code{nil}, an empty | |
4218 list, is returned by evaluating the expression. | |
4219 | |
4220 You can see this by evaluating the two expressions in the following examples. | |
4221 | |
4222 In the first example, the number 4 is evaluated as the test in the | |
4223 @code{if} expression and returns itself; consequently, the then-part | |
4224 of the expression is evaluated and returned: @samp{true} appears in | |
4225 the echo area. In the second example, the @code{nil} indicates false; | |
4226 consequently, the else-part of the expression is evaluated and | |
4227 returned: @samp{false} appears in the echo area. | |
4228 | |
4229 @smallexample | |
4230 @group | |
4231 (if 4 | |
4232 'true | |
4233 'false) | |
4234 @end group | |
4235 | |
4236 @group | |
4237 (if nil | |
4238 'true | |
4239 'false) | |
4240 @end group | |
4241 @end smallexample | |
4242 | |
4243 @need 1250 | |
4244 Incidentally, if some other useful value is not available for a test that | |
4245 returns true, then the Lisp interpreter will return the symbol @code{t} | |
4246 for true. For example, the expression @code{(> 5 4)} returns @code{t} | |
4247 when evaluated, as you can see by evaluating it in the usual way: | |
4248 | |
4249 @smallexample | |
4250 (> 5 4) | |
4251 @end smallexample | |
4252 | |
4253 @need 1250 | |
4254 @noindent | |
4255 On the other hand, this function returns @code{nil} if the test is false. | |
4256 | |
4257 @smallexample | |
4258 (> 4 5) | |
4259 @end smallexample | |
4260 | |
4261 @node save-excursion, Review, Truth & Falsehood, Writing Defuns | |
4262 @comment node-name, next, previous, up | |
4263 @section @code{save-excursion} | |
4264 @findex save-excursion | |
4265 @cindex Region, what it is | |
4266 @cindex Preserving point, mark, and buffer | |
4267 @cindex Point, mark, buffer preservation | |
4268 @findex point | |
4269 @findex mark | |
4270 | |
4271 The @code{save-excursion} function is the fourth and final special form | |
4272 that we will discuss in this chapter. | |
4273 | |
4274 In Emacs Lisp programs used for editing, the @code{save-excursion} | |
4275 function is very common. It saves the location of point and mark, | |
4276 executes the body of the function, and then restores point and mark to | |
4277 their previous positions if their locations were changed. Its primary | |
4278 purpose is to keep the user from being surprised and disturbed by | |
4279 unexpected movement of point or mark. | |
4280 | |
4281 @menu | |
4282 * Point and mark:: A review of various locations. | |
4283 * Template for save-excursion:: | |
4284 @end menu | |
4285 | |
4286 @node Point and mark, Template for save-excursion, save-excursion, save-excursion | |
4287 @ifnottex | |
4288 @unnumberedsubsec Point and Mark | |
4289 @end ifnottex | |
4290 | |
4291 Before discussing @code{save-excursion}, however, it may be useful | |
4292 first to review what point and mark are in GNU Emacs. @dfn{Point} is | |
4293 the current location of the cursor. Wherever the cursor | |
4294 is, that is point. More precisely, on terminals where the cursor | |
4295 appears to be on top of a character, point is immediately before the | |
4296 character. In Emacs Lisp, point is an integer. The first character in | |
4297 a buffer is number one, the second is number two, and so on. The | |
4298 function @code{point} returns the current position of the cursor as a | |
4299 number. Each buffer has its own value for point. | |
4300 | |
4301 The @dfn{mark} is another position in the buffer; its value can be set | |
4302 with a command such as @kbd{C-@key{SPC}} (@code{set-mark-command}). If | |
4303 a mark has been set, you can use the command @kbd{C-x C-x} | |
4304 (@code{exchange-point-and-mark}) to cause the cursor to jump to the mark | |
4305 and set the mark to be the previous position of point. In addition, if | |
4306 you set another mark, the position of the previous mark is saved in the | |
4307 mark ring. Many mark positions can be saved this way. You can jump the | |
4308 cursor to a saved mark by typing @kbd{C-u C-@key{SPC}} one or more | |
4309 times. | |
4310 | |
4311 The part of the buffer between point and mark is called @dfn{the | |
4312 region}. Numerous commands work on the region, including | |
4313 @code{center-region}, @code{count-lines-region}, @code{kill-region}, and | |
4314 @code{print-region}. | |
4315 | |
4316 The @code{save-excursion} special form saves the locations of point and | |
4317 mark and restores those positions after the code within the body of the | |
4318 special form is evaluated by the Lisp interpreter. Thus, if point were | |
4319 in the beginning of a piece of text and some code moved point to the end | |
4320 of the buffer, the @code{save-excursion} would put point back to where | |
4321 it was before, after the expressions in the body of the function were | |
4322 evaluated. | |
4323 | |
4324 In Emacs, a function frequently moves point as part of its internal | |
4325 workings even though a user would not expect this. For example, | |
4326 @code{count-lines-region} moves point. To prevent the user from being | |
4327 bothered by jumps that are both unexpected and (from the user's point of | |
4328 view) unnecessary, @code{save-excursion} is often used to keep point and | |
4329 mark in the location expected by the user. The use of | |
4330 @code{save-excursion} is good housekeeping. | |
4331 | |
4332 To make sure the house stays clean, @code{save-excursion} restores the | |
4333 values of point and mark even if something goes wrong in the code inside | |
4334 of it (or, to be more precise and to use the proper jargon, ``in case of | |
4335 abnormal exit''). This feature is very helpful. | |
4336 | |
4337 In addition to recording the values of point and mark, | |
4338 @code{save-excursion} keeps track of the current buffer, and restores | |
4339 it, too. This means you can write code that will change the buffer and | |
4340 have @code{save-excursion} switch you back to the original buffer. | |
4341 This is how @code{save-excursion} is used in @code{append-to-buffer}. | |
4342 (@xref{append-to-buffer, , The Definition of @code{append-to-buffer}}.) | |
4343 | |
4344 @node Template for save-excursion, , Point and mark, save-excursion | |
4345 @comment node-name, next, previous, up | |
4346 @subsection Template for a @code{save-excursion} Expression | |
4347 | |
4348 @need 800 | |
4349 The template for code using @code{save-excursion} is simple: | |
4350 | |
4351 @smallexample | |
4352 @group | |
4353 (save-excursion | |
4354 @var{body}@dots{}) | |
4355 @end group | |
4356 @end smallexample | |
4357 | |
4358 @noindent | |
4359 The body of the function is one or more expressions that will be | |
4360 evaluated in sequence by the Lisp interpreter. If there is more than | |
4361 one expression in the body, the value of the last one will be returned | |
4362 as the value of the @code{save-excursion} function. The other | |
4363 expressions in the body are evaluated only for their side effects; and | |
4364 @code{save-excursion} itself is used only for its side effect (which | |
4365 is restoring the positions of point and mark). | |
4366 | |
4367 @need 1250 | |
4368 In more detail, the template for a @code{save-excursion} expression | |
4369 looks like this: | |
4370 | |
4371 @smallexample | |
4372 @group | |
4373 (save-excursion | |
4374 @var{first-expression-in-body} | |
4375 @var{second-expression-in-body} | |
4376 @var{third-expression-in-body} | |
4377 @dots{} | |
4378 @var{last-expression-in-body}) | |
4379 @end group | |
4380 @end smallexample | |
4381 | |
4382 @noindent | |
4383 An expression, of course, may be a symbol on its own or a list. | |
4384 | |
4385 In Emacs Lisp code, a @code{save-excursion} expression often occurs | |
4386 within the body of a @code{let} expression. It looks like this: | |
4387 | |
4388 @smallexample | |
4389 @group | |
4390 (let @var{varlist} | |
4391 (save-excursion | |
4392 @var{body}@dots{})) | |
4393 @end group | |
4394 @end smallexample | |
4395 | |
4396 @node Review, defun Exercises, save-excursion, Writing Defuns | |
4397 @comment node-name, next, previous, up | |
4398 @section Review | |
4399 | |
4400 In the last few chapters we have introduced a fair number of functions | |
4401 and special forms. Here they are described in brief, along with a few | |
4402 similar functions that have not been mentioned yet. | |
4403 | |
4404 @table @code | |
4405 @item eval-last-sexp | |
4406 Evaluate the last symbolic expression before the current location of | |
4407 point. The value is printed in the echo area unless the function is | |
4408 invoked with an argument; in that case, the output is printed in the | |
4409 current buffer. This command is normally bound to @kbd{C-x C-e}. | |
4410 | |
4411 @item defun | |
4412 Define function. This special form has up to five parts: the name, | |
4413 a template for the arguments that will be passed to the function, | |
4414 documentation, an optional interactive declaration, and the body of the | |
4415 definition. | |
4416 | |
4417 @need 1250 | |
4418 For example, in an early version of Emacs, the function definition was | |
4419 as follows. (It is slightly more complex now that it seeks the first | |
4420 non-whitespace character rather than the first visible character.) | |
4421 | |
4422 @smallexample | |
4423 @group | |
4424 (defun back-to-indentation () | |
4425 "Move point to first visible character on line." | |
4426 (interactive) | |
4427 (beginning-of-line 1) | |
4428 (skip-chars-forward " \t")) | |
4429 @end group | |
4430 @end smallexample | |
4431 | |
4432 @ignore | |
4433 In GNU Emacs 22, | |
4434 | |
4435 (defun backward-to-indentation (&optional arg) | |
4436 "Move backward ARG lines and position at first nonblank character." | |
4437 (interactive "p") | |
4438 (forward-line (- (or arg 1))) | |
4439 (skip-chars-forward " \t")) | |
4440 | |
4441 (defun back-to-indentation () | |
4442 "Move point to the first non-whitespace character on this line." | |
4443 (interactive) | |
4444 (beginning-of-line 1) | |
4445 (skip-syntax-forward " " (line-end-position)) | |
4446 ;; Move back over chars that have whitespace syntax but have the p flag. | |
4447 (backward-prefix-chars)) | |
4448 @end ignore | |
4449 | |
4450 @item interactive | |
4451 Declare to the interpreter that the function can be used | |
4452 interactively. This special form may be followed by a string with one | |
4453 or more parts that pass the information to the arguments of the | |
4454 function, in sequence. These parts may also tell the interpreter to | |
4455 prompt for information. Parts of the string are separated by | |
4456 newlines, @samp{\n}. | |
4457 | |
4458 @need 1000 | |
4459 Common code characters are: | |
4460 | |
4461 @table @code | |
4462 @item b | |
4463 The name of an existing buffer. | |
4464 | |
4465 @item f | |
4466 The name of an existing file. | |
4467 | |
4468 @item p | |
4469 The numeric prefix argument. (Note that this `p' is lower case.) | |
4470 | |
4471 @item r | |
4472 Point and the mark, as two numeric arguments, smallest first. This | |
4473 is the only code letter that specifies two successive arguments | |
4474 rather than one. | |
4475 @end table | |
4476 | |
4477 @xref{Interactive Codes, , Code Characters for @samp{interactive}, | |
4478 elisp, The GNU Emacs Lisp Reference Manual}, for a complete list of | |
4479 code characters. | |
4480 | |
4481 @item let | |
4482 Declare that a list of variables is for use within the body of the | |
4483 @code{let} and give them an initial value, either @code{nil} or a | |
4484 specified value; then evaluate the rest of the expressions in the body | |
4485 of the @code{let} and return the value of the last one. Inside the | |
4486 body of the @code{let}, the Lisp interpreter does not see the values of | |
4487 the variables of the same names that are bound outside of the | |
4488 @code{let}. | |
4489 | |
4490 @need 1250 | |
4491 For example, | |
4492 | |
4493 @smallexample | |
4494 @group | |
4495 (let ((foo (buffer-name)) | |
4496 (bar (buffer-size))) | |
4497 (message | |
4498 "This buffer is %s and has %d characters." | |
4499 foo bar)) | |
4500 @end group | |
4501 @end smallexample | |
4502 | |
4503 @item save-excursion | |
4504 Record the values of point and mark and the current buffer before | |
4505 evaluating the body of this special form. Restore the values of point | |
4506 and mark and buffer afterward. | |
4507 | |
4508 @need 1250 | |
4509 For example, | |
4510 | |
4511 @smallexample | |
4512 @group | |
4513 (message "We are %d characters into this buffer." | |
4514 (- (point) | |
4515 (save-excursion | |
4516 (goto-char (point-min)) (point)))) | |
4517 @end group | |
4518 @end smallexample | |
4519 | |
4520 @item if | |
4521 Evaluate the first argument to the function; if it is true, evaluate | |
4522 the second argument; else evaluate the third argument, if there is one. | |
4523 | |
4524 The @code{if} special form is called a @dfn{conditional}. There are | |
4525 other conditionals in Emacs Lisp, but @code{if} is perhaps the most | |
4526 commonly used. | |
4527 | |
4528 @need 1250 | |
4529 For example, | |
4530 | |
4531 @smallexample | |
4532 @group | |
4533 (if (= 22 emacs-major-version) | |
4534 (message "This is version 22 Emacs") | |
4535 (message "This is not version 22 Emacs")) | |
4536 @end group | |
4537 @end smallexample | |
4538 | |
4539 @need 1250 | |
4540 @item < | |
4541 @itemx > | |
4542 @itemx <= | |
4543 @itemx >= | |
4544 The @code{<} function tests whether its first argument is smaller than | |
4545 its second argument. A corresponding function, @code{>}, tests whether | |
4546 the first argument is greater than the second. Likewise, @code{<=} | |
4547 tests whether the first argument is less than or equal to the second and | |
4548 @code{>=} tests whether the first argument is greater than or equal to | |
4549 the second. In all cases, both arguments must be numbers or markers | |
4550 (markers indicate positions in buffers). | |
4551 | |
4552 @need 800 | |
4553 @item = | |
4554 The @code{=} function tests whether two arguments, both numbers or | |
4555 markers, are equal. | |
4556 | |
4557 @need 1250 | |
4558 @item equal | |
4559 @itemx eq | |
4560 Test whether two objects are the same. @code{equal} uses one meaning | |
4561 of the word `same' and @code{eq} uses another: @code{equal} returns | |
4562 true if the two objects have a similar structure and contents, such as | |
4563 two copies of the same book. On the other hand, @code{eq}, returns | |
4564 true if both arguments are actually the same object. | |
4565 @findex equal | |
4566 @findex eq | |
4567 | |
4568 @need 1250 | |
4569 @item string< | |
4570 @itemx string-lessp | |
4571 @itemx string= | |
4572 @itemx string-equal | |
4573 The @code{string-lessp} function tests whether its first argument is | |
4574 smaller than the second argument. A shorter, alternative name for the | |
4575 same function (a @code{defalias}) is @code{string<}. | |
4576 | |
4577 The arguments to @code{string-lessp} must be strings or symbols; the | |
4578 ordering is lexicographic, so case is significant. The print names of | |
4579 symbols are used instead of the symbols themselves. | |
4580 | |
4581 @cindex @samp{empty string} defined | |
4582 An empty string, @samp{""}, a string with no characters in it, is | |
4583 smaller than any string of characters. | |
4584 | |
4585 @code{string-equal} provides the corresponding test for equality. Its | |
4586 shorter, alternative name is @code{string=}. There are no string test | |
4587 functions that correspond to @var{>}, @code{>=}, or @code{<=}. | |
4588 | |
4589 @item message | |
4590 Print a message in the echo area. The first argument is a string that | |
4591 can contain @samp{%s}, @samp{%d}, or @samp{%c} to print the value of | |
4592 arguments that follow the string. The argument used by @samp{%s} must | |
4593 be a string or a symbol; the argument used by @samp{%d} must be a | |
4594 number. The argument used by @samp{%c} must be an @sc{ascii} code | |
4595 number; it will be printed as the character with that @sc{ascii} code. | |
4596 (Various other %-sequences have not been mentioned.) | |
4597 | |
4598 @item setq | |
4599 @itemx set | |
4600 The @code{setq} function sets the value of its first argument to the | |
4601 value of the second argument. The first argument is automatically | |
4602 quoted by @code{setq}. It does the same for succeeding pairs of | |
4603 arguments. Another function, @code{set}, takes only two arguments and | |
4604 evaluates both of them before setting the value returned by its first | |
4605 argument to the value returned by its second argument. | |
4606 | |
4607 @item buffer-name | |
4608 Without an argument, return the name of the buffer, as a string. | |
4609 | |
4610 @itemx buffer-file-name | |
4611 Without an argument, return the name of the file the buffer is | |
4612 visiting. | |
4613 | |
4614 @item current-buffer | |
4615 Return the buffer in which Emacs is active; it may not be | |
4616 the buffer that is visible on the screen. | |
4617 | |
4618 @item other-buffer | |
4619 Return the most recently selected buffer (other than the buffer passed | |
4620 to @code{other-buffer} as an argument and other than the current | |
4621 buffer). | |
4622 | |
4623 @item switch-to-buffer | |
4624 Select a buffer for Emacs to be active in and display it in the current | |
4625 window so users can look at it. Usually bound to @kbd{C-x b}. | |
4626 | |
4627 @item set-buffer | |
4628 Switch Emacs' attention to a buffer on which programs will run. Don't | |
4629 alter what the window is showing. | |
4630 | |
4631 @item buffer-size | |
4632 Return the number of characters in the current buffer. | |
4633 | |
4634 @item point | |
4635 Return the value of the current position of the cursor, as an | |
4636 integer counting the number of characters from the beginning of the | |
4637 buffer. | |
4638 | |
4639 @item point-min | |
4640 Return the minimum permissible value of point in | |
4641 the current buffer. This is 1, unless narrowing is in effect. | |
4642 | |
4643 @item point-max | |
4644 Return the value of the maximum permissible value of point in the | |
4645 current buffer. This is the end of the buffer, unless narrowing is in | |
4646 effect. | |
4647 @end table | |
4648 | |
4649 @need 1500 | |
4650 @node defun Exercises, , Review, Writing Defuns | |
4651 @section Exercises | |
4652 | |
4653 @itemize @bullet | |
4654 @item | |
4655 Write a non-interactive function that doubles the value of its | |
4656 argument, a number. Make that function interactive. | |
4657 | |
4658 @item | |
4659 Write a function that tests whether the current value of | |
4660 @code{fill-column} is greater than the argument passed to the function, | |
4661 and if so, prints an appropriate message. | |
4662 @end itemize | |
4663 | |
4664 @node Buffer Walk Through, More Complex, Writing Defuns, Top | |
4665 @comment node-name, next, previous, up | |
4666 @chapter A Few Buffer--Related Functions | |
4667 | |
4668 In this chapter we study in detail several of the functions used in GNU | |
4669 Emacs. This is called a ``walk-through''. These functions are used as | |
4670 examples of Lisp code, but are not imaginary examples; with the | |
4671 exception of the first, simplified function definition, these functions | |
4672 show the actual code used in GNU Emacs. You can learn a great deal from | |
4673 these definitions. The functions described here are all related to | |
4674 buffers. Later, we will study other functions. | |
4675 | |
4676 @menu | |
4677 * Finding More:: How to find more information. | |
4678 * simplified-beginning-of-buffer:: Shows @code{goto-char}, | |
4679 @code{point-min}, and @code{push-mark}. | |
4680 * mark-whole-buffer:: Almost the same as @code{beginning-of-buffer}. | |
4681 * append-to-buffer:: Uses @code{save-excursion} and | |
4682 @code{insert-buffer-substring}. | |
4683 * Buffer Related Review:: Review. | |
4684 * Buffer Exercises:: | |
4685 @end menu | |
4686 | |
4687 @node Finding More, simplified-beginning-of-buffer, Buffer Walk Through, Buffer Walk Through | |
4688 @section Finding More Information | |
4689 | |
4690 @findex describe-function, @r{introduced} | |
4691 @cindex Find function documentation | |
4692 In this walk-through, I will describe each new function as we come to | |
4693 it, sometimes in detail and sometimes briefly. If you are interested, | |
4694 you can get the full documentation of any Emacs Lisp function at any | |
4695 time by typing @kbd{C-h f} and then the name of the function (and then | |
4696 @key{RET}). Similarly, you can get the full documentation for a | |
4697 variable by typing @kbd{C-h v} and then the name of the variable (and | |
4698 then @key{RET}). | |
4699 | |
4700 @cindex Find source of function | |
4701 @c In version 22, tells location both of C and of Emacs Lisp | |
4702 Also, @code{describe-function} will tell you the location of the | |
4703 function definition. | |
4704 | |
4705 Put point into the name of the file that contains the function and | |
4706 press the @key{RET} key. In this case, @key{RET} means | |
4707 @code{push-button} rather than `return' or `enter'. Emacs will take | |
4708 you directly to the function definition. | |
4709 | |
4710 @ignore | |
4711 Not In version 22 | |
4712 | |
4713 If you move point over the file name and press | |
4714 the @key{RET} key, which in this case means @code{help-follow} rather | |
4715 than `return' or `enter', Emacs will take you directly to the function | |
4716 definition. | |
4717 @end ignore | |
4718 | |
4719 More generally, if you want to see a function in its original source | |
4720 file, you can use the @code{find-tags} function to jump to it. | |
4721 @code{find-tags} works with a wide variety of languages, not just | |
4722 Lisp, and C, and it works with non-programming text as well. For | |
4723 example, @code{find-tags} will jump to the various nodes in the | |
4724 Texinfo source file of this document. | |
4725 The @code{find-tags} function depends on `tags tables' that record | |
4726 the locations of the functions, variables, and other items to which | |
4727 @code{find-tags} jumps. | |
4728 | |
4729 To use the @code{find-tags} command, type @kbd{M-.} (i.e., press the | |
4730 period key while holding down the @key{META} key, or else type the | |
4731 @key{ESC} key and then type the period key), and then, at the prompt, | |
4732 type in the name of the function whose source code you want to see, | |
4733 such as @code{mark-whole-buffer}, and then type @key{RET}. Emacs will | |
4734 switch buffers and display the source code for the function on your | |
4735 screen. To switch back to your current buffer, type @kbd{C-x b | |
4736 @key{RET}}. (On some keyboards, the @key{META} key is labelled | |
4737 @key{ALT}.) | |
4738 | |
4739 @c !!! 22.1.1 tags table location in this paragraph | |
4740 @cindex TAGS table, specifying | |
4741 @findex find-tags | |
4742 Depending on how the initial default values of your copy of Emacs are | |
4743 set, you may also need to specify the location of your `tags table', | |
4744 which is a file called @file{TAGS}. For example, if you are | |
4745 interested in Emacs sources, the tags table you will most likely want, | |
4746 if it has already been created for you, will be in a subdirectory of | |
4747 the @file{/usr/local/share/emacs/} directory; thus you would use the | |
4748 @code{M-x visit-tags-table} command and specify a pathname such as | |
4749 @file{/usr/local/share/emacs/22.1.1/lisp/TAGS}. If the tags table | |
4750 has not already been created, you will have to create it yourself. It | |
4751 will in a file such as @file{/usr/local/src/emacs/src/TAGS}. | |
4752 | |
4753 @need 1250 | |
4754 To create a @file{TAGS} file in a specific directory, switch to that | |
4755 directory in Emacs using @kbd{M-x cd} command, or list the directory | |
4756 with @kbd{C-x d} (@code{dired}). Then run the compile command, with | |
4757 @w{@code{etags *.el}} as the command to execute: | |
4758 | |
4759 @smallexample | |
4760 M-x compile RET etags *.el RET | |
4761 @end smallexample | |
4762 | |
4763 For more information, see @ref{etags, , Create Your Own @file{TAGS} File}. | |
4764 | |
4765 After you become more familiar with Emacs Lisp, you will find that you will | |
4766 frequently use @code{find-tags} to navigate your way around source code; | |
4767 and you will create your own @file{TAGS} tables. | |
4768 | |
4769 @cindex Library, as term for `file' | |
4770 Incidentally, the files that contain Lisp code are conventionally | |
4771 called @dfn{libraries}. The metaphor is derived from that of a | |
4772 specialized library, such as a law library or an engineering library, | |
4773 rather than a general library. Each library, or file, contains | |
4774 functions that relate to a particular topic or activity, such as | |
4775 @file{abbrev.el} for handling abbreviations and other typing | |
4776 shortcuts, and @file{help.el} for on-line help. (Sometimes several | |
4777 libraries provide code for a single activity, as the various | |
4778 @file{rmail@dots{}} files provide code for reading electronic mail.) | |
4779 In @cite{The GNU Emacs Manual}, you will see sentences such as ``The | |
4780 @kbd{C-h p} command lets you search the standard Emacs Lisp libraries | |
4781 by topic keywords.'' | |
4782 | |
4783 @node simplified-beginning-of-buffer, mark-whole-buffer, Finding More, Buffer Walk Through | |
4784 @comment node-name, next, previous, up | |
4785 @section A Simplified @code{beginning-of-buffer} Definition | |
4786 @findex simplified-beginning-of-buffer | |
4787 | |
4788 The @code{beginning-of-buffer} command is a good function to start with | |
4789 since you are likely to be familiar with it and it is easy to | |
4790 understand. Used as an interactive command, @code{beginning-of-buffer} | |
4791 moves the cursor to the beginning of the buffer, leaving the mark at the | |
4792 previous position. It is generally bound to @kbd{M-<}. | |
4793 | |
4794 In this section, we will discuss a shortened version of the function | |
4795 that shows how it is most frequently used. This shortened function | |
4796 works as written, but it does not contain the code for a complex option. | |
4797 In another section, we will describe the entire function. | |
4798 (@xref{beginning-of-buffer, , Complete Definition of | |
4799 @code{beginning-of-buffer}}.) | |
4800 | |
4801 Before looking at the code, let's consider what the function | |
4802 definition has to contain: it must include an expression that makes | |
4803 the function interactive so it can be called by typing @kbd{M-x | |
4804 beginning-of-buffer} or by typing a keychord such as @kbd{M-<}; it | |
4805 must include code to leave a mark at the original position in the | |
4806 buffer; and it must include code to move the cursor to the beginning | |
4807 of the buffer. | |
4808 | |
4809 @need 1250 | |
4810 Here is the complete text of the shortened version of the function: | |
4811 | |
4812 @smallexample | |
4813 @group | |
4814 (defun simplified-beginning-of-buffer () | |
4815 "Move point to the beginning of the buffer; | |
4816 leave mark at previous position." | |
4817 (interactive) | |
4818 (push-mark) | |
4819 (goto-char (point-min))) | |
4820 @end group | |
4821 @end smallexample | |
4822 | |
4823 Like all function definitions, this definition has five parts following | |
4824 the special form @code{defun}: | |
4825 | |
4826 @enumerate | |
4827 @item | |
4828 The name: in this example, @code{simplified-beginning-of-buffer}. | |
4829 | |
4830 @item | |
4831 A list of the arguments: in this example, an empty list, @code{()}, | |
4832 | |
4833 @item | |
4834 The documentation string. | |
4835 | |
4836 @item | |
4837 The interactive expression. | |
4838 | |
4839 @item | |
4840 The body. | |
4841 @end enumerate | |
4842 | |
4843 @noindent | |
4844 In this function definition, the argument list is empty; this means that | |
4845 this function does not require any arguments. (When we look at the | |
4846 definition for the complete function, we will see that it may be passed | |
4847 an optional argument.) | |
4848 | |
4849 The interactive expression tells Emacs that the function is intended to | |
4850 be used interactively. In this example, @code{interactive} does not have | |
4851 an argument because @code{simplified-beginning-of-buffer} does not | |
4852 require one. | |
4853 | |
4854 @need 800 | |
4855 The body of the function consists of the two lines: | |
4856 | |
4857 @smallexample | |
4858 @group | |
4859 (push-mark) | |
4860 (goto-char (point-min)) | |
4861 @end group | |
4862 @end smallexample | |
4863 | |
4864 The first of these lines is the expression, @code{(push-mark)}. When | |
4865 this expression is evaluated by the Lisp interpreter, it sets a mark at | |
4866 the current position of the cursor, wherever that may be. The position | |
4867 of this mark is saved in the mark ring. | |
4868 | |
4869 The next line is @code{(goto-char (point-min))}. This expression | |
4870 jumps the cursor to the minimum point in the buffer, that is, to the | |
4871 beginning of the buffer (or to the beginning of the accessible portion | |
4872 of the buffer if it is narrowed. @xref{Narrowing & Widening, , | |
4873 Narrowing and Widening}.) | |
4874 | |
4875 The @code{push-mark} command sets a mark at the place where the cursor | |
4876 was located before it was moved to the beginning of the buffer by the | |
4877 @code{(goto-char (point-min))} expression. Consequently, you can, if | |
4878 you wish, go back to where you were originally by typing @kbd{C-x C-x}. | |
4879 | |
4880 That is all there is to the function definition! | |
4881 | |
4882 @findex describe-function | |
4883 When you are reading code such as this and come upon an unfamiliar | |
4884 function, such as @code{goto-char}, you can find out what it does by | |
4885 using the @code{describe-function} command. To use this command, type | |
4886 @kbd{C-h f} and then type in the name of the function and press | |
4887 @key{RET}. The @code{describe-function} command will print the | |
4888 function's documentation string in a @file{*Help*} window. For | |
4889 example, the documentation for @code{goto-char} is: | |
4890 | |
4891 @smallexample | |
4892 @group | |
4893 Set point to POSITION, a number or marker. | |
4894 Beginning of buffer is position (point-min), end is (point-max). | |
4895 @end group | |
4896 @end smallexample | |
4897 | |
4898 @noindent | |
4899 The function's one argument is the desired position. | |
4900 | |
4901 @noindent | |
4902 (The prompt for @code{describe-function} will offer you the symbol | |
4903 under or preceding the cursor, so you can save typing by positioning | |
4904 the cursor right over or after the function and then typing @kbd{C-h f | |
4905 @key{RET}}.) | |
4906 | |
4907 The @code{end-of-buffer} function definition is written in the same way as | |
4908 the @code{beginning-of-buffer} definition except that the body of the | |
4909 function contains the expression @code{(goto-char (point-max))} in place | |
4910 of @code{(goto-char (point-min))}. | |
4911 | |
4912 @node mark-whole-buffer, append-to-buffer, simplified-beginning-of-buffer, Buffer Walk Through | |
4913 @comment node-name, next, previous, up | |
4914 @section The Definition of @code{mark-whole-buffer} | |
4915 @findex mark-whole-buffer | |
4916 | |
4917 The @code{mark-whole-buffer} function is no harder to understand than the | |
4918 @code{simplified-beginning-of-buffer} function. In this case, however, | |
4919 we will look at the complete function, not a shortened version. | |
4920 | |
4921 The @code{mark-whole-buffer} function is not as commonly used as the | |
4922 @code{beginning-of-buffer} function, but is useful nonetheless: it | |
4923 marks a whole buffer as a region by putting point at the beginning and | |
4924 a mark at the end of the buffer. It is generally bound to @kbd{C-x | |
4925 h}. | |
4926 | |
4927 @menu | |
4928 * mark-whole-buffer overview:: | |
4929 * Body of mark-whole-buffer:: Only three lines of code. | |
4930 @end menu | |
4931 | |
4932 @node mark-whole-buffer overview, Body of mark-whole-buffer, mark-whole-buffer, mark-whole-buffer | |
4933 @ifnottex | |
4934 @unnumberedsubsec An overview of @code{mark-whole-buffer} | |
4935 @end ifnottex | |
4936 | |
4937 @need 1250 | |
4938 In GNU Emacs 22, the code for the complete function looks like this: | |
4939 | |
4940 @smallexample | |
4941 @group | |
4942 (defun mark-whole-buffer () | |
4943 "Put point at beginning and mark at end of buffer. | |
4944 You probably should not use this function in Lisp programs; | |
4945 it is usually a mistake for a Lisp function to use any subroutine | |
4946 that uses or sets the mark." | |
4947 (interactive) | |
4948 (push-mark (point)) | |
4949 (push-mark (point-max) nil t) | |
4950 (goto-char (point-min))) | |
4951 @end group | |
4952 @end smallexample | |
4953 | |
4954 @need 1250 | |
4955 Like all other functions, the @code{mark-whole-buffer} function fits | |
4956 into the template for a function definition. The template looks like | |
4957 this: | |
4958 | |
4959 @smallexample | |
4960 @group | |
4961 (defun @var{name-of-function} (@var{argument-list}) | |
4962 "@var{documentation}@dots{}" | |
4963 (@var{interactive-expression}@dots{}) | |
4964 @var{body}@dots{}) | |
4965 @end group | |
4966 @end smallexample | |
4967 | |
4968 Here is how the function works: the name of the function is | |
4969 @code{mark-whole-buffer}; it is followed by an empty argument list, | |
4970 @samp{()}, which means that the function does not require arguments. | |
4971 The documentation comes next. | |
4972 | |
4973 The next line is an @code{(interactive)} expression that tells Emacs | |
4974 that the function will be used interactively. These details are similar | |
4975 to the @code{simplified-beginning-of-buffer} function described in the | |
4976 previous section. | |
4977 | |
4978 @need 1250 | |
4979 @node Body of mark-whole-buffer, , mark-whole-buffer overview, mark-whole-buffer | |
4980 @comment node-name, next, previous, up | |
4981 @subsection Body of @code{mark-whole-buffer} | |
4982 | |
4983 The body of the @code{mark-whole-buffer} function consists of three | |
4984 lines of code: | |
4985 | |
4986 @c GNU Emacs 22 | |
4987 @smallexample | |
4988 @group | |
4989 (push-mark (point)) | |
4990 (push-mark (point-max) nil t) | |
4991 (goto-char (point-min)) | |
4992 @end group | |
4993 @end smallexample | |
4994 | |
4995 The first of these lines is the expression, @code{(push-mark (point))}. | |
4996 | |
4997 This line does exactly the same job as the first line of the body of | |
4998 the @code{simplified-beginning-of-buffer} function, which is written | |
4999 @code{(push-mark)}. In both cases, the Lisp interpreter sets a mark | |
5000 at the current position of the cursor. | |
5001 | |
5002 I don't know why the expression in @code{mark-whole-buffer} is written | |
5003 @code{(push-mark (point))} and the expression in | |
5004 @code{beginning-of-buffer} is written @code{(push-mark)}. Perhaps | |
5005 whoever wrote the code did not know that the arguments for | |
5006 @code{push-mark} are optional and that if @code{push-mark} is not | |
5007 passed an argument, the function automatically sets mark at the | |
5008 location of point by default. Or perhaps the expression was written | |
5009 so as to parallel the structure of the next line. In any case, the | |
5010 line causes Emacs to determine the position of point and set a mark | |
5011 there. | |
5012 | |
5013 In earlier versions of GNU Emacs, the next line of | |
5014 @code{mark-whole-buffer} was @code{(push-mark (point-max))}. This | |
5015 expression sets a mark at the point in the buffer that has the highest | |
5016 number. This will be the end of the buffer (or, if the buffer is | |
5017 narrowed, the end of the accessible portion of the buffer. | |
5018 @xref{Narrowing & Widening, , Narrowing and Widening}, for more about | |
5019 narrowing.) After this mark has been set, the previous mark, the one | |
5020 set at point, is no longer set, but Emacs remembers its position, just | |
5021 as all other recent marks are always remembered. This means that you | |
5022 can, if you wish, go back to that position by typing @kbd{C-u | |
5023 C-@key{SPC}} twice. | |
5024 | |
5025 @need 1250 | |
5026 In GNU Emacs 22, the @code{(point-max)} is slightly more complicated. | |
5027 The line reads | |
5028 | |
5029 @smallexample | |
5030 (push-mark (point-max) nil t) | |
5031 @end smallexample | |
5032 | |
5033 @noindent | |
5034 The expression works nearly the same as before. It sets a mark at the | |
5035 highest numbered place in the buffer that it can. However, in this | |
5036 version, @code{push-mark} has two additional arguments. The second | |
5037 argument to @code{push-mark} is @code{nil}. This tells the function | |
5038 it @emph{should} display a message that says `Mark set' when it pushes | |
5039 the mark. The third argument is @code{t}. This tells | |
5040 @code{push-mark} to activate the mark when Transient Mark mode is | |
5041 turned on. Transient Mark mode highlights the currently active | |
5042 region. It is often turned off. | |
5043 | |
5044 Finally, the last line of the function is @code{(goto-char | |
5045 (point-min)))}. This is written exactly the same way as it is written | |
5046 in @code{beginning-of-buffer}. The expression moves the cursor to | |
5047 the minimum point in the buffer, that is, to the beginning of the buffer | |
5048 (or to the beginning of the accessible portion of the buffer). As a | |
5049 result of this, point is placed at the beginning of the buffer and mark | |
5050 is set at the end of the buffer. The whole buffer is, therefore, the | |
5051 region. | |
5052 | |
5053 @node append-to-buffer, Buffer Related Review, mark-whole-buffer, Buffer Walk Through | |
5054 @comment node-name, next, previous, up | |
5055 @section The Definition of @code{append-to-buffer} | |
5056 @findex append-to-buffer | |
5057 | |
5058 The @code{append-to-buffer} command is more complex than the | |
5059 @code{mark-whole-buffer} command. What it does is copy the region | |
5060 (that is, the part of the buffer between point and mark) from the | |
5061 current buffer to a specified buffer. | |
5062 | |
5063 @menu | |
5064 * append-to-buffer overview:: | |
5065 * append interactive:: A two part interactive expression. | |
5066 * append-to-buffer body:: Incorporates a @code{let} expression. | |
5067 * append save-excursion:: How the @code{save-excursion} works. | |
5068 @end menu | |
5069 | |
5070 @node append-to-buffer overview, append interactive, append-to-buffer, append-to-buffer | |
5071 @ifnottex | |
5072 @unnumberedsubsec An Overview of @code{append-to-buffer} | |
5073 @end ifnottex | |
5074 | |
5075 @findex insert-buffer-substring | |
5076 The @code{append-to-buffer} command uses the | |
5077 @code{insert-buffer-substring} function to copy the region. | |
5078 @code{insert-buffer-substring} is described by its name: it takes a | |
5079 string of characters from part of a buffer, a ``substring'', and | |
5080 inserts them into another buffer. | |
5081 | |
5082 Most of @code{append-to-buffer} is | |
5083 concerned with setting up the conditions for | |
5084 @code{insert-buffer-substring} to work: the code must specify both the | |
5085 buffer to which the text will go, the window it comes from and goes | |
5086 to, and the region that will be copied. | |
5087 | |
5088 @need 1250 | |
5089 Here is the complete text of the function: | |
5090 | |
5091 @smallexample | |
5092 @group | |
5093 (defun append-to-buffer (buffer start end) | |
5094 "Append to specified buffer the text of the region. | |
5095 It is inserted into that buffer before its point. | |
5096 @end group | |
5097 | |
5098 @group | |
5099 When calling from a program, give three arguments: | |
5100 BUFFER (or buffer name), START and END. | |
5101 START and END specify the portion of the current buffer to be copied." | |
5102 (interactive | |
5103 (list (read-buffer "Append to buffer: " (other-buffer | |
5104 (current-buffer) t)) | |
5105 (region-beginning) (region-end))) | |
5106 @end group | |
5107 @group | |
5108 (let ((oldbuf (current-buffer))) | |
5109 (save-excursion | |
5110 (let* ((append-to (get-buffer-create buffer)) | |
5111 (windows (get-buffer-window-list append-to t t)) | |
5112 point) | |
5113 (set-buffer append-to) | |
5114 (setq point (point)) | |
5115 (barf-if-buffer-read-only) | |
5116 (insert-buffer-substring oldbuf start end) | |
5117 (dolist (window windows) | |
5118 (when (= (window-point window) point) | |
5119 (set-window-point window (point)))))))) | |
5120 @end group | |
5121 @end smallexample | |
5122 | |
5123 The function can be understood by looking at it as a series of | |
5124 filled-in templates. | |
5125 | |
5126 The outermost template is for the function definition. In this | |
5127 function, it looks like this (with several slots filled in): | |
5128 | |
5129 @smallexample | |
5130 @group | |
5131 (defun append-to-buffer (buffer start end) | |
5132 "@var{documentation}@dots{}" | |
5133 (interactive @dots{}) | |
5134 @var{body}@dots{}) | |
5135 @end group | |
5136 @end smallexample | |
5137 | |
5138 The first line of the function includes its name and three arguments. | |
5139 The arguments are the @code{buffer} to which the text will be copied, and | |
5140 the @code{start} and @code{end} of the region in the current buffer that | |
5141 will be copied. | |
5142 | |
5143 The next part of the function is the documentation, which is clear and | |
5144 complete. As is conventional, the three arguments are written in | |
5145 upper case so you will notice them easily. Even better, they are | |
5146 described in the same order as in the argument list. | |
5147 | |
5148 Note that the documentation distinguishes between a buffer and its | |
5149 name. (The function can handle either.) | |
5150 | |
5151 @node append interactive, append-to-buffer body, append-to-buffer overview, append-to-buffer | |
5152 @comment node-name, next, previous, up | |
5153 @subsection The @code{append-to-buffer} Interactive Expression | |
5154 | |
5155 Since the @code{append-to-buffer} function will be used interactively, | |
5156 the function must have an @code{interactive} expression. (For a | |
5157 review of @code{interactive}, see @ref{Interactive, , Making a | |
5158 Function Interactive}.) The expression reads as follows: | |
5159 | |
5160 @smallexample | |
5161 @group | |
5162 (interactive | |
5163 (list (read-buffer | |
5164 "Append to buffer: " | |
5165 (other-buffer (current-buffer) t)) | |
5166 (region-beginning) | |
5167 (region-end))) | |
5168 @end group | |
5169 @end smallexample | |
5170 | |
5171 @noindent | |
5172 This expression is not one with letters standing for parts, as | |
5173 described earlier. Instead, it starts a list with these parts: | |
5174 | |
5175 The first part of the list is an expression to read the name of a | |
5176 buffer and return it as a string. That is @code{read-buffer}. The | |
5177 function requires a prompt as its first argument, @samp{"Append to | |
5178 buffer: "}. Its second argument tells the command what value to | |
5179 provide if you don't specify anything. | |
5180 | |
5181 In this case that second argument is an expression containing the | |
5182 function @code{other-buffer}, an exception, and a @samp{t}, standing | |
5183 for true. | |
5184 | |
5185 The first argument to @code{other-buffer}, the exception, is yet | |
5186 another function, @code{current-buffer}. That is not going to be | |
5187 returned. The second argument is the symbol for true, @code{t}. that | |
5188 tells @code{other-buffer} that it may show visible buffers (except in | |
5189 this case, it will not show the current buffer, which makes sense). | |
5190 | |
5191 @need 1250 | |
5192 The expression looks like this: | |
5193 | |
5194 @smallexample | |
5195 (other-buffer (current-buffer) t) | |
5196 @end smallexample | |
5197 | |
5198 The second and third arguments to the @code{list} expression are | |
5199 @code{(region-beginning)} and @code{(region-end)}. These two | |
5200 functions specify the beginning and end of the text to be appended. | |
5201 | |
5202 @need 1250 | |
5203 Originally, the command used the letters @samp{B} and @samp{r}. | |
5204 The whole @code{interactive} expression looked like this: | |
5205 | |
5206 @smallexample | |
5207 (interactive "BAppend to buffer:@: \nr") | |
5208 @end smallexample | |
5209 | |
5210 @noindent | |
5211 But when that was done, the default value of the buffer switched to | |
5212 was invisible. That was not wanted. | |
5213 | |
5214 (The prompt was separated from the second argument with a newline, | |
5215 @samp{\n}. It was followed by an @samp{r} that told Emacs to bind the | |
5216 two arguments that follow the symbol @code{buffer} in the function's | |
5217 argument list (that is, @code{start} and @code{end}) to the values of | |
5218 point and mark. That argument worked fine.) | |
5219 | |
5220 @node append-to-buffer body, append save-excursion, append interactive, append-to-buffer | |
5221 @comment node-name, next, previous, up | |
5222 @subsection The Body of @code{append-to-buffer} | |
5223 | |
5224 @ignore | |
5225 in GNU Emacs 22 in /usr/local/src/emacs/lisp/simple.el | |
5226 | |
5227 (defun append-to-buffer (buffer start end) | |
5228 "Append to specified buffer the text of the region. | |
5229 It is inserted into that buffer before its point. | |
5230 | |
5231 When calling from a program, give three arguments: | |
5232 BUFFER (or buffer name), START and END. | |
5233 START and END specify the portion of the current buffer to be copied." | |
5234 (interactive | |
5235 (list (read-buffer "Append to buffer: " (other-buffer (current-buffer) t)) | |
5236 (region-beginning) (region-end))) | |
5237 (let ((oldbuf (current-buffer))) | |
5238 (save-excursion | |
5239 (let* ((append-to (get-buffer-create buffer)) | |
5240 (windows (get-buffer-window-list append-to t t)) | |
5241 point) | |
5242 (set-buffer append-to) | |
5243 (setq point (point)) | |
5244 (barf-if-buffer-read-only) | |
5245 (insert-buffer-substring oldbuf start end) | |
5246 (dolist (window windows) | |
5247 (when (= (window-point window) point) | |
5248 (set-window-point window (point)))))))) | |
5249 @end ignore | |
5250 | |
5251 The body of the @code{append-to-buffer} function begins with @code{let}. | |
5252 | |
5253 As we have seen before (@pxref{let, , @code{let}}), the purpose of a | |
5254 @code{let} expression is to create and give initial values to one or | |
5255 more variables that will only be used within the body of the | |
5256 @code{let}. This means that such a variable will not be confused with | |
5257 any variable of the same name outside the @code{let} expression. | |
5258 | |
5259 We can see how the @code{let} expression fits into the function as a | |
5260 whole by showing a template for @code{append-to-buffer} with the | |
5261 @code{let} expression in outline: | |
5262 | |
5263 @smallexample | |
5264 @group | |
5265 (defun append-to-buffer (buffer start end) | |
5266 "@var{documentation}@dots{}" | |
5267 (interactive @dots{}) | |
5268 (let ((@var{variable} @var{value})) | |
5269 @var{body}@dots{}) | |
5270 @end group | |
5271 @end smallexample | |
5272 | |
5273 The @code{let} expression has three elements: | |
5274 | |
5275 @enumerate | |
5276 @item | |
5277 The symbol @code{let}; | |
5278 | |
5279 @item | |
5280 A varlist containing, in this case, a single two-element list, | |
5281 @code{(@var{variable} @var{value})}; | |
5282 | |
5283 @item | |
5284 The body of the @code{let} expression. | |
5285 @end enumerate | |
5286 | |
5287 @need 800 | |
5288 In the @code{append-to-buffer} function, the varlist looks like this: | |
5289 | |
5290 @smallexample | |
5291 (oldbuf (current-buffer)) | |
5292 @end smallexample | |
5293 | |
5294 @noindent | |
5295 In this part of the @code{let} expression, the one variable, | |
5296 @code{oldbuf}, is bound to the value returned by the | |
5297 @code{(current-buffer)} expression. The variable, @code{oldbuf}, is | |
5298 used to keep track of the buffer in which you are working and from | |
5299 which you will copy. | |
5300 | |
5301 The element or elements of a varlist are surrounded by a set of | |
5302 parentheses so the Lisp interpreter can distinguish the varlist from | |
5303 the body of the @code{let}. As a consequence, the two-element list | |
5304 within the varlist is surrounded by a circumscribing set of parentheses. | |
5305 The line looks like this: | |
5306 | |
5307 @smallexample | |
5308 @group | |
5309 (let ((oldbuf (current-buffer))) | |
5310 @dots{} ) | |
5311 @end group | |
5312 @end smallexample | |
5313 | |
5314 @noindent | |
5315 The two parentheses before @code{oldbuf} might surprise you if you did | |
5316 not realize that the first parenthesis before @code{oldbuf} marks the | |
5317 boundary of the varlist and the second parenthesis marks the beginning | |
5318 of the two-element list, @code{(oldbuf (current-buffer))}. | |
5319 | |
5320 @node append save-excursion, , append-to-buffer body, append-to-buffer | |
5321 @comment node-name, next, previous, up | |
5322 @subsection @code{save-excursion} in @code{append-to-buffer} | |
5323 | |
5324 The body of the @code{let} expression in @code{append-to-buffer} | |
5325 consists of a @code{save-excursion} expression. | |
5326 | |
5327 The @code{save-excursion} function saves the locations of point and | |
5328 mark, and restores them to those positions after the expressions in the | |
5329 body of the @code{save-excursion} complete execution. In addition, | |
5330 @code{save-excursion} keeps track of the original buffer, and | |
5331 restores it. This is how @code{save-excursion} is used in | |
5332 @code{append-to-buffer}. | |
5333 | |
5334 @need 1500 | |
5335 @cindex Indentation for formatting | |
5336 @cindex Formatting convention | |
5337 Incidentally, it is worth noting here that a Lisp function is normally | |
5338 formatted so that everything that is enclosed in a multi-line spread is | |
5339 indented more to the right than the first symbol. In this function | |
5340 definition, the @code{let} is indented more than the @code{defun}, and | |
5341 the @code{save-excursion} is indented more than the @code{let}, like | |
5342 this: | |
5343 | |
5344 @smallexample | |
5345 @group | |
5346 (defun @dots{} | |
5347 @dots{} | |
5348 @dots{} | |
5349 (let@dots{} | |
5350 (save-excursion | |
5351 @dots{} | |
5352 @end group | |
5353 @end smallexample | |
5354 | |
5355 @need 1500 | |
5356 @noindent | |
5357 This formatting convention makes it easy to see that the lines in | |
5358 the body of the @code{save-excursion} are enclosed by the parentheses | |
5359 associated with @code{save-excursion}, just as the | |
5360 @code{save-excursion} itself is enclosed by the parentheses associated | |
5361 with the @code{let}: | |
5362 | |
5363 @smallexample | |
5364 @group | |
5365 (let ((oldbuf (current-buffer))) | |
5366 (save-excursion | |
5367 @dots{} | |
5368 (set-buffer @dots{}) | |
5369 (insert-buffer-substring oldbuf start end) | |
5370 @dots{})) | |
5371 @end group | |
5372 @end smallexample | |
5373 | |
5374 @need 1200 | |
5375 The use of the @code{save-excursion} function can be viewed as a process | |
5376 of filling in the slots of a template: | |
5377 | |
5378 @smallexample | |
5379 @group | |
5380 (save-excursion | |
5381 @var{first-expression-in-body} | |
5382 @var{second-expression-in-body} | |
5383 @dots{} | |
5384 @var{last-expression-in-body}) | |
5385 @end group | |
5386 @end smallexample | |
5387 | |
5388 @need 1200 | |
5389 @noindent | |
5390 In this function, the body of the @code{save-excursion} contains only | |
5391 one expression, the @code{let*} expression. You know about a | |
5392 @code{let} function. The @code{let*} function is different. It has a | |
5393 @samp{*} in its name. It enables Emacs to set each variable in its | |
5394 varlist in sequence, one after another. | |
5395 | |
5396 Its critical feature is that variables later in the varlist can make | |
5397 use of the values to which Emacs set variables earlier in the varlist. | |
5398 @xref{fwd-para let, , The @code{let*} expression}. | |
5399 | |
5400 We will skip functions like @code{let*} and focus on two: the | |
5401 @code{set-buffer} function and the @code{insert-buffer-substring} | |
5402 function. | |
5403 | |
5404 @need 1250 | |
5405 In the old days, the @code{set-buffer} expression was simply | |
5406 | |
5407 @smallexample | |
5408 (set-buffer (get-buffer-create buffer)) | |
5409 @end smallexample | |
5410 | |
5411 @need 1250 | |
5412 @noindent | |
5413 but now it is | |
5414 | |
5415 @smallexample | |
5416 (set-buffer append-to) | |
5417 @end smallexample | |
5418 | |
5419 @noindent | |
5420 @code{append-to} is bound to @code{(get-buffer-create buffer)} earlier | |
5421 on in the @code{let*} expression. That extra binding would not be | |
5422 necessary except for that @code{append-to} is used later in the | |
5423 varlist as an argument to @code{get-buffer-window-list}. | |
5424 | |
5425 @ignore | |
5426 in GNU Emacs 22 | |
5427 | |
5428 (let ((oldbuf (current-buffer))) | |
5429 (save-excursion | |
5430 (let* ((append-to (get-buffer-create buffer)) | |
5431 (windows (get-buffer-window-list append-to t t)) | |
5432 point) | |
5433 (set-buffer append-to) | |
5434 (setq point (point)) | |
5435 (barf-if-buffer-read-only) | |
5436 (insert-buffer-substring oldbuf start end) | |
5437 (dolist (window windows) | |
5438 (when (= (window-point window) point) | |
5439 (set-window-point window (point)))))))) | |
5440 @end ignore | |
5441 | |
5442 The @code{append-to-buffer} function definition inserts text from the | |
5443 buffer in which you are currently to a named buffer. It happens that | |
5444 @code{insert-buffer-substring} copies text from another buffer to the | |
5445 current buffer, just the reverse---that is why the | |
5446 @code{append-to-buffer} definition starts out with a @code{let} that | |
5447 binds the local symbol @code{oldbuf} to the value returned by | |
5448 @code{current-buffer}. | |
5449 | |
5450 @need 1250 | |
5451 The @code{insert-buffer-substring} expression looks like this: | |
5452 | |
5453 @smallexample | |
5454 (insert-buffer-substring oldbuf start end) | |
5455 @end smallexample | |
5456 | |
5457 @noindent | |
5458 The @code{insert-buffer-substring} function copies a string | |
5459 @emph{from} the buffer specified as its first argument and inserts the | |
5460 string into the present buffer. In this case, the argument to | |
5461 @code{insert-buffer-substring} is the value of the variable created | |
5462 and bound by the @code{let}, namely the value of @code{oldbuf}, which | |
5463 was the current buffer when you gave the @code{append-to-buffer} | |
5464 command. | |
5465 | |
5466 After @code{insert-buffer-substring} has done its work, | |
5467 @code{save-excursion} will restore the action to the original buffer | |
5468 and @code{append-to-buffer} will have done its job. | |
5469 | |
5470 @need 800 | |
5471 Written in skeletal form, the workings of the body look like this: | |
5472 | |
5473 @smallexample | |
5474 @group | |
5475 (let (@var{bind-}@code{oldbuf}@var{-to-value-of-}@code{current-buffer}) | |
5476 (save-excursion ; @r{Keep track of buffer.} | |
5477 @var{change-buffer} | |
5478 @var{insert-substring-from-}@code{oldbuf}@var{-into-buffer}) | |
5479 | |
5480 @var{change-back-to-original-buffer-when-finished} | |
5481 @var{let-the-local-meaning-of-}@code{oldbuf}@var{-disappear-when-finished} | |
5482 @end group | |
5483 @end smallexample | |
5484 | |
5485 In summary, @code{append-to-buffer} works as follows: it saves the | |
5486 value of the current buffer in the variable called @code{oldbuf}. It | |
5487 gets the new buffer (creating one if need be) and switches Emacs' | |
5488 attention to it. Using the value of @code{oldbuf}, it inserts the | |
5489 region of text from the old buffer into the new buffer; and then using | |
5490 @code{save-excursion}, it brings you back to your original buffer. | |
5491 | |
5492 In looking at @code{append-to-buffer}, you have explored a fairly | |
5493 complex function. It shows how to use @code{let} and | |
5494 @code{save-excursion}, and how to change to and come back from another | |
5495 buffer. Many function definitions use @code{let}, | |
5496 @code{save-excursion}, and @code{set-buffer} this way. | |
5497 | |
5498 @node Buffer Related Review, Buffer Exercises, append-to-buffer, Buffer Walk Through | |
5499 @comment node-name, next, previous, up | |
5500 @section Review | |
5501 | |
5502 Here is a brief summary of the various functions discussed in this chapter. | |
5503 | |
5504 @table @code | |
5505 @item describe-function | |
5506 @itemx describe-variable | |
5507 Print the documentation for a function or variable. | |
5508 Conventionally bound to @kbd{C-h f} and @kbd{C-h v}. | |
5509 | |
5510 @item find-tag | |
5511 Find the file containing the source for a function or variable and | |
5512 switch buffers to it, positioning point at the beginning of the item. | |
5513 Conventionally bound to @kbd{M-.} (that's a period following the | |
5514 @key{META} key). | |
5515 | |
5516 @item save-excursion | |
5517 Save the location of point and mark and restore their values after the | |
5518 arguments to @code{save-excursion} have been evaluated. Also, remember | |
5519 the current buffer and return to it. | |
5520 | |
5521 @item push-mark | |
5522 Set mark at a location and record the value of the previous mark on the | |
5523 mark ring. The mark is a location in the buffer that will keep its | |
5524 relative position even if text is added to or removed from the buffer. | |
5525 | |
5526 @item goto-char | |
5527 Set point to the location specified by the value of the argument, which | |
5528 can be a number, a marker, or an expression that returns the number of | |
5529 a position, such as @code{(point-min)}. | |
5530 | |
5531 @item insert-buffer-substring | |
5532 Copy a region of text from a buffer that is passed to the function as | |
5533 an argument and insert the region into the current buffer. | |
5534 | |
5535 @item mark-whole-buffer | |
5536 Mark the whole buffer as a region. Normally bound to @kbd{C-x h}. | |
5537 | |
5538 @item set-buffer | |
5539 Switch the attention of Emacs to another buffer, but do not change the | |
5540 window being displayed. Used when the program rather than a human is | |
5541 to work on a different buffer. | |
5542 | |
5543 @item get-buffer-create | |
5544 @itemx get-buffer | |
5545 Find a named buffer or create one if a buffer of that name does not | |
5546 exist. The @code{get-buffer} function returns @code{nil} if the named | |
5547 buffer does not exist. | |
5548 @end table | |
5549 | |
5550 @need 1500 | |
5551 @node Buffer Exercises, , Buffer Related Review, Buffer Walk Through | |
5552 @section Exercises | |
5553 | |
5554 @itemize @bullet | |
5555 @item | |
5556 Write your own @code{simplified-end-of-buffer} function definition; | |
5557 then test it to see whether it works. | |
5558 | |
5559 @item | |
5560 Use @code{if} and @code{get-buffer} to write a function that prints a | |
5561 message telling you whether a buffer exists. | |
5562 | |
5563 @item | |
5564 Using @code{find-tag}, find the source for the @code{copy-to-buffer} | |
5565 function. | |
5566 @end itemize | |
5567 | |
5568 @node More Complex, Narrowing & Widening, Buffer Walk Through, Top | |
5569 @comment node-name, next, previous, up | |
5570 @chapter A Few More Complex Functions | |
5571 | |
5572 In this chapter, we build on what we have learned in previous chapters | |
5573 by looking at more complex functions. The @code{copy-to-buffer} | |
5574 function illustrates use of two @code{save-excursion} expressions in | |
5575 one definition, while the @code{insert-buffer} function illustrates | |
5576 use of an asterisk in an @code{interactive} expression, use of | |
5577 @code{or}, and the important distinction between a name and the object | |
5578 to which the name refers. | |
5579 | |
5580 @menu | |
5581 * copy-to-buffer:: With @code{set-buffer}, @code{get-buffer-create}. | |
5582 * insert-buffer:: Read-only, and with @code{or}. | |
5583 * beginning-of-buffer:: Shows @code{goto-char}, | |
5584 @code{point-min}, and @code{push-mark}. | |
5585 * Second Buffer Related Review:: | |
5586 * optional Exercise:: | |
5587 @end menu | |
5588 | |
5589 @node copy-to-buffer, insert-buffer, More Complex, More Complex | |
5590 @comment node-name, next, previous, up | |
5591 @section The Definition of @code{copy-to-buffer} | |
5592 @findex copy-to-buffer | |
5593 | |
5594 After understanding how @code{append-to-buffer} works, it is easy to | |
5595 understand @code{copy-to-buffer}. This function copies text into a | |
5596 buffer, but instead of adding to the second buffer, it replaces all the | |
5597 previous text in the second buffer. | |
5598 | |
5599 @need 800 | |
5600 The body of @code{copy-to-buffer} looks like this, | |
5601 | |
5602 @smallexample | |
5603 @group | |
5604 @dots{} | |
5605 (interactive "BCopy to buffer: \nr") | |
5606 (let ((oldbuf (current-buffer))) | |
5607 (with-current-buffer (get-buffer-create buffer) | |
5608 (barf-if-buffer-read-only) | |
5609 (erase-buffer) | |
5610 (save-excursion | |
5611 (insert-buffer-substring oldbuf start end))))) | |
5612 @end group | |
5613 @end smallexample | |
5614 | |
5615 The @code{copy-to-buffer} function has a simpler @code{interactive} | |
5616 expression than @code{append-to-buffer}. | |
5617 | |
5618 @need 800 | |
5619 The definition then says | |
5620 | |
5621 @smallexample | |
5622 (with-current-buffer (get-buffer-create buffer) @dots{} | |
5623 @end smallexample | |
5624 | |
5625 First, look at the earliest inner expression; that is evaluated first. | |
5626 That expression starts with @code{get-buffer-create buffer}. The | |
5627 function tells the computer to use the buffer with the name specified | |
5628 as the one to which you are copying, or if such a buffer does not | |
5629 exist, to create it. Then, the @code{with-current-buffer} function | |
5630 evaluates its body with that buffer temporarily current. | |
5631 | |
5632 (This demonstrates another way to shift the computer's attention but | |
5633 not the user's. The @code{append-to-buffer} function showed how to do | |
5634 the same with @code{save-excursion} and @code{set-buffer}. | |
5635 @code{with-current-buffer} is a newer, and arguably easier, | |
5636 mechanism.) | |
5637 | |
5638 The @code{barf-if-buffer-read-only} function sends you an error | |
5639 message saying the buffer is read-only if you cannot modify it. | |
5640 | |
5641 The next line has the @code{erase-buffer} function as its sole | |
5642 contents. That function erases the buffer. | |
5643 | |
5644 Finally, the last two lines contain the @code{save-excursion} | |
5645 expression with @code{insert-buffer-substring} as its body. | |
5646 The @code{insert-buffer-substring} expression copies the text from | |
5647 the buffer you are in (and you have not seen the computer shift its | |
5648 attention, so you don't know that that buffer is now called | |
5649 @code{oldbuf}). | |
5650 | |
5651 Incidentally, this is what is meant by `replacement'. To replace text, | |
5652 Emacs erases the previous text and then inserts new text. | |
5653 | |
5654 @need 1250 | |
5655 In outline, the body of @code{copy-to-buffer} looks like this: | |
5656 | |
5657 @smallexample | |
5658 @group | |
5659 (let (@var{bind-}@code{oldbuf}@var{-to-value-of-}@code{current-buffer}) | |
5660 (@var{with-the-buffer-you-are-copying-to} | |
5661 (@var{but-do-not-erase-or-copy-to-a-read-only-buffer}) | |
5662 (erase-buffer) | |
5663 (save-excursion | |
5664 @var{insert-substring-from-}@code{oldbuf}@var{-into-buffer}))) | |
5665 @end group | |
5666 @end smallexample | |
5667 | |
5668 @node insert-buffer, beginning-of-buffer, copy-to-buffer, More Complex | |
5669 @comment node-name, next, previous, up | |
5670 @section The Definition of @code{insert-buffer} | |
5671 @findex insert-buffer | |
5672 | |
5673 @code{insert-buffer} is yet another buffer-related function. This | |
5674 command copies another buffer @emph{into} the current buffer. It is the | |
5675 reverse of @code{append-to-buffer} or @code{copy-to-buffer}, since they | |
5676 copy a region of text @emph{from} the current buffer to another buffer. | |
5677 | |
5678 Here is a discussion based on the original code. The code was | |
5679 simplified in 2003 and is harder to understand. | |
5680 | |
5681 (@xref{New insert-buffer, , New Body for @code{insert-buffer}}, to see | |
5682 a discussion of the new body.) | |
5683 | |
5684 In addition, this code illustrates the use of @code{interactive} with a | |
5685 buffer that might be @dfn{read-only} and the important distinction | |
5686 between the name of an object and the object actually referred to. | |
5687 | |
5688 @menu | |
5689 * insert-buffer code:: | |
5690 * insert-buffer interactive:: When you can read, but not write. | |
5691 * insert-buffer body:: The body has an @code{or} and a @code{let}. | |
5692 * if & or:: Using an @code{if} instead of an @code{or}. | |
5693 * Insert or:: How the @code{or} expression works. | |
5694 * Insert let:: Two @code{save-excursion} expressions. | |
5695 * New insert-buffer:: | |
5696 @end menu | |
5697 | |
5698 @node insert-buffer code, insert-buffer interactive, insert-buffer, insert-buffer | |
5699 @ifnottex | |
5700 @unnumberedsubsec The Code for @code{insert-buffer} | |
5701 @end ifnottex | |
5702 | |
5703 @need 800 | |
5704 Here is the earlier code: | |
5705 | |
5706 @smallexample | |
5707 @group | |
5708 (defun insert-buffer (buffer) | |
5709 "Insert after point the contents of BUFFER. | |
5710 Puts mark after the inserted text. | |
5711 BUFFER may be a buffer or a buffer name." | |
5712 (interactive "*bInsert buffer:@: ") | |
5713 @end group | |
5714 @group | |
5715 (or (bufferp buffer) | |
5716 (setq buffer (get-buffer buffer))) | |
5717 (let (start end newmark) | |
5718 (save-excursion | |
5719 (save-excursion | |
5720 (set-buffer buffer) | |
5721 (setq start (point-min) end (point-max))) | |
5722 @end group | |
5723 @group | |
5724 (insert-buffer-substring buffer start end) | |
5725 (setq newmark (point))) | |
5726 (push-mark newmark))) | |
5727 @end group | |
5728 @end smallexample | |
5729 | |
5730 @need 1200 | |
5731 As with other function definitions, you can use a template to see an | |
5732 outline of the function: | |
5733 | |
5734 @smallexample | |
5735 @group | |
5736 (defun insert-buffer (buffer) | |
5737 "@var{documentation}@dots{}" | |
5738 (interactive "*bInsert buffer:@: ") | |
5739 @var{body}@dots{}) | |
5740 @end group | |
5741 @end smallexample | |
5742 | |
5743 @node insert-buffer interactive, insert-buffer body, insert-buffer code, insert-buffer | |
5744 @comment node-name, next, previous, up | |
5745 @subsection The Interactive Expression in @code{insert-buffer} | |
5746 @findex interactive, @r{example use of} | |
5747 | |
5748 In @code{insert-buffer}, the argument to the @code{interactive} | |
5749 declaration has two parts, an asterisk, @samp{*}, and @samp{bInsert | |
5750 buffer:@: }. | |
5751 | |
5752 @menu | |
5753 * Read-only buffer:: When a buffer cannot be modified. | |
5754 * b for interactive:: An existing buffer or else its name. | |
5755 @end menu | |
5756 | |
5757 @node Read-only buffer, b for interactive, insert-buffer interactive, insert-buffer interactive | |
5758 @comment node-name, next, previous, up | |
5759 @unnumberedsubsubsec A Read-only Buffer | |
5760 @cindex Read-only buffer | |
5761 @cindex Asterisk for read-only buffer | |
5762 @findex * @r{for read-only buffer} | |
5763 | |
5764 The asterisk is for the situation when the current buffer is a | |
5765 read-only buffer---a buffer that cannot be modified. If | |
5766 @code{insert-buffer} is called when the current buffer is read-only, a | |
5767 message to this effect is printed in the echo area and the terminal | |
5768 may beep or blink at you; you will not be permitted to insert anything | |
5769 into current buffer. The asterisk does not need to be followed by a | |
5770 newline to separate it from the next argument. | |
5771 | |
5772 @node b for interactive, , Read-only buffer, insert-buffer interactive | |
5773 @comment node-name, next, previous, up | |
5774 @unnumberedsubsubsec @samp{b} in an Interactive Expression | |
5775 | |
5776 The next argument in the interactive expression starts with a lower | |
5777 case @samp{b}. (This is different from the code for | |
5778 @code{append-to-buffer}, which uses an upper-case @samp{B}. | |
5779 @xref{append-to-buffer, , The Definition of @code{append-to-buffer}}.) | |
5780 The lower-case @samp{b} tells the Lisp interpreter that the argument | |
5781 for @code{insert-buffer} should be an existing buffer or else its | |
5782 name. (The upper-case @samp{B} option provides for the possibility | |
5783 that the buffer does not exist.) Emacs will prompt you for the name | |
5784 of the buffer, offering you a default buffer, with name completion | |
5785 enabled. If the buffer does not exist, you receive a message that | |
5786 says ``No match''; your terminal may beep at you as well. | |
5787 | |
5788 The new and simplified code generates a list for @code{interactive}. | |
5789 It uses the @code{barf-if-buffer-read-only} and @code{read-buffer} | |
5790 functions with which we are already familiar and the @code{progn} | |
5791 special form with which we are not. (It will be described later.) | |
5792 | |
5793 @node insert-buffer body, if & or, insert-buffer interactive, insert-buffer | |
5794 @comment node-name, next, previous, up | |
5795 @subsection The Body of the @code{insert-buffer} Function | |
5796 | |
5797 The body of the @code{insert-buffer} function has two major parts: an | |
5798 @code{or} expression and a @code{let} expression. The purpose of the | |
5799 @code{or} expression is to ensure that the argument @code{buffer} is | |
5800 bound to a buffer and not just the name of a buffer. The body of the | |
5801 @code{let} expression contains the code which copies the other buffer | |
5802 into the current buffer. | |
5803 | |
5804 @need 1250 | |
5805 In outline, the two expressions fit into the @code{insert-buffer} | |
5806 function like this: | |
5807 | |
5808 @smallexample | |
5809 @group | |
5810 (defun insert-buffer (buffer) | |
5811 "@var{documentation}@dots{}" | |
5812 (interactive "*bInsert buffer:@: ") | |
5813 (or @dots{} | |
5814 @dots{} | |
5815 @end group | |
5816 @group | |
5817 (let (@var{varlist}) | |
5818 @var{body-of-}@code{let}@dots{} ) | |
5819 @end group | |
5820 @end smallexample | |
5821 | |
5822 To understand how the @code{or} expression ensures that the argument | |
5823 @code{buffer} is bound to a buffer and not to the name of a buffer, it | |
5824 is first necessary to understand the @code{or} function. | |
5825 | |
5826 Before doing this, let me rewrite this part of the function using | |
5827 @code{if} so that you can see what is done in a manner that will be familiar. | |
5828 | |
5829 @node if & or, Insert or, insert-buffer body, insert-buffer | |
5830 @comment node-name, next, previous, up | |
5831 @subsection @code{insert-buffer} With an @code{if} Instead of an @code{or} | |
5832 | |
5833 The job to be done is to make sure the value of @code{buffer} is a | |
5834 buffer itself and not the name of a buffer. If the value is the name, | |
5835 then the buffer itself must be got. | |
5836 | |
5837 You can imagine yourself at a conference where an usher is wandering | |
5838 around holding a list with your name on it and looking for you: the | |
5839 usher is ``bound'' to your name, not to you; but when the usher finds | |
5840 you and takes your arm, the usher becomes ``bound'' to you. | |
5841 | |
5842 @need 800 | |
5843 In Lisp, you might describe this situation like this: | |
5844 | |
5845 @smallexample | |
5846 @group | |
5847 (if (not (holding-on-to-guest)) | |
5848 (find-and-take-arm-of-guest)) | |
5849 @end group | |
5850 @end smallexample | |
5851 | |
5852 We want to do the same thing with a buffer---if we do not have the | |
5853 buffer itself, we want to get it. | |
5854 | |
5855 @need 1200 | |
5856 Using a predicate called @code{bufferp} that tells us whether we have a | |
5857 buffer (rather than its name), we can write the code like this: | |
5858 | |
5859 @smallexample | |
5860 @group | |
5861 (if (not (bufferp buffer)) ; @r{if-part} | |
5862 (setq buffer (get-buffer buffer))) ; @r{then-part} | |
5863 @end group | |
5864 @end smallexample | |
5865 | |
5866 @noindent | |
5867 Here, the true-or-false-test of the @code{if} expression is | |
5868 @w{@code{(not (bufferp buffer))}}; and the then-part is the expression | |
5869 @w{@code{(setq buffer (get-buffer buffer))}}. | |
5870 | |
5871 In the test, the function @code{bufferp} returns true if its argument is | |
5872 a buffer---but false if its argument is the name of the buffer. (The | |
5873 last character of the function name @code{bufferp} is the character | |
5874 @samp{p}; as we saw earlier, such use of @samp{p} is a convention that | |
5875 indicates that the function is a predicate, which is a term that means | |
5876 that the function will determine whether some property is true or false. | |
5877 @xref{Wrong Type of Argument, , Using the Wrong Type Object as an | |
5878 Argument}.) | |
5879 | |
5880 @need 1200 | |
5881 The function @code{not} precedes the expression @code{(bufferp buffer)}, | |
5882 so the true-or-false-test looks like this: | |
5883 | |
5884 @smallexample | |
5885 (not (bufferp buffer)) | |
5886 @end smallexample | |
5887 | |
5888 @noindent | |
5889 @code{not} is a function that returns true if its argument is false | |
5890 and false if its argument is true. So if @code{(bufferp buffer)} | |
5891 returns true, the @code{not} expression returns false and vice-verse: | |
5892 what is ``not true'' is false and what is ``not false'' is true. | |
5893 | |
5894 Using this test, the @code{if} expression works as follows: when the | |
5895 value of the variable @code{buffer} is actually a buffer rather than | |
5896 its name, the true-or-false-test returns false and the @code{if} | |
5897 expression does not evaluate the then-part. This is fine, since we do | |
5898 not need to do anything to the variable @code{buffer} if it really is | |
5899 a buffer. | |
5900 | |
5901 On the other hand, when the value of @code{buffer} is not a buffer | |
5902 itself, but the name of a buffer, the true-or-false-test returns true | |
5903 and the then-part of the expression is evaluated. In this case, the | |
5904 then-part is @code{(setq buffer (get-buffer buffer))}. This | |
5905 expression uses the @code{get-buffer} function to return an actual | |
5906 buffer itself, given its name. The @code{setq} then sets the variable | |
5907 @code{buffer} to the value of the buffer itself, replacing its previous | |
5908 value (which was the name of the buffer). | |
5909 | |
5910 @node Insert or, Insert let, if & or, insert-buffer | |
5911 @comment node-name, next, previous, up | |
5912 @subsection The @code{or} in the Body | |
5913 | |
5914 The purpose of the @code{or} expression in the @code{insert-buffer} | |
5915 function is to ensure that the argument @code{buffer} is bound to a | |
5916 buffer and not just to the name of a buffer. The previous section shows | |
5917 how the job could have been done using an @code{if} expression. | |
5918 However, the @code{insert-buffer} function actually uses @code{or}. | |
5919 To understand this, it is necessary to understand how @code{or} works. | |
5920 | |
5921 @findex or | |
5922 An @code{or} function can have any number of arguments. It evaluates | |
5923 each argument in turn and returns the value of the first of its | |
5924 arguments that is not @code{nil}. Also, and this is a crucial feature | |
5925 of @code{or}, it does not evaluate any subsequent arguments after | |
5926 returning the first non-@code{nil} value. | |
5927 | |
5928 @need 800 | |
5929 The @code{or} expression looks like this: | |
5930 | |
5931 @smallexample | |
5932 @group | |
5933 (or (bufferp buffer) | |
5934 (setq buffer (get-buffer buffer))) | |
5935 @end group | |
5936 @end smallexample | |
5937 | |
5938 @noindent | |
5939 The first argument to @code{or} is the expression @code{(bufferp buffer)}. | |
5940 This expression returns true (a non-@code{nil} value) if the buffer is | |
5941 actually a buffer, and not just the name of a buffer. In the @code{or} | |
5942 expression, if this is the case, the @code{or} expression returns this | |
5943 true value and does not evaluate the next expression---and this is fine | |
5944 with us, since we do not want to do anything to the value of | |
5945 @code{buffer} if it really is a buffer. | |
5946 | |
5947 On the other hand, if the value of @code{(bufferp buffer)} is @code{nil}, | |
5948 which it will be if the value of @code{buffer} is the name of a buffer, | |
5949 the Lisp interpreter evaluates the next element of the @code{or} | |
5950 expression. This is the expression @code{(setq buffer (get-buffer | |
5951 buffer))}. This expression returns a non-@code{nil} value, which | |
5952 is the value to which it sets the variable @code{buffer}---and this | |
5953 value is a buffer itself, not the name of a buffer. | |
5954 | |
5955 The result of all this is that the symbol @code{buffer} is always | |
5956 bound to a buffer itself rather than to the name of a buffer. All | |
5957 this is necessary because the @code{set-buffer} function in a | |
5958 following line only works with a buffer itself, not with the name to a | |
5959 buffer. | |
5960 | |
5961 @need 1250 | |
5962 Incidentally, using @code{or}, the situation with the usher would be | |
5963 written like this: | |
5964 | |
5965 @smallexample | |
5966 (or (holding-on-to-guest) (find-and-take-arm-of-guest)) | |
5967 @end smallexample | |
5968 | |
5969 @node Insert let, New insert-buffer, Insert or, insert-buffer | |
5970 @comment node-name, next, previous, up | |
5971 @subsection The @code{let} Expression in @code{insert-buffer} | |
5972 | |
5973 After ensuring that the variable @code{buffer} refers to a buffer itself | |
5974 and not just to the name of a buffer, the @code{insert-buffer function} | |
5975 continues with a @code{let} expression. This specifies three local | |
5976 variables, @code{start}, @code{end}, and @code{newmark} and binds them | |
5977 to the initial value @code{nil}. These variables are used inside the | |
5978 remainder of the @code{let} and temporarily hide any other occurrence of | |
5979 variables of the same name in Emacs until the end of the @code{let}. | |
5980 | |
5981 @need 1200 | |
5982 The body of the @code{let} contains two @code{save-excursion} | |
5983 expressions. First, we will look at the inner @code{save-excursion} | |
5984 expression in detail. The expression looks like this: | |
5985 | |
5986 @smallexample | |
5987 @group | |
5988 (save-excursion | |
5989 (set-buffer buffer) | |
5990 (setq start (point-min) end (point-max))) | |
5991 @end group | |
5992 @end smallexample | |
5993 | |
5994 @noindent | |
5995 The expression @code{(set-buffer buffer)} changes Emacs' attention | |
5996 from the current buffer to the one from which the text will copied. | |
5997 In that buffer, the variables @code{start} and @code{end} are set to | |
5998 the beginning and end of the buffer, using the commands | |
5999 @code{point-min} and @code{point-max}. Note that we have here an | |
6000 illustration of how @code{setq} is able to set two variables in the | |
6001 same expression. The first argument of @code{setq} is set to the | |
6002 value of its second, and its third argument is set to the value of its | |
6003 fourth. | |
6004 | |
6005 After the body of the inner @code{save-excursion} is evaluated, the | |
6006 @code{save-excursion} restores the original buffer, but @code{start} and | |
6007 @code{end} remain set to the values of the beginning and end of the | |
6008 buffer from which the text will be copied. | |
6009 | |
6010 @need 1250 | |
6011 The outer @code{save-excursion} expression looks like this: | |
6012 | |
6013 @smallexample | |
6014 @group | |
6015 (save-excursion | |
6016 (@var{inner-}@code{save-excursion}@var{-expression} | |
6017 (@var{go-to-new-buffer-and-set-}@code{start}@var{-and-}@code{end}) | |
6018 (insert-buffer-substring buffer start end) | |
6019 (setq newmark (point))) | |
6020 @end group | |
6021 @end smallexample | |
6022 | |
6023 @noindent | |
6024 The @code{insert-buffer-substring} function copies the text | |
6025 @emph{into} the current buffer @emph{from} the region indicated by | |
6026 @code{start} and @code{end} in @code{buffer}. Since the whole of the | |
6027 second buffer lies between @code{start} and @code{end}, the whole of | |
6028 the second buffer is copied into the buffer you are editing. Next, | |
6029 the value of point, which will be at the end of the inserted text, is | |
6030 recorded in the variable @code{newmark}. | |
6031 | |
6032 After the body of the outer @code{save-excursion} is evaluated, point | |
6033 and mark are relocated to their original places. | |
6034 | |
6035 However, it is convenient to locate a mark at the end of the newly | |
6036 inserted text and locate point at its beginning. The @code{newmark} | |
6037 variable records the end of the inserted text. In the last line of | |
6038 the @code{let} expression, the @code{(push-mark newmark)} expression | |
6039 function sets a mark to this location. (The previous location of the | |
6040 mark is still accessible; it is recorded on the mark ring and you can | |
6041 go back to it with @kbd{C-u C-@key{SPC}}.) Meanwhile, point is | |
6042 located at the beginning of the inserted text, which is where it was | |
6043 before you called the insert function, the position of which was saved | |
6044 by the first @code{save-excursion}. | |
6045 | |
6046 @need 1250 | |
6047 The whole @code{let} expression looks like this: | |
6048 | |
6049 @smallexample | |
6050 @group | |
6051 (let (start end newmark) | |
6052 (save-excursion | |
6053 (save-excursion | |
6054 (set-buffer buffer) | |
6055 (setq start (point-min) end (point-max))) | |
6056 (insert-buffer-substring buffer start end) | |
6057 (setq newmark (point))) | |
6058 (push-mark newmark)) | |
6059 @end group | |
6060 @end smallexample | |
6061 | |
6062 Like the @code{append-to-buffer} function, the @code{insert-buffer} | |
6063 function uses @code{let}, @code{save-excursion}, and | |
6064 @code{set-buffer}. In addition, the function illustrates one way to | |
6065 use @code{or}. All these functions are building blocks that we will | |
6066 find and use again and again. | |
6067 | |
6068 @node New insert-buffer, , Insert let, insert-buffer | |
6069 @comment node-name, next, previous, up | |
6070 @subsection New Body for @code{insert-buffer} | |
6071 @findex insert-buffer, new version body | |
6072 @findex new version body for insert-buffer | |
6073 | |
6074 The body in the GNU Emacs 22 version is more confusing than the original. | |
6075 | |
6076 @need 1250 | |
6077 It consists of two expressions, | |
6078 | |
6079 @smallexample | |
6080 @group | |
6081 (push-mark | |
6082 (save-excursion | |
6083 (insert-buffer-substring (get-buffer buffer)) | |
6084 (point))) | |
6085 | |
6086 nil | |
6087 @end group | |
6088 @end smallexample | |
6089 | |
6090 @noindent | |
6091 except, and this is what confuses novices, very important work is done | |
6092 inside the @code{push-mark} expression. | |
6093 | |
6094 The @code{get-buffer} function returns a buffer with the name | |
6095 provided. You will note that the function is @emph{not} called | |
6096 @code{get-buffer-create}; it does not create a buffer if one does not | |
6097 already exist. The buffer returned by @code{get-buffer}, an existing | |
6098 buffer, is passed to @code{insert-buffer-substring}, which inserts the | |
6099 whole of the buffer (since you did not specify anything else). | |
6100 | |
6101 The location into which the buffer is inserted is recorded by | |
6102 @code{push-mark}. Then the function returns @code{nil}, the value of | |
6103 its last command. Put another way, the @code{insert-buffer} function | |
6104 exists only to produce a side effect, inserting another buffer, not to | |
6105 return any value. | |
6106 | |
6107 @node beginning-of-buffer, Second Buffer Related Review, insert-buffer, More Complex | |
6108 @comment node-name, next, previous, up | |
6109 @section Complete Definition of @code{beginning-of-buffer} | |
6110 @findex beginning-of-buffer | |
6111 | |
6112 The basic structure of the @code{beginning-of-buffer} function has | |
6113 already been discussed. (@xref{simplified-beginning-of-buffer, , A | |
6114 Simplified @code{beginning-of-buffer} Definition}.) | |
6115 This section describes the complex part of the definition. | |
6116 | |
6117 As previously described, when invoked without an argument, | |
6118 @code{beginning-of-buffer} moves the cursor to the beginning of the | |
6119 buffer (in truth, the beginning of the accessible portion of the | |
6120 buffer), leaving the mark at the previous position. However, when the | |
6121 command is invoked with a number between one and ten, the function | |
6122 considers that number to be a fraction of the length of the buffer, | |
6123 measured in tenths, and Emacs moves the cursor that fraction of the | |
6124 way from the beginning of the buffer. Thus, you can either call this | |
6125 function with the key command @kbd{M-<}, which will move the cursor to | |
6126 the beginning of the buffer, or with a key command such as @kbd{C-u 7 | |
6127 M-<} which will move the cursor to a point 70% of the way through the | |
6128 buffer. If a number bigger than ten is used for the argument, it | |
6129 moves to the end of the buffer. | |
6130 | |
6131 The @code{beginning-of-buffer} function can be called with or without an | |
6132 argument. The use of the argument is optional. | |
6133 | |
6134 @menu | |
6135 * Optional Arguments:: | |
6136 * beginning-of-buffer opt arg:: Example with optional argument. | |
6137 * beginning-of-buffer complete:: | |
6138 @end menu | |
6139 | |
6140 @node Optional Arguments, beginning-of-buffer opt arg, beginning-of-buffer, beginning-of-buffer | |
6141 @subsection Optional Arguments | |
6142 | |
6143 Unless told otherwise, Lisp expects that a function with an argument in | |
6144 its function definition will be called with a value for that argument. | |
6145 If that does not happen, you get an error and a message that says | |
6146 @samp{Wrong number of arguments}. | |
6147 | |
6148 @cindex Optional arguments | |
6149 @cindex Keyword | |
6150 @findex optional | |
6151 However, optional arguments are a feature of Lisp: a particular | |
6152 @dfn{keyword} is used to tell the Lisp interpreter that an argument is | |
6153 optional. The keyword is @code{&optional}. (The @samp{&} in front of | |
6154 @samp{optional} is part of the keyword.) In a function definition, if | |
6155 an argument follows the keyword @code{&optional}, no value need be | |
6156 passed to that argument when the function is called. | |
6157 | |
6158 @need 1200 | |
6159 The first line of the function definition of @code{beginning-of-buffer} | |
6160 therefore looks like this: | |
6161 | |
6162 @smallexample | |
6163 (defun beginning-of-buffer (&optional arg) | |
6164 @end smallexample | |
6165 | |
6166 @need 1250 | |
6167 In outline, the whole function looks like this: | |
6168 | |
6169 @smallexample | |
6170 @group | |
6171 (defun beginning-of-buffer (&optional arg) | |
6172 "@var{documentation}@dots{}" | |
6173 (interactive "P") | |
6174 (or (@var{is-the-argument-a-cons-cell} arg) | |
6175 (and @var{are-both-transient-mark-mode-and-mark-active-true}) | |
6176 (push-mark)) | |
6177 (let (@var{determine-size-and-set-it}) | |
6178 (goto-char | |
6179 (@var{if-there-is-an-argument} | |
6180 @var{figure-out-where-to-go} | |
6181 @var{else-go-to} | |
6182 (point-min)))) | |
6183 @var{do-nicety} | |
6184 @end group | |
6185 @end smallexample | |
6186 | |
6187 The function is similar to the @code{simplified-beginning-of-buffer} | |
6188 function except that the @code{interactive} expression has @code{"P"} | |
6189 as an argument and the @code{goto-char} function is followed by an | |
6190 if-then-else expression that figures out where to put the cursor if | |
6191 there is an argument that is not a cons cell. | |
6192 | |
6193 (Since I do not explain a cons cell for many more chapters, please | |
6194 consider ignoring the function @code{consp}. @xref{List | |
6195 Implementation, , How Lists are Implemented}, and @ref{Cons Cell Type, | |
6196 , Cons Cell and List Types, elisp, The GNU Emacs Lisp Reference | |
6197 Manual}.) | |
6198 | |
6199 The @code{"P"} in the @code{interactive} expression tells Emacs to | |
6200 pass a prefix argument, if there is one, to the function in raw form. | |
6201 A prefix argument is made by typing the @key{META} key followed by a | |
6202 number, or by typing @kbd{C-u} and then a number. (If you don't type | |
6203 a number, @kbd{C-u} defaults to a cons cell with a 4. A lowercase | |
6204 @code{"p"} in the @code{interactive} expression causes the function to | |
6205 convert a prefix arg to a number.) | |
6206 | |
6207 The true-or-false-test of the @code{if} expression looks complex, but | |
6208 it is not: it checks whether @code{arg} has a value that is not | |
6209 @code{nil} and whether it is a cons cell. (That is what @code{consp} | |
6210 does; it checks whether its argument is a cons cell.) If @code{arg} | |
6211 has a value that is not @code{nil} (and is not a cons cell), which | |
6212 will be the case if @code{beginning-of-buffer} is called with a | |
6213 numeric argument, then this true-or-false-test will return true and | |
6214 the then-part of the @code{if} expression will be evaluated. On the | |
6215 other hand, if @code{beginning-of-buffer} is not called with an | |
6216 argument, the value of @code{arg} will be @code{nil} and the else-part | |
6217 of the @code{if} expression will be evaluated. The else-part is | |
6218 simply @code{point-min}, and when this is the outcome, the whole | |
6219 @code{goto-char} expression is @code{(goto-char (point-min))}, which | |
6220 is how we saw the @code{beginning-of-buffer} function in its | |
6221 simplified form. | |
6222 | |
6223 @node beginning-of-buffer opt arg, beginning-of-buffer complete, Optional Arguments, beginning-of-buffer | |
6224 @subsection @code{beginning-of-buffer} with an Argument | |
6225 | |
6226 When @code{beginning-of-buffer} is called with an argument, an | |
6227 expression is evaluated which calculates what value to pass to | |
6228 @code{goto-char}. This expression is rather complicated at first sight. | |
6229 It includes an inner @code{if} expression and much arithmetic. It looks | |
6230 like this: | |
6231 | |
6232 @smallexample | |
6233 @group | |
6234 (if (> (buffer-size) 10000) | |
6235 ;; @r{Avoid overflow for large buffer sizes!} | |
6236 (* (prefix-numeric-value arg) | |
6237 (/ size 10)) | |
6238 (/ | |
6239 (+ 10 | |
6240 (* | |
6241 size (prefix-numeric-value arg))) 10))) | |
6242 @end group | |
6243 @end smallexample | |
6244 | |
6245 @menu | |
6246 * Disentangle beginning-of-buffer:: | |
6247 * Large buffer case:: | |
6248 * Small buffer case:: | |
6249 @end menu | |
6250 | |
6251 @node Disentangle beginning-of-buffer, Large buffer case, beginning-of-buffer opt arg, beginning-of-buffer opt arg | |
6252 @ifnottex | |
6253 @unnumberedsubsubsec Disentangle @code{beginning-of-buffer} | |
6254 @end ifnottex | |
6255 | |
6256 Like other complex-looking expressions, the conditional expression | |
6257 within @code{beginning-of-buffer} can be disentangled by looking at it | |
6258 as parts of a template, in this case, the template for an if-then-else | |
6259 expression. In skeletal form, the expression looks like this: | |
6260 | |
6261 @smallexample | |
6262 @group | |
6263 (if (@var{buffer-is-large} | |
6264 @var{divide-buffer-size-by-10-and-multiply-by-arg} | |
6265 @var{else-use-alternate-calculation} | |
6266 @end group | |
6267 @end smallexample | |
6268 | |
6269 The true-or-false-test of this inner @code{if} expression checks the | |
6270 size of the buffer. The reason for this is that the old version 18 | |
6271 Emacs used numbers that are no bigger than eight million or so and in | |
6272 the computation that followed, the programmer feared that Emacs might | |
6273 try to use over-large numbers if the buffer were large. The term | |
6274 `overflow', mentioned in the comment, means numbers that are over | |
6275 large. More recent versions of Emacs use larger numbers, but this | |
6276 code has not been touched, if only because people now look at buffers | |
6277 that are far, far larger than ever before. | |
6278 | |
6279 There are two cases: if the buffer is large and if it is not. | |
6280 | |
6281 @node Large buffer case, Small buffer case, Disentangle beginning-of-buffer, beginning-of-buffer opt arg | |
6282 @comment node-name, next, previous, up | |
6283 @unnumberedsubsubsec What happens in a large buffer | |
6284 | |
6285 In @code{beginning-of-buffer}, the inner @code{if} expression tests | |
6286 whether the size of the buffer is greater than 10,000 characters. To do | |
6287 this, it uses the @code{>} function and the computation of @code{size} | |
6288 that comes from the let expression. | |
6289 | |
6290 In the old days, the function @code{buffer-size} was used. Not only | |
6291 was that function called several times, it gave the size of the whole | |
6292 buffer, not the accessible part. The computation makes much more | |
6293 sense when it handles just the accessible part. (@xref{Narrowing & | |
6294 Widening, , Narrowing and Widening}, for more information on focusing | |
6295 attention to an `accessible' part.) | |
6296 | |
6297 @need 800 | |
6298 The line looks like this: | |
6299 | |
6300 @smallexample | |
6301 (if (> size 10000) | |
6302 @end smallexample | |
6303 | |
6304 @need 1200 | |
6305 @noindent | |
6306 When the buffer is large, the then-part of the @code{if} expression is | |
6307 evaluated. It reads like this (after formatting for easy reading): | |
6308 | |
6309 @smallexample | |
6310 @group | |
6311 (* | |
6312 (prefix-numeric-value arg) | |
6313 (/ size 10)) | |
6314 @end group | |
6315 @end smallexample | |
6316 | |
6317 @noindent | |
6318 This expression is a multiplication, with two arguments to the function | |
6319 @code{*}. | |
6320 | |
6321 The first argument is @code{(prefix-numeric-value arg)}. When | |
6322 @code{"P"} is used as the argument for @code{interactive}, the value | |
6323 passed to the function as its argument is passed a ``raw prefix | |
6324 argument'', and not a number. (It is a number in a list.) To perform | |
6325 the arithmetic, a conversion is necessary, and | |
6326 @code{prefix-numeric-value} does the job. | |
6327 | |
6328 @findex / @r{(division)} | |
6329 @cindex Division | |
6330 The second argument is @code{(/ size 10)}. This expression divides | |
6331 the numeric value by ten --- the numeric value of the size of the | |
6332 accessible portion of the buffer. This produces a number that tells | |
6333 how many characters make up one tenth of the buffer size. (In Lisp, | |
6334 @code{/} is used for division, just as @code{*} is used for | |
6335 multiplication.) | |
6336 | |
6337 @need 1200 | |
6338 In the multiplication expression as a whole, this amount is multiplied | |
6339 by the value of the prefix argument---the multiplication looks like this: | |
6340 | |
6341 @smallexample | |
6342 @group | |
6343 (* @var{numeric-value-of-prefix-arg} | |
6344 @var{number-of-characters-in-one-tenth-of-the-accessible-buffer}) | |
6345 @end group | |
6346 @end smallexample | |
6347 | |
6348 @noindent | |
6349 If, for example, the prefix argument is @samp{7}, the one-tenth value | |
6350 will be multiplied by 7 to give a position 70% of the way through. | |
6351 | |
6352 @need 1200 | |
6353 The result of all this is that if the accessible portion of the buffer | |
6354 is large, the @code{goto-char} expression reads like this: | |
6355 | |
6356 @smallexample | |
6357 @group | |
6358 (goto-char (* (prefix-numeric-value arg) | |
6359 (/ size 10))) | |
6360 @end group | |
6361 @end smallexample | |
6362 | |
6363 This puts the cursor where we want it. | |
6364 | |
6365 @node Small buffer case, , Large buffer case, beginning-of-buffer opt arg | |
6366 @comment node-name, next, previous, up | |
6367 @unnumberedsubsubsec What happens in a small buffer | |
6368 | |
6369 If the buffer contains fewer than 10,000 characters, a slightly | |
6370 different computation is performed. You might think this is not | |
6371 necessary, since the first computation could do the job. However, in | |
6372 a small buffer, the first method may not put the cursor on exactly the | |
6373 desired line; the second method does a better job. | |
6374 | |
6375 @need 800 | |
6376 The code looks like this: | |
6377 | |
6378 @c Keep this on one line. | |
6379 @smallexample | |
6380 (/ (+ 10 (* size (prefix-numeric-value arg))) 10)) | |
6381 @end smallexample | |
6382 | |
6383 @need 1200 | |
6384 @noindent | |
6385 This is code in which you figure out what happens by discovering how the | |
6386 functions are embedded in parentheses. It is easier to read if you | |
6387 reformat it with each expression indented more deeply than its | |
6388 enclosing expression: | |
6389 | |
6390 @smallexample | |
6391 @group | |
6392 (/ | |
6393 (+ 10 | |
6394 (* | |
6395 size | |
6396 (prefix-numeric-value arg))) | |
6397 10)) | |
6398 @end group | |
6399 @end smallexample | |
6400 | |
6401 @need 1200 | |
6402 @noindent | |
6403 Looking at parentheses, we see that the innermost operation is | |
6404 @code{(prefix-numeric-value arg)}, which converts the raw argument to | |
6405 a number. In the following expression, this number is multiplied by | |
6406 the size of the accessible portion of the buffer: | |
6407 | |
6408 @smallexample | |
6409 (* size (prefix-numeric-value arg)) | |
6410 @end smallexample | |
6411 | |
6412 @noindent | |
6413 This multiplication creates a number that may be larger than the size of | |
6414 the buffer---seven times larger if the argument is 7, for example. Ten | |
6415 is then added to this number and finally the large number is divided by | |
6416 ten to provide a value that is one character larger than the percentage | |
6417 position in the buffer. | |
6418 | |
6419 The number that results from all this is passed to @code{goto-char} and | |
6420 the cursor is moved to that point. | |
6421 | |
6422 @need 1500 | |
6423 @node beginning-of-buffer complete, , beginning-of-buffer opt arg, beginning-of-buffer | |
6424 @comment node-name, next, previous, up | |
6425 @subsection The Complete @code{beginning-of-buffer} | |
6426 | |
6427 @need 1000 | |
6428 Here is the complete text of the @code{beginning-of-buffer} function: | |
6429 @sp 1 | |
6430 | |
6431 @c In GNU Emacs 22 | |
6432 @smallexample | |
6433 @group | |
6434 (defun beginning-of-buffer (&optional arg) | |
6435 "Move point to the beginning of the buffer; | |
6436 leave mark at previous position. | |
6437 With \\[universal-argument] prefix, | |
6438 do not set mark at previous position. | |
6439 With numeric arg N, | |
6440 put point N/10 of the way from the beginning. | |
6441 | |
6442 If the buffer is narrowed, | |
6443 this command uses the beginning and size | |
6444 of the accessible part of the buffer. | |
6445 @end group | |
6446 | |
6447 @group | |
6448 Don't use this command in Lisp programs! | |
6449 \(goto-char (point-min)) is faster | |
6450 and avoids clobbering the mark." | |
6451 (interactive "P") | |
6452 (or (consp arg) | |
6453 (and transient-mark-mode mark-active) | |
6454 (push-mark)) | |
6455 @end group | |
6456 @group | |
6457 (let ((size (- (point-max) (point-min)))) | |
6458 (goto-char (if (and arg (not (consp arg))) | |
6459 (+ (point-min) | |
6460 (if (> size 10000) | |
6461 ;; Avoid overflow for large buffer sizes! | |
6462 (* (prefix-numeric-value arg) | |
6463 (/ size 10)) | |
6464 (/ (+ 10 (* size (prefix-numeric-value arg))) 10))) | |
6465 (point-min)))) | |
6466 (if arg (forward-line 1))) | |
6467 @end group | |
6468 @end smallexample | |
6469 | |
6470 @ignore | |
6471 From before GNU Emacs 22 | |
6472 @smallexample | |
6473 @group | |
6474 (defun beginning-of-buffer (&optional arg) | |
6475 "Move point to the beginning of the buffer; | |
6476 leave mark at previous position. | |
6477 With arg N, put point N/10 of the way | |
6478 from the true beginning. | |
6479 @end group | |
6480 @group | |
6481 Don't use this in Lisp programs! | |
6482 \(goto-char (point-min)) is faster | |
6483 and does not set the mark." | |
6484 (interactive "P") | |
6485 (push-mark) | |
6486 @end group | |
6487 @group | |
6488 (goto-char | |
6489 (if arg | |
6490 (if (> (buffer-size) 10000) | |
6491 ;; @r{Avoid overflow for large buffer sizes!} | |
6492 (* (prefix-numeric-value arg) | |
6493 (/ (buffer-size) 10)) | |
6494 @end group | |
6495 @group | |
6496 (/ (+ 10 (* (buffer-size) | |
6497 (prefix-numeric-value arg))) | |
6498 10)) | |
6499 (point-min))) | |
6500 (if arg (forward-line 1))) | |
6501 @end group | |
6502 @end smallexample | |
6503 @end ignore | |
6504 | |
6505 @noindent | |
6506 Except for two small points, the previous discussion shows how this | |
6507 function works. The first point deals with a detail in the | |
6508 documentation string, and the second point concerns the last line of | |
6509 the function. | |
6510 | |
6511 @need 800 | |
6512 In the documentation string, there is reference to an expression: | |
6513 | |
6514 @smallexample | |
6515 \\[universal-argument] | |
6516 @end smallexample | |
6517 | |
6518 @noindent | |
6519 A @samp{\\} is used before the first square bracket of this | |
6520 expression. This @samp{\\} tells the Lisp interpreter to substitute | |
6521 whatever key is currently bound to the @samp{[@dots{}]}. In the case | |
6522 of @code{universal-argument}, that is usually @kbd{C-u}, but it might | |
6523 be different. (@xref{Documentation Tips, , Tips for Documentation | |
6524 Strings, elisp, The GNU Emacs Lisp Reference Manual}, for more | |
6525 information.) | |
6526 | |
6527 @need 1200 | |
6528 Finally, the last line of the @code{beginning-of-buffer} command says | |
6529 to move point to the beginning of the next line if the command is | |
6530 invoked with an argument: | |
6531 | |
6532 @smallexample | |
6533 (if arg (forward-line 1))) | |
6534 @end smallexample | |
6535 | |
6536 @noindent | |
6537 This puts the cursor at the beginning of the first line after the | |
6538 appropriate tenths position in the buffer. This is a flourish that | |
6539 means that the cursor is always located @emph{at least} the requested | |
6540 tenths of the way through the buffer, which is a nicety that is, | |
6541 perhaps, not necessary, but which, if it did not occur, would be sure | |
6542 to draw complaints. | |
6543 | |
6544 On the other hand, it also means that if you specify the command with | |
6545 a @kbd{C-u}, but without a number, that is to say, if the `raw prefix | |
6546 argument' is simply a cons cell, then the command puts you at the | |
6547 beginning of the second line @dots{} I don't know whether this is | |
6548 intended or whether no one has dealt with the code to avoid this | |
6549 happening. | |
6550 | |
6551 @node Second Buffer Related Review, optional Exercise, beginning-of-buffer, More Complex | |
6552 @comment node-name, next, previous, up | |
6553 @section Review | |
6554 | |
6555 Here is a brief summary of some of the topics covered in this chapter. | |
6556 | |
6557 @table @code | |
6558 @item or | |
6559 Evaluate each argument in sequence, and return the value of the first | |
6560 argument that is not @code{nil}; if none return a value that is not | |
6561 @code{nil}, return @code{nil}. In brief, return the first true value | |
6562 of the arguments; return a true value if one @emph{or} any of the | |
6563 others are true. | |
6564 | |
6565 @item and | |
6566 Evaluate each argument in sequence, and if any are @code{nil}, return | |
6567 @code{nil}; if none are @code{nil}, return the value of the last | |
6568 argument. In brief, return a true value only if all the arguments are | |
6569 true; return a true value if one @emph{and} each of the others is | |
6570 true. | |
6571 | |
6572 @item &optional | |
6573 A keyword used to indicate that an argument to a function definition | |
6574 is optional; this means that the function can be evaluated without the | |
6575 argument, if desired. | |
6576 | |
6577 @item prefix-numeric-value | |
6578 Convert the `raw prefix argument' produced by @code{(interactive | |
6579 "P")} to a numeric value. | |
6580 | |
6581 @item forward-line | |
6582 Move point forward to the beginning of the next line, or if the argument | |
6583 is greater than one, forward that many lines. If it can't move as far | |
6584 forward as it is supposed to, @code{forward-line} goes forward as far as | |
6585 it can and then returns a count of the number of additional lines it was | |
6586 supposed to move but couldn't. | |
6587 | |
6588 @item erase-buffer | |
6589 Delete the entire contents of the current buffer. | |
6590 | |
6591 @item bufferp | |
6592 Return @code{t} if its argument is a buffer; otherwise return @code{nil}. | |
6593 @end table | |
6594 | |
6595 @node optional Exercise, , Second Buffer Related Review, More Complex | |
6596 @section @code{optional} Argument Exercise | |
6597 | |
6598 Write an interactive function with an optional argument that tests | |
6599 whether its argument, a number, is greater than or equal to, or else, | |
6600 less than the value of @code{fill-column}, and tells you which, in a | |
6601 message. However, if you do not pass an argument to the function, use | |
6602 56 as a default value. | |
6603 | |
6604 @node Narrowing & Widening, car cdr & cons, More Complex, Top | |
6605 @comment node-name, next, previous, up | |
6606 @chapter Narrowing and Widening | |
6607 @cindex Focusing attention (narrowing) | |
6608 @cindex Narrowing | |
6609 @cindex Widening | |
6610 | |
6611 Narrowing is a feature of Emacs that makes it possible for you to focus | |
6612 on a specific part of a buffer, and work without accidentally changing | |
6613 other parts. Narrowing is normally disabled since it can confuse | |
6614 novices. | |
6615 | |
6616 @menu | |
6617 * Narrowing advantages:: The advantages of narrowing | |
6618 * save-restriction:: The @code{save-restriction} special form. | |
6619 * what-line:: The number of the line that point is on. | |
6620 * narrow Exercise:: | |
6621 @end menu | |
6622 | |
6623 @node Narrowing advantages, save-restriction, Narrowing & Widening, Narrowing & Widening | |
6624 @ifnottex | |
6625 @unnumberedsec The Advantages of Narrowing | |
6626 @end ifnottex | |
6627 | |
6628 With narrowing, the rest of a buffer is made invisible, as if it weren't | |
6629 there. This is an advantage if, for example, you want to replace a word | |
6630 in one part of a buffer but not in another: you narrow to the part you want | |
6631 and the replacement is carried out only in that section, not in the rest | |
6632 of the buffer. Searches will only work within a narrowed region, not | |
6633 outside of one, so if you are fixing a part of a document, you can keep | |
6634 yourself from accidentally finding parts you do not need to fix by | |
6635 narrowing just to the region you want. | |
6636 (The key binding for @code{narrow-to-region} is @kbd{C-x n n}.) | |
6637 | |
6638 However, narrowing does make the rest of the buffer invisible, which | |
6639 can scare people who inadvertently invoke narrowing and think they | |
6640 have deleted a part of their file. Moreover, the @code{undo} command | |
6641 (which is usually bound to @kbd{C-x u}) does not turn off narrowing | |
6642 (nor should it), so people can become quite desperate if they do not | |
6643 know that they can return the rest of a buffer to visibility with the | |
6644 @code{widen} command. | |
6645 (The key binding for @code{widen} is @kbd{C-x n w}.) | |
6646 | |
6647 Narrowing is just as useful to the Lisp interpreter as to a human. | |
6648 Often, an Emacs Lisp function is designed to work on just part of a | |
6649 buffer; or conversely, an Emacs Lisp function needs to work on all of a | |
6650 buffer that has been narrowed. The @code{what-line} function, for | |
6651 example, removes the narrowing from a buffer, if it has any narrowing | |
6652 and when it has finished its job, restores the narrowing to what it was. | |
6653 On the other hand, the @code{count-lines} function, which is called by | |
6654 @code{what-line}, uses narrowing to restrict itself to just that portion | |
6655 of the buffer in which it is interested and then restores the previous | |
6656 situation. | |
6657 | |
6658 @node save-restriction, what-line, Narrowing advantages, Narrowing & Widening | |
6659 @comment node-name, next, previous, up | |
6660 @section The @code{save-restriction} Special Form | |
6661 @findex save-restriction | |
6662 | |
6663 In Emacs Lisp, you can use the @code{save-restriction} special form to | |
6664 keep track of whatever narrowing is in effect, if any. When the Lisp | |
6665 interpreter meets with @code{save-restriction}, it executes the code | |
6666 in the body of the @code{save-restriction} expression, and then undoes | |
6667 any changes to narrowing that the code caused. If, for example, the | |
6668 buffer is narrowed and the code that follows @code{save-restriction} | |
6669 gets rid of the narrowing, @code{save-restriction} returns the buffer | |
6670 to its narrowed region afterwards. In the @code{what-line} command, | |
6671 any narrowing the buffer may have is undone by the @code{widen} | |
6672 command that immediately follows the @code{save-restriction} command. | |
6673 Any original narrowing is restored just before the completion of the | |
6674 function. | |
6675 | |
6676 @need 1250 | |
6677 The template for a @code{save-restriction} expression is simple: | |
6678 | |
6679 @smallexample | |
6680 @group | |
6681 (save-restriction | |
6682 @var{body}@dots{} ) | |
6683 @end group | |
6684 @end smallexample | |
6685 | |
6686 @noindent | |
6687 The body of the @code{save-restriction} is one or more expressions that | |
6688 will be evaluated in sequence by the Lisp interpreter. | |
6689 | |
6690 Finally, a point to note: when you use both @code{save-excursion} and | |
6691 @code{save-restriction}, one right after the other, you should use | |
6692 @code{save-excursion} outermost. If you write them in reverse order, | |
6693 you may fail to record narrowing in the buffer to which Emacs switches | |
6694 after calling @code{save-excursion}. Thus, when written together, | |
6695 @code{save-excursion} and @code{save-restriction} should be written | |
6696 like this: | |
6697 | |
6698 @smallexample | |
6699 @group | |
6700 (save-excursion | |
6701 (save-restriction | |
6702 @var{body}@dots{})) | |
6703 @end group | |
6704 @end smallexample | |
6705 | |
6706 In other circumstances, when not written together, the | |
6707 @code{save-excursion} and @code{save-restriction} special forms must | |
6708 be written in the order appropriate to the function. | |
6709 | |
6710 @need 1250 | |
6711 For example, | |
6712 | |
6713 @smallexample | |
6714 @group | |
6715 (save-restriction | |
6716 (widen) | |
6717 (save-excursion | |
6718 @var{body}@dots{})) | |
6719 @end group | |
6720 @end smallexample | |
6721 | |
6722 @ignore | |
6723 Emacs 22 | |
6724 /usr/local/src/emacs/lisp/simple.el | |
6725 | |
6726 (defun what-line () | |
6727 "Print the current buffer line number and narrowed line number of point." | |
6728 (interactive) | |
6729 (let ((start (point-min)) | |
6730 (n (line-number-at-pos))) | |
6731 (if (= start 1) | |
6732 (message "Line %d" n) | |
6733 (save-excursion | |
6734 (save-restriction | |
6735 (widen) | |
6736 (message "line %d (narrowed line %d)" | |
6737 (+ n (line-number-at-pos start) -1) n)))))) | |
6738 | |
6739 (defun line-number-at-pos (&optional pos) | |
6740 "Return (narrowed) buffer line number at position POS. | |
6741 If POS is nil, use current buffer location. | |
6742 Counting starts at (point-min), so the value refers | |
6743 to the contents of the accessible portion of the buffer." | |
6744 (let ((opoint (or pos (point))) start) | |
6745 (save-excursion | |
6746 (goto-char (point-min)) | |
6747 (setq start (point)) | |
6748 (goto-char opoint) | |
6749 (forward-line 0) | |
6750 (1+ (count-lines start (point)))))) | |
6751 | |
6752 (defun count-lines (start end) | |
6753 "Return number of lines between START and END. | |
6754 This is usually the number of newlines between them, | |
6755 but can be one more if START is not equal to END | |
6756 and the greater of them is not at the start of a line." | |
6757 (save-excursion | |
6758 (save-restriction | |
6759 (narrow-to-region start end) | |
6760 (goto-char (point-min)) | |
6761 (if (eq selective-display t) | |
6762 (save-match-data | |
6763 (let ((done 0)) | |
6764 (while (re-search-forward "[\n\C-m]" nil t 40) | |
6765 (setq done (+ 40 done))) | |
6766 (while (re-search-forward "[\n\C-m]" nil t 1) | |
6767 (setq done (+ 1 done))) | |
6768 (goto-char (point-max)) | |
6769 (if (and (/= start end) | |
6770 (not (bolp))) | |
6771 (1+ done) | |
6772 done))) | |
6773 (- (buffer-size) (forward-line (buffer-size))))))) | |
6774 @end ignore | |
6775 | |
6776 @node what-line, narrow Exercise, save-restriction, Narrowing & Widening | |
6777 @comment node-name, next, previous, up | |
6778 @section @code{what-line} | |
6779 @findex what-line | |
6780 @cindex Widening, example of | |
6781 | |
6782 The @code{what-line} command tells you the number of the line in which | |
6783 the cursor is located. The function illustrates the use of the | |
6784 @code{save-restriction} and @code{save-excursion} commands. Here is the | |
6785 original text of the function: | |
6786 | |
6787 @smallexample | |
6788 @group | |
6789 (defun what-line () | |
6790 "Print the current line number (in the buffer) of point." | |
6791 (interactive) | |
6792 (save-restriction | |
6793 (widen) | |
6794 (save-excursion | |
6795 (beginning-of-line) | |
6796 (message "Line %d" | |
6797 (1+ (count-lines 1 (point))))))) | |
6798 @end group | |
6799 @end smallexample | |
6800 | |
6801 (In recent versions of GNU Emacs, the @code{what-line} function has | |
6802 been expanded to tell you your line number in a narrowed buffer as | |
6803 well as your line number in a widened buffer. The recent version is | |
6804 more complex than the version shown here. If you feel adventurous, | |
6805 you might want to look at it after figuring out how this version | |
6806 works. You will probably need to use @kbd{C-h f} | |
6807 (@code{describe-function}). The newer version uses a conditional to | |
6808 determine whether the buffer has been narrowed. | |
6809 | |
6810 (Also, it uses @code{line-number-at-pos}, which among other simple | |
6811 expressions, such as @code{(goto-char (point-min))}, moves point to | |
6812 the beginning of the current line with @code{(forward-line 0)} rather | |
6813 than @code{beginning-of-line}.) | |
6814 | |
6815 The @code{what-line} function as shown here has a documentation line | |
6816 and is interactive, as you would expect. The next two lines use the | |
6817 functions @code{save-restriction} and @code{widen}. | |
6818 | |
6819 The @code{save-restriction} special form notes whatever narrowing is in | |
6820 effect, if any, in the current buffer and restores that narrowing after | |
6821 the code in the body of the @code{save-restriction} has been evaluated. | |
6822 | |
6823 The @code{save-restriction} special form is followed by @code{widen}. | |
6824 This function undoes any narrowing the current buffer may have had | |
6825 when @code{what-line} was called. (The narrowing that was there is | |
6826 the narrowing that @code{save-restriction} remembers.) This widening | |
6827 makes it possible for the line counting commands to count from the | |
6828 beginning of the buffer. Otherwise, they would have been limited to | |
6829 counting within the accessible region. Any original narrowing is | |
6830 restored just before the completion of the function by the | |
6831 @code{save-restriction} special form. | |
6832 | |
6833 The call to @code{widen} is followed by @code{save-excursion}, which | |
6834 saves the location of the cursor (i.e., of point) and of the mark, and | |
6835 restores them after the code in the body of the @code{save-excursion} | |
6836 uses the @code{beginning-of-line} function to move point. | |
6837 | |
6838 (Note that the @code{(widen)} expression comes between the | |
6839 @code{save-restriction} and @code{save-excursion} special forms. When | |
6840 you write the two @code{save- @dots{}} expressions in sequence, write | |
6841 @code{save-excursion} outermost.) | |
6842 | |
6843 @need 1200 | |
6844 The last two lines of the @code{what-line} function are functions to | |
6845 count the number of lines in the buffer and then print the number in the | |
6846 echo area. | |
6847 | |
6848 @smallexample | |
6849 @group | |
6850 (message "Line %d" | |
6851 (1+ (count-lines 1 (point))))))) | |
6852 @end group | |
6853 @end smallexample | |
6854 | |
6855 The @code{message} function prints a one-line message at the bottom of | |
6856 the Emacs screen. The first argument is inside of quotation marks and | |
6857 is printed as a string of characters. However, it may contain a | |
6858 @samp{%d} expression to print a following argument. @samp{%d} prints | |
6859 the argument as a decimal, so the message will say something such as | |
6860 @samp{Line 243}. | |
6861 | |
6862 @need 1200 | |
6863 The number that is printed in place of the @samp{%d} is computed by the | |
6864 last line of the function: | |
6865 | |
6866 @smallexample | |
6867 (1+ (count-lines 1 (point))) | |
6868 @end smallexample | |
6869 | |
6870 @ignore | |
6871 GNU Emacs 22 | |
6872 | |
6873 (defun count-lines (start end) | |
6874 "Return number of lines between START and END. | |
6875 This is usually the number of newlines between them, | |
6876 but can be one more if START is not equal to END | |
6877 and the greater of them is not at the start of a line." | |
6878 (save-excursion | |
6879 (save-restriction | |
6880 (narrow-to-region start end) | |
6881 (goto-char (point-min)) | |
6882 (if (eq selective-display t) | |
6883 (save-match-data | |
6884 (let ((done 0)) | |
6885 (while (re-search-forward "[\n\C-m]" nil t 40) | |
6886 (setq done (+ 40 done))) | |
6887 (while (re-search-forward "[\n\C-m]" nil t 1) | |
6888 (setq done (+ 1 done))) | |
6889 (goto-char (point-max)) | |
6890 (if (and (/= start end) | |
6891 (not (bolp))) | |
6892 (1+ done) | |
6893 done))) | |
6894 (- (buffer-size) (forward-line (buffer-size))))))) | |
6895 @end ignore | |
6896 | |
6897 @noindent | |
6898 What this does is count the lines from the first position of the | |
6899 buffer, indicated by the @code{1}, up to @code{(point)}, and then add | |
6900 one to that number. (The @code{1+} function adds one to its | |
6901 argument.) We add one to it because line 2 has only one line before | |
6902 it, and @code{count-lines} counts only the lines @emph{before} the | |
6903 current line. | |
6904 | |
6905 After @code{count-lines} has done its job, and the message has been | |
6906 printed in the echo area, the @code{save-excursion} restores point and | |
6907 mark to their original positions; and @code{save-restriction} restores | |
6908 the original narrowing, if any. | |
6909 | |
6910 @node narrow Exercise, , what-line, Narrowing & Widening | |
6911 @section Exercise with Narrowing | |
6912 | |
6913 Write a function that will display the first 60 characters of the | |
6914 current buffer, even if you have narrowed the buffer to its latter | |
6915 half so that the first line is inaccessible. Restore point, mark, and | |
6916 narrowing. For this exercise, you need to use a whole potpourri of | |
6917 functions, including @code{save-restriction}, @code{widen}, | |
6918 @code{goto-char}, @code{point-min}, @code{message}, and | |
6919 @code{buffer-substring}. | |
6920 | |
6921 @cindex Properties, mention of @code{buffer-substring-no-properties} | |
6922 (@code{buffer-substring} is a previously unmentioned function you will | |
6923 have to investigate yourself; or perhaps you will have to use | |
6924 @code{buffer-substring-no-properties} or | |
6925 @code{filter-buffer-substring} @dots{}, yet other functions. Text | |
6926 properties are a feature otherwise not discussed here. @xref{Text | |
6927 Properties, , Text Properties, elisp, The GNU Emacs Lisp Reference | |
6928 Manual}.) | |
6929 | |
6930 Additionally, do you really need @code{goto-char} or @code{point-min}? | |
6931 Or can you write the function without them? | |
6932 | |
6933 @node car cdr & cons, Cutting & Storing Text, Narrowing & Widening, Top | |
6934 @comment node-name, next, previous, up | |
6935 @chapter @code{car}, @code{cdr}, @code{cons}: Fundamental Functions | |
6936 @findex car, @r{introduced} | |
6937 @findex cdr, @r{introduced} | |
6938 | |
6939 In Lisp, @code{car}, @code{cdr}, and @code{cons} are fundamental | |
6940 functions. The @code{cons} function is used to construct lists, and | |
6941 the @code{car} and @code{cdr} functions are used to take them apart. | |
6942 | |
6943 In the walk through of the @code{copy-region-as-kill} function, we | |
6944 will see @code{cons} as well as two variants on @code{cdr}, | |
6945 namely, @code{setcdr} and @code{nthcdr}. (@xref{copy-region-as-kill}.) | |
6946 | |
6947 @menu | |
6948 * Strange Names:: An historical aside: why the strange names? | |
6949 * car & cdr:: Functions for extracting part of a list. | |
6950 * cons:: Constructing a list. | |
6951 * nthcdr:: Calling @code{cdr} repeatedly. | |
6952 * nth:: | |
6953 * setcar:: Changing the first element of a list. | |
6954 * setcdr:: Changing the rest of a list. | |
6955 * cons Exercise:: | |
6956 @end menu | |
6957 | |
6958 @node Strange Names, car & cdr, car cdr & cons, car cdr & cons | |
6959 @ifnottex | |
6960 @unnumberedsec Strange Names | |
6961 @end ifnottex | |
6962 | |
6963 The name of the @code{cons} function is not unreasonable: it is an | |
6964 abbreviation of the word `construct'. The origins of the names for | |
6965 @code{car} and @code{cdr}, on the other hand, are esoteric: @code{car} | |
6966 is an acronym from the phrase `Contents of the Address part of the | |
6967 Register'; and @code{cdr} (pronounced `could-er') is an acronym from | |
6968 the phrase `Contents of the Decrement part of the Register'. These | |
6969 phrases refer to specific pieces of hardware on the very early | |
6970 computer on which the original Lisp was developed. Besides being | |
6971 obsolete, the phrases have been completely irrelevant for more than 25 | |
6972 years to anyone thinking about Lisp. Nonetheless, although a few | |
6973 brave scholars have begun to use more reasonable names for these | |
6974 functions, the old terms are still in use. In particular, since the | |
6975 terms are used in the Emacs Lisp source code, we will use them in this | |
6976 introduction. | |
6977 | |
6978 @node car & cdr, cons, Strange Names, car cdr & cons | |
6979 @comment node-name, next, previous, up | |
6980 @section @code{car} and @code{cdr} | |
6981 | |
6982 The @sc{car} of a list is, quite simply, the first item in the list. | |
6983 Thus the @sc{car} of the list @code{(rose violet daisy buttercup)} is | |
6984 @code{rose}. | |
6985 | |
6986 @need 1200 | |
6987 If you are reading this in Info in GNU Emacs, you can see this by | |
6988 evaluating the following: | |
6989 | |
6990 @smallexample | |
6991 (car '(rose violet daisy buttercup)) | |
6992 @end smallexample | |
6993 | |
6994 @noindent | |
6995 After evaluating the expression, @code{rose} will appear in the echo | |
6996 area. | |
6997 | |
6998 Clearly, a more reasonable name for the @code{car} function would be | |
6999 @code{first} and this is often suggested. | |
7000 | |
7001 @code{car} does not remove the first item from the list; it only reports | |
7002 what it is. After @code{car} has been applied to a list, the list is | |
7003 still the same as it was. In the jargon, @code{car} is | |
7004 `non-destructive'. This feature turns out to be important. | |
7005 | |
7006 The @sc{cdr} of a list is the rest of the list, that is, the | |
7007 @code{cdr} function returns the part of the list that follows the | |
7008 first item. Thus, while the @sc{car} of the list @code{'(rose violet | |
7009 daisy buttercup)} is @code{rose}, the rest of the list, the value | |
7010 returned by the @code{cdr} function, is @code{(violet daisy | |
7011 buttercup)}. | |
7012 | |
7013 @need 800 | |
7014 You can see this by evaluating the following in the usual way: | |
7015 | |
7016 @smallexample | |
7017 (cdr '(rose violet daisy buttercup)) | |
7018 @end smallexample | |
7019 | |
7020 @noindent | |
7021 When you evaluate this, @code{(violet daisy buttercup)} will appear in | |
7022 the echo area. | |
7023 | |
7024 Like @code{car}, @code{cdr} does not remove any elements from the | |
7025 list---it just returns a report of what the second and subsequent | |
7026 elements are. | |
7027 | |
7028 Incidentally, in the example, the list of flowers is quoted. If it were | |
7029 not, the Lisp interpreter would try to evaluate the list by calling | |
7030 @code{rose} as a function. In this example, we do not want to do that. | |
7031 | |
7032 Clearly, a more reasonable name for @code{cdr} would be @code{rest}. | |
7033 | |
7034 (There is a lesson here: when you name new functions, consider very | |
7035 carefully what you are doing, since you may be stuck with the names | |
7036 for far longer than you expect. The reason this document perpetuates | |
7037 these names is that the Emacs Lisp source code uses them, and if I did | |
7038 not use them, you would have a hard time reading the code; but do, | |
7039 please, try to avoid using these terms yourself. The people who come | |
7040 after you will be grateful to you.) | |
7041 | |
7042 When @code{car} and @code{cdr} are applied to a list made up of symbols, | |
7043 such as the list @code{(pine fir oak maple)}, the element of the list | |
7044 returned by the function @code{car} is the symbol @code{pine} without | |
7045 any parentheses around it. @code{pine} is the first element in the | |
7046 list. However, the @sc{cdr} of the list is a list itself, @code{(fir | |
7047 oak maple)}, as you can see by evaluating the following expressions in | |
7048 the usual way: | |
7049 | |
7050 @smallexample | |
7051 @group | |
7052 (car '(pine fir oak maple)) | |
7053 | |
7054 (cdr '(pine fir oak maple)) | |
7055 @end group | |
7056 @end smallexample | |
7057 | |
7058 On the other hand, in a list of lists, the first element is itself a | |
7059 list. @code{car} returns this first element as a list. For example, | |
7060 the following list contains three sub-lists, a list of carnivores, a | |
7061 list of herbivores and a list of sea mammals: | |
7062 | |
7063 @smallexample | |
7064 @group | |
7065 (car '((lion tiger cheetah) | |
7066 (gazelle antelope zebra) | |
7067 (whale dolphin seal))) | |
7068 @end group | |
7069 @end smallexample | |
7070 | |
7071 @noindent | |
7072 In this example, the first element or @sc{car} of the list is the list of | |
7073 carnivores, @code{(lion tiger cheetah)}, and the rest of the list is | |
7074 @code{((gazelle antelope zebra) (whale dolphin seal))}. | |
7075 | |
7076 @smallexample | |
7077 @group | |
7078 (cdr '((lion tiger cheetah) | |
7079 (gazelle antelope zebra) | |
7080 (whale dolphin seal))) | |
7081 @end group | |
7082 @end smallexample | |
7083 | |
7084 It is worth saying again that @code{car} and @code{cdr} are | |
7085 non-destructive---that is, they do not modify or change lists to which | |
7086 they are applied. This is very important for how they are used. | |
7087 | |
7088 Also, in the first chapter, in the discussion about atoms, I said that | |
7089 in Lisp, ``certain kinds of atom, such as an array, can be separated | |
7090 into parts; but the mechanism for doing this is different from the | |
7091 mechanism for splitting a list. As far as Lisp is concerned, the | |
7092 atoms of a list are unsplittable.'' (@xref{Lisp Atoms}.) The | |
7093 @code{car} and @code{cdr} functions are used for splitting lists and | |
7094 are considered fundamental to Lisp. Since they cannot split or gain | |
7095 access to the parts of an array, an array is considered an atom. | |
7096 Conversely, the other fundamental function, @code{cons}, can put | |
7097 together or construct a list, but not an array. (Arrays are handled | |
7098 by array-specific functions. @xref{Arrays, , Arrays, elisp, The GNU | |
7099 Emacs Lisp Reference Manual}.) | |
7100 | |
7101 @node cons, nthcdr, car & cdr, car cdr & cons | |
7102 @comment node-name, next, previous, up | |
7103 @section @code{cons} | |
7104 @findex cons, @r{introduced} | |
7105 | |
7106 The @code{cons} function constructs lists; it is the inverse of | |
7107 @code{car} and @code{cdr}. For example, @code{cons} can be used to make | |
7108 a four element list from the three element list, @code{(fir oak maple)}: | |
7109 | |
7110 @smallexample | |
7111 (cons 'pine '(fir oak maple)) | |
7112 @end smallexample | |
7113 | |
7114 @need 800 | |
7115 @noindent | |
7116 After evaluating this list, you will see | |
7117 | |
7118 @smallexample | |
7119 (pine fir oak maple) | |
7120 @end smallexample | |
7121 | |
7122 @noindent | |
7123 appear in the echo area. @code{cons} causes the creation of a new | |
7124 list in which the element is followed by the elements of the original | |
7125 list. | |
7126 | |
7127 We often say that `@code{cons} puts a new element at the beginning of | |
7128 a list; it attaches or pushes elements onto the list', but this | |
7129 phrasing can be misleading, since @code{cons} does not change an | |
7130 existing list, but creates a new one. | |
7131 | |
7132 Like @code{car} and @code{cdr}, @code{cons} is non-destructive. | |
7133 | |
7134 @menu | |
7135 * Build a list:: | |
7136 * length:: How to find the length of a list. | |
7137 @end menu | |
7138 | |
7139 @node Build a list, length, cons, cons | |
7140 @ifnottex | |
7141 @unnumberedsubsec Build a list | |
7142 @end ifnottex | |
7143 | |
7144 @code{cons} must have a list to attach to.@footnote{Actually, you can | |
7145 @code{cons} an element to an atom to produce a dotted pair. Dotted | |
7146 pairs are not discussed here; see @ref{Dotted Pair Notation, , Dotted | |
7147 Pair Notation, elisp, The GNU Emacs Lisp Reference Manual}.} You | |
7148 cannot start from absolutely nothing. If you are building a list, you | |
7149 need to provide at least an empty list at the beginning. Here is a | |
7150 series of @code{cons} expressions that build up a list of flowers. If | |
7151 you are reading this in Info in GNU Emacs, you can evaluate each of | |
7152 the expressions in the usual way; the value is printed in this text | |
7153 after @samp{@result{}}, which you may read as `evaluates to'. | |
7154 | |
7155 @smallexample | |
7156 @group | |
7157 (cons 'buttercup ()) | |
7158 @result{} (buttercup) | |
7159 @end group | |
7160 | |
7161 @group | |
7162 (cons 'daisy '(buttercup)) | |
7163 @result{} (daisy buttercup) | |
7164 @end group | |
7165 | |
7166 @group | |
7167 (cons 'violet '(daisy buttercup)) | |
7168 @result{} (violet daisy buttercup) | |
7169 @end group | |
7170 | |
7171 @group | |
7172 (cons 'rose '(violet daisy buttercup)) | |
7173 @result{} (rose violet daisy buttercup) | |
7174 @end group | |
7175 @end smallexample | |
7176 | |
7177 @noindent | |
7178 In the first example, the empty list is shown as @code{()} and a list | |
7179 made up of @code{buttercup} followed by the empty list is constructed. | |
7180 As you can see, the empty list is not shown in the list that was | |
7181 constructed. All that you see is @code{(buttercup)}. The empty list is | |
7182 not counted as an element of a list because there is nothing in an empty | |
7183 list. Generally speaking, an empty list is invisible. | |
7184 | |
7185 The second example, @code{(cons 'daisy '(buttercup))} constructs a new, | |
7186 two element list by putting @code{daisy} in front of @code{buttercup}; | |
7187 and the third example constructs a three element list by putting | |
7188 @code{violet} in front of @code{daisy} and @code{buttercup}. | |
7189 | |
7190 @node length, , Build a list, cons | |
7191 @comment node-name, next, previous, up | |
7192 @subsection Find the Length of a List: @code{length} | |
7193 @findex length | |
7194 | |
7195 You can find out how many elements there are in a list by using the Lisp | |
7196 function @code{length}, as in the following examples: | |
7197 | |
7198 @smallexample | |
7199 @group | |
7200 (length '(buttercup)) | |
7201 @result{} 1 | |
7202 @end group | |
7203 | |
7204 @group | |
7205 (length '(daisy buttercup)) | |
7206 @result{} 2 | |
7207 @end group | |
7208 | |
7209 @group | |
7210 (length (cons 'violet '(daisy buttercup))) | |
7211 @result{} 3 | |
7212 @end group | |
7213 @end smallexample | |
7214 | |
7215 @noindent | |
7216 In the third example, the @code{cons} function is used to construct a | |
7217 three element list which is then passed to the @code{length} function as | |
7218 its argument. | |
7219 | |
7220 @need 1200 | |
7221 We can also use @code{length} to count the number of elements in an | |
7222 empty list: | |
7223 | |
7224 @smallexample | |
7225 @group | |
7226 (length ()) | |
7227 @result{} 0 | |
7228 @end group | |
7229 @end smallexample | |
7230 | |
7231 @noindent | |
7232 As you would expect, the number of elements in an empty list is zero. | |
7233 | |
7234 An interesting experiment is to find out what happens if you try to find | |
7235 the length of no list at all; that is, if you try to call @code{length} | |
7236 without giving it an argument, not even an empty list: | |
7237 | |
7238 @smallexample | |
7239 (length ) | |
7240 @end smallexample | |
7241 | |
7242 @need 800 | |
7243 @noindent | |
7244 What you see, if you evaluate this, is the error message | |
7245 | |
7246 @smallexample | |
7247 Lisp error: (wrong-number-of-arguments length 0) | |
7248 @end smallexample | |
7249 | |
7250 @noindent | |
7251 This means that the function receives the wrong number of | |
7252 arguments, zero, when it expects some other number of arguments. In | |
7253 this case, one argument is expected, the argument being a list whose | |
7254 length the function is measuring. (Note that @emph{one} list is | |
7255 @emph{one} argument, even if the list has many elements inside it.) | |
7256 | |
7257 The part of the error message that says @samp{length} is the name of | |
7258 the function. | |
7259 | |
7260 @ignore | |
7261 @code{length} is still a subroutine, but you need C-h f to discover that. | |
7262 | |
7263 In an earlier version: | |
7264 This is written with a special notation, @samp{#<subr}, | |
7265 that indicates that the function @code{length} is one of the primitive | |
7266 functions written in C rather than in Emacs Lisp. (@samp{subr} is an | |
7267 abbreviation for `subroutine'.) @xref{What Is a Function, , What Is a | |
7268 Function?, elisp , The GNU Emacs Lisp Reference Manual}, for more | |
7269 about subroutines. | |
7270 @end ignore | |
7271 | |
7272 @node nthcdr, nth, cons, car cdr & cons | |
7273 @comment node-name, next, previous, up | |
7274 @section @code{nthcdr} | |
7275 @findex nthcdr | |
7276 | |
7277 The @code{nthcdr} function is associated with the @code{cdr} function. | |
7278 What it does is take the @sc{cdr} of a list repeatedly. | |
7279 | |
7280 If you take the @sc{cdr} of the list @code{(pine fir | |
7281 oak maple)}, you will be returned the list @code{(fir oak maple)}. If you | |
7282 repeat this on what was returned, you will be returned the list | |
7283 @code{(oak maple)}. (Of course, repeated @sc{cdr}ing on the original | |
7284 list will just give you the original @sc{cdr} since the function does | |
7285 not change the list. You need to evaluate the @sc{cdr} of the | |
7286 @sc{cdr} and so on.) If you continue this, eventually you will be | |
7287 returned an empty list, which in this case, instead of being shown as | |
7288 @code{()} is shown as @code{nil}. | |
7289 | |
7290 @need 1200 | |
7291 For review, here is a series of repeated @sc{cdr}s, the text following | |
7292 the @samp{@result{}} shows what is returned. | |
7293 | |
7294 @smallexample | |
7295 @group | |
7296 (cdr '(pine fir oak maple)) | |
7297 @result{}(fir oak maple) | |
7298 @end group | |
7299 | |
7300 @group | |
7301 (cdr '(fir oak maple)) | |
7302 @result{} (oak maple) | |
7303 @end group | |
7304 | |
7305 @group | |
7306 (cdr '(oak maple)) | |
7307 @result{}(maple) | |
7308 @end group | |
7309 | |
7310 @group | |
7311 (cdr '(maple)) | |
7312 @result{} nil | |
7313 @end group | |
7314 | |
7315 @group | |
7316 (cdr 'nil) | |
7317 @result{} nil | |
7318 @end group | |
7319 | |
7320 @group | |
7321 (cdr ()) | |
7322 @result{} nil | |
7323 @end group | |
7324 @end smallexample | |
7325 | |
7326 @need 1200 | |
7327 You can also do several @sc{cdr}s without printing the values in | |
7328 between, like this: | |
7329 | |
7330 @smallexample | |
7331 @group | |
7332 (cdr (cdr '(pine fir oak maple))) | |
7333 @result{} (oak maple) | |
7334 @end group | |
7335 @end smallexample | |
7336 | |
7337 @noindent | |
7338 In this example, the Lisp interpreter evaluates the innermost list first. | |
7339 The innermost list is quoted, so it just passes the list as it is to the | |
7340 innermost @code{cdr}. This @code{cdr} passes a list made up of the | |
7341 second and subsequent elements of the list to the outermost @code{cdr}, | |
7342 which produces a list composed of the third and subsequent elements of | |
7343 the original list. In this example, the @code{cdr} function is repeated | |
7344 and returns a list that consists of the original list without its | |
7345 first two elements. | |
7346 | |
7347 The @code{nthcdr} function does the same as repeating the call to | |
7348 @code{cdr}. In the following example, the argument 2 is passed to the | |
7349 function @code{nthcdr}, along with the list, and the value returned is | |
7350 the list without its first two items, which is exactly the same | |
7351 as repeating @code{cdr} twice on the list: | |
7352 | |
7353 @smallexample | |
7354 @group | |
7355 (nthcdr 2 '(pine fir oak maple)) | |
7356 @result{} (oak maple) | |
7357 @end group | |
7358 @end smallexample | |
7359 | |
7360 @need 1200 | |
7361 Using the original four element list, we can see what happens when | |
7362 various numeric arguments are passed to @code{nthcdr}, including 0, 1, | |
7363 and 5: | |
7364 | |
7365 @smallexample | |
7366 @group | |
7367 ;; @r{Leave the list as it was.} | |
7368 (nthcdr 0 '(pine fir oak maple)) | |
7369 @result{} (pine fir oak maple) | |
7370 @end group | |
7371 | |
7372 @group | |
7373 ;; @r{Return a copy without the first element.} | |
7374 (nthcdr 1 '(pine fir oak maple)) | |
7375 @result{} (fir oak maple) | |
7376 @end group | |
7377 | |
7378 @group | |
7379 ;; @r{Return a copy of the list without three elements.} | |
7380 (nthcdr 3 '(pine fir oak maple)) | |
7381 @result{} (maple) | |
7382 @end group | |
7383 | |
7384 @group | |
7385 ;; @r{Return a copy lacking all four elements.} | |
7386 (nthcdr 4 '(pine fir oak maple)) | |
7387 @result{} nil | |
7388 @end group | |
7389 | |
7390 @group | |
7391 ;; @r{Return a copy lacking all elements.} | |
7392 (nthcdr 5 '(pine fir oak maple)) | |
7393 @result{} nil | |
7394 @end group | |
7395 @end smallexample | |
7396 | |
7397 @node nth, setcar, nthcdr, car cdr & cons | |
7398 @comment node-name, next, previous, up | |
7399 @section @code{nth} | |
7400 @findex nth | |
7401 | |
7402 The @code{nthcdr} function takes the @sc{cdr} of a list repeatedly. | |
7403 The @code{nth} function takes the @sc{car} of the result returned by | |
7404 @code{nthcdr}. It returns the Nth element of the list. | |
7405 | |
7406 @need 1500 | |
7407 Thus, if it were not defined in C for speed, the definition of | |
7408 @code{nth} would be: | |
7409 | |
7410 @smallexample | |
7411 @group | |
7412 (defun nth (n list) | |
7413 "Returns the Nth element of LIST. | |
7414 N counts from zero. If LIST is not that long, nil is returned." | |
7415 (car (nthcdr n list))) | |
7416 @end group | |
7417 @end smallexample | |
7418 | |
7419 @noindent | |
7420 (Originally, @code{nth} was defined in Emacs Lisp in @file{subr.el}, | |
7421 but its definition was redone in C in the 1980s.) | |
7422 | |
7423 The @code{nth} function returns a single element of a list. | |
7424 This can be very convenient. | |
7425 | |
7426 Note that the elements are numbered from zero, not one. That is to | |
7427 say, the first element of a list, its @sc{car} is the zeroth element. | |
7428 This is called `zero-based' counting and often bothers people who | |
7429 are accustomed to the first element in a list being number one, which | |
7430 is `one-based'. | |
7431 | |
7432 @need 1250 | |
7433 For example: | |
7434 | |
7435 @smallexample | |
7436 @group | |
7437 (nth 0 '("one" "two" "three")) | |
7438 @result{} "one" | |
7439 | |
7440 (nth 1 '("one" "two" "three")) | |
7441 @result{} "two" | |
7442 @end group | |
7443 @end smallexample | |
7444 | |
7445 It is worth mentioning that @code{nth}, like @code{nthcdr} and | |
7446 @code{cdr}, does not change the original list---the function is | |
7447 non-destructive. This is in sharp contrast to the @code{setcar} and | |
7448 @code{setcdr} functions. | |
7449 | |
7450 @node setcar, setcdr, nth, car cdr & cons | |
7451 @comment node-name, next, previous, up | |
7452 @section @code{setcar} | |
7453 @findex setcar | |
7454 | |
7455 As you might guess from their names, the @code{setcar} and @code{setcdr} | |
7456 functions set the @sc{car} or the @sc{cdr} of a list to a new value. | |
7457 They actually change the original list, unlike @code{car} and @code{cdr} | |
7458 which leave the original list as it was. One way to find out how this | |
7459 works is to experiment. We will start with the @code{setcar} function. | |
7460 | |
7461 @need 1200 | |
7462 First, we can make a list and then set the value of a variable to the | |
7463 list, using the @code{setq} function. Here is a list of animals: | |
7464 | |
7465 @smallexample | |
7466 (setq animals '(antelope giraffe lion tiger)) | |
7467 @end smallexample | |
7468 | |
7469 @noindent | |
7470 If you are reading this in Info inside of GNU Emacs, you can evaluate | |
7471 this expression in the usual fashion, by positioning the cursor after | |
7472 the expression and typing @kbd{C-x C-e}. (I'm doing this right here | |
7473 as I write this. This is one of the advantages of having the | |
7474 interpreter built into the computing environment. Incidentally, when | |
7475 there is nothing on the line after the final parentheses, such as a | |
7476 comment, point can be on the next line. Thus, if your cursor is in | |
7477 the first column of the next line, you do not need to move it. | |
7478 Indeed, Emacs permits any amount of white space after the final | |
7479 parenthesis.) | |
7480 | |
7481 @need 1200 | |
7482 When we evaluate the variable @code{animals}, we see that it is bound to | |
7483 the list @code{(antelope giraffe lion tiger)}: | |
7484 | |
7485 @smallexample | |
7486 @group | |
7487 animals | |
7488 @result{} (antelope giraffe lion tiger) | |
7489 @end group | |
7490 @end smallexample | |
7491 | |
7492 @noindent | |
7493 Put another way, the variable @code{animals} points to the list | |
7494 @code{(antelope giraffe lion tiger)}. | |
7495 | |
7496 Next, evaluate the function @code{setcar} while passing it two | |
7497 arguments, the variable @code{animals} and the quoted symbol | |
7498 @code{hippopotamus}; this is done by writing the three element list | |
7499 @code{(setcar animals 'hippopotamus)} and then evaluating it in the | |
7500 usual fashion: | |
7501 | |
7502 @smallexample | |
7503 (setcar animals 'hippopotamus) | |
7504 @end smallexample | |
7505 | |
7506 @need 1200 | |
7507 @noindent | |
7508 After evaluating this expression, evaluate the variable @code{animals} | |
7509 again. You will see that the list of animals has changed: | |
7510 | |
7511 @smallexample | |
7512 @group | |
7513 animals | |
7514 @result{} (hippopotamus giraffe lion tiger) | |
7515 @end group | |
7516 @end smallexample | |
7517 | |
7518 @noindent | |
7519 The first element on the list, @code{antelope} is replaced by | |
7520 @code{hippopotamus}. | |
7521 | |
7522 So we can see that @code{setcar} did not add a new element to the list | |
7523 as @code{cons} would have; it replaced @code{antelope} with | |
7524 @code{hippopotamus}; it @emph{changed} the list. | |
7525 | |
7526 @node setcdr, cons Exercise, setcar, car cdr & cons | |
7527 @comment node-name, next, previous, up | |
7528 @section @code{setcdr} | |
7529 @findex setcdr | |
7530 | |
7531 The @code{setcdr} function is similar to the @code{setcar} function, | |
7532 except that the function replaces the second and subsequent elements of | |
7533 a list rather than the first element. | |
7534 | |
7535 (To see how to change the last element of a list, look ahead to | |
7536 @ref{kill-new function, , The @code{kill-new} function}, which uses | |
7537 the @code{nthcdr} and @code{setcdr} functions.) | |
7538 | |
7539 @need 1200 | |
7540 To see how this works, set the value of the variable to a list of | |
7541 domesticated animals by evaluating the following expression: | |
7542 | |
7543 @smallexample | |
7544 (setq domesticated-animals '(horse cow sheep goat)) | |
7545 @end smallexample | |
7546 | |
7547 @need 1200 | |
7548 @noindent | |
7549 If you now evaluate the list, you will be returned the list | |
7550 @code{(horse cow sheep goat)}: | |
7551 | |
7552 @smallexample | |
7553 @group | |
7554 domesticated-animals | |
7555 @result{} (horse cow sheep goat) | |
7556 @end group | |
7557 @end smallexample | |
7558 | |
7559 @need 1200 | |
7560 Next, evaluate @code{setcdr} with two arguments, the name of the | |
7561 variable which has a list as its value, and the list to which the | |
7562 @sc{cdr} of the first list will be set; | |
7563 | |
7564 @smallexample | |
7565 (setcdr domesticated-animals '(cat dog)) | |
7566 @end smallexample | |
7567 | |
7568 @noindent | |
7569 If you evaluate this expression, the list @code{(cat dog)} will appear | |
7570 in the echo area. This is the value returned by the function. The | |
7571 result we are interested in is the ``side effect'', which we can see by | |
7572 evaluating the variable @code{domesticated-animals}: | |
7573 | |
7574 @smallexample | |
7575 @group | |
7576 domesticated-animals | |
7577 @result{} (horse cat dog) | |
7578 @end group | |
7579 @end smallexample | |
7580 | |
7581 @noindent | |
7582 Indeed, the list is changed from @code{(horse cow sheep goat)} to | |
7583 @code{(horse cat dog)}. The @sc{cdr} of the list is changed from | |
7584 @code{(cow sheep goat)} to @code{(cat dog)}. | |
7585 | |
7586 @node cons Exercise, , setcdr, car cdr & cons | |
7587 @section Exercise | |
7588 | |
7589 Construct a list of four birds by evaluating several expressions with | |
7590 @code{cons}. Find out what happens when you @code{cons} a list onto | |
7591 itself. Replace the first element of the list of four birds with a | |
7592 fish. Replace the rest of that list with a list of other fish. | |
7593 | |
7594 @node Cutting & Storing Text, List Implementation, car cdr & cons, Top | |
7595 @comment node-name, next, previous, up | |
7596 @chapter Cutting and Storing Text | |
7597 @cindex Cutting and storing text | |
7598 @cindex Storing and cutting text | |
7599 @cindex Killing text | |
7600 @cindex Clipping text | |
7601 @cindex Erasing text | |
7602 @cindex Deleting text | |
7603 | |
7604 Whenever you cut or clip text out of a buffer with a `kill' command in | |
7605 GNU Emacs, it is stored in a list and you can bring it back with a | |
7606 `yank' command. | |
7607 | |
7608 (The use of the word `kill' in Emacs for processes which specifically | |
7609 @emph{do not} destroy the values of the entities is an unfortunate | |
7610 historical accident. A much more appropriate word would be `clip' since | |
7611 that is what the kill commands do; they clip text out of a buffer and | |
7612 put it into storage from which it can be brought back. I have often | |
7613 been tempted to replace globally all occurrences of `kill' in the Emacs | |
7614 sources with `clip' and all occurrences of `killed' with `clipped'.) | |
7615 | |
7616 @menu | |
7617 * Storing Text:: Text is stored in a list. | |
7618 * zap-to-char:: Cutting out text up to a character. | |
7619 * kill-region:: Cutting text out of a region. | |
7620 * copy-region-as-kill:: A definition for copying text. | |
7621 * Digression into C:: Minor note on C programming language macros. | |
7622 * defvar:: How to give a variable an initial value. | |
7623 * cons & search-fwd Review:: | |
7624 * search Exercises:: | |
7625 @end menu | |
7626 | |
7627 @node Storing Text, zap-to-char, Cutting & Storing Text, Cutting & Storing Text | |
7628 @ifnottex | |
7629 @unnumberedsec Storing Text in a List | |
7630 @end ifnottex | |
7631 | |
7632 When text is cut out of a buffer, it is stored on a list. Successive | |
7633 pieces of text are stored on the list successively, so the list might | |
7634 look like this: | |
7635 | |
7636 @smallexample | |
7637 ("a piece of text" "previous piece") | |
7638 @end smallexample | |
7639 | |
7640 @need 1200 | |
7641 @noindent | |
7642 The function @code{cons} can be used to create a new list from a piece | |
7643 of text (an `atom', to use the jargon) and an existing list, like | |
7644 this: | |
7645 | |
7646 @smallexample | |
7647 @group | |
7648 (cons "another piece" | |
7649 '("a piece of text" "previous piece")) | |
7650 @end group | |
7651 @end smallexample | |
7652 | |
7653 @need 1200 | |
7654 @noindent | |
7655 If you evaluate this expression, a list of three elements will appear in | |
7656 the echo area: | |
7657 | |
7658 @smallexample | |
7659 ("another piece" "a piece of text" "previous piece") | |
7660 @end smallexample | |
7661 | |
7662 With the @code{car} and @code{nthcdr} functions, you can retrieve | |
7663 whichever piece of text you want. For example, in the following code, | |
7664 @code{nthcdr 1 @dots{}} returns the list with the first item removed; | |
7665 and the @code{car} returns the first element of that remainder---the | |
7666 second element of the original list: | |
7667 | |
7668 @smallexample | |
7669 @group | |
7670 (car (nthcdr 1 '("another piece" | |
7671 "a piece of text" | |
7672 "previous piece"))) | |
7673 @result{} "a piece of text" | |
7674 @end group | |
7675 @end smallexample | |
7676 | |
7677 The actual functions in Emacs are more complex than this, of course. | |
7678 The code for cutting and retrieving text has to be written so that | |
7679 Emacs can figure out which element in the list you want---the first, | |
7680 second, third, or whatever. In addition, when you get to the end of | |
7681 the list, Emacs should give you the first element of the list, rather | |
7682 than nothing at all. | |
7683 | |
7684 The list that holds the pieces of text is called the @dfn{kill ring}. | |
7685 This chapter leads up to a description of the kill ring and how it is | |
7686 used by first tracing how the @code{zap-to-char} function works. This | |
7687 function uses (or `calls') a function that invokes a function that | |
7688 manipulates the kill ring. Thus, before reaching the mountains, we | |
7689 climb the foothills. | |
7690 | |
7691 A subsequent chapter describes how text that is cut from the buffer is | |
7692 retrieved. @xref{Yanking, , Yanking Text Back}. | |
7693 | |
7694 @node zap-to-char, kill-region, Storing Text, Cutting & Storing Text | |
7695 @comment node-name, next, previous, up | |
7696 @section @code{zap-to-char} | |
7697 @findex zap-to-char | |
7698 | |
7699 The @code{zap-to-char} function changed little between GNU Emacs | |
7700 version 19 and GNU Emacs version 22. However, @code{zap-to-char} | |
7701 calls another function, @code{kill-region}, which enjoyed a major | |
7702 rewrite. | |
7703 | |
7704 The @code{kill-region} function in Emacs 19 is complex, but does not | |
7705 use code that is important at this time. We will skip it. | |
7706 | |
7707 The @code{kill-region} function in Emacs 22 is easier to read than the | |
7708 same function in Emacs 19 and introduces a very important concept, | |
7709 that of error handling. We will walk through the function. | |
7710 | |
7711 But first, let us look at the interactive @code{zap-to-char} function. | |
7712 | |
7713 @menu | |
7714 * Complete zap-to-char:: The complete implementation. | |
7715 * zap-to-char interactive:: A three part interactive expression. | |
7716 * zap-to-char body:: A short overview. | |
7717 * search-forward:: How to search for a string. | |
7718 * progn:: The @code{progn} special form. | |
7719 * Summing up zap-to-char:: Using @code{point} and @code{search-forward}. | |
7720 @end menu | |
7721 | |
7722 @node Complete zap-to-char, zap-to-char interactive, zap-to-char, zap-to-char | |
7723 @ifnottex | |
7724 @unnumberedsubsec The Complete @code{zap-to-char} Implementation | |
7725 @end ifnottex | |
7726 | |
7727 The @code{zap-to-char} function removes the text in the region between | |
7728 the location of the cursor (i.e., of point) up to and including the | |
7729 next occurrence of a specified character. The text that | |
7730 @code{zap-to-char} removes is put in the kill ring; and it can be | |
7731 retrieved from the kill ring by typing @kbd{C-y} (@code{yank}). If | |
7732 the command is given an argument, it removes text through that number | |
7733 of occurrences. Thus, if the cursor were at the beginning of this | |
7734 sentence and the character were @samp{s}, @samp{Thus} would be | |
7735 removed. If the argument were two, @samp{Thus, if the curs} would be | |
7736 removed, up to and including the @samp{s} in @samp{cursor}. | |
7737 | |
7738 If the specified character is not found, @code{zap-to-char} will say | |
7739 ``Search failed'', tell you the character you typed, and not remove | |
7740 any text. | |
7741 | |
7742 In order to determine how much text to remove, @code{zap-to-char} uses | |
7743 a search function. Searches are used extensively in code that | |
7744 manipulates text, and we will focus attention on them as well as on the | |
7745 deletion command. | |
7746 | |
7747 @ignore | |
7748 @c GNU Emacs version 19 | |
7749 (defun zap-to-char (arg char) ; version 19 implementation | |
7750 "Kill up to and including ARG'th occurrence of CHAR. | |
7751 Goes backward if ARG is negative; error if CHAR not found." | |
7752 (interactive "*p\ncZap to char: ") | |
7753 (kill-region (point) | |
7754 (progn | |
7755 (search-forward | |
7756 (char-to-string char) nil nil arg) | |
7757 (point)))) | |
7758 @end ignore | |
7759 | |
7760 @need 1250 | |
7761 Here is the complete text of the version 22 implementation of the function: | |
7762 | |
7763 @c GNU Emacs 22 | |
7764 @smallexample | |
7765 @group | |
7766 (defun zap-to-char (arg char) | |
7767 "Kill up to and including ARG'th occurrence of CHAR. | |
7768 Case is ignored if `case-fold-search' is non-nil in the current buffer. | |
7769 Goes backward if ARG is negative; error if CHAR not found." | |
7770 (interactive "p\ncZap to char: ") | |
7771 (if (char-table-p translation-table-for-input) | |
7772 (setq char (or (aref translation-table-for-input char) char))) | |
7773 (kill-region (point) (progn | |
7774 (search-forward (char-to-string char) nil nil arg) | |
7775 (point)))) | |
7776 @end group | |
7777 @end smallexample | |
7778 | |
7779 The documentation is thorough. You do need to know the jargon meaning | |
7780 of the word `kill'. | |
7781 | |
7782 @node zap-to-char interactive, zap-to-char body, Complete zap-to-char, zap-to-char | |
7783 @comment node-name, next, previous, up | |
7784 @subsection The @code{interactive} Expression | |
7785 | |
7786 @need 800 | |
7787 The interactive expression in the @code{zap-to-char} command looks like | |
7788 this: | |
7789 | |
7790 @smallexample | |
7791 (interactive "p\ncZap to char: ") | |
7792 @end smallexample | |
7793 | |
7794 The part within quotation marks, @code{"p\ncZap to char:@: "}, specifies | |
7795 two different things. First, and most simply, is the @samp{p}. | |
7796 This part is separated from the next part by a newline, @samp{\n}. | |
7797 The @samp{p} means that the first argument to the function will be | |
7798 passed the value of a `processed prefix'. The prefix argument is | |
7799 passed by typing @kbd{C-u} and a number, or @kbd{M-} and a number. If | |
7800 the function is called interactively without a prefix, 1 is passed to | |
7801 this argument. | |
7802 | |
7803 The second part of @code{"p\ncZap to char:@: "} is | |
7804 @samp{cZap to char:@: }. In this part, the lower case @samp{c} | |
7805 indicates that @code{interactive} expects a prompt and that the | |
7806 argument will be a character. The prompt follows the @samp{c} and is | |
7807 the string @samp{Zap to char:@: } (with a space after the colon to | |
7808 make it look good). | |
7809 | |
7810 What all this does is prepare the arguments to @code{zap-to-char} so they | |
7811 are of the right type, and give the user a prompt. | |
7812 | |
7813 In a read-only buffer, the @code{zap-to-char} function copies the text | |
7814 to the kill ring, but does not remove it. The echo area displays a | |
7815 message saying that the buffer is read-only. Also, the terminal may | |
7816 beep or blink at you. | |
7817 | |
7818 @node zap-to-char body, search-forward, zap-to-char interactive, zap-to-char | |
7819 @comment node-name, next, previous, up | |
7820 @subsection The Body of @code{zap-to-char} | |
7821 | |
7822 The body of the @code{zap-to-char} function contains the code that | |
7823 kills (that is, removes) the text in the region from the current | |
7824 position of the cursor up to and including the specified character. | |
7825 | |
7826 The first part of the code looks like this: | |
7827 | |
7828 @smallexample | |
7829 (if (char-table-p translation-table-for-input) | |
7830 (setq char (or (aref translation-table-for-input char) char))) | |
7831 (kill-region (point) (progn | |
7832 (search-forward (char-to-string char) nil nil arg) | |
7833 (point))) | |
7834 @end smallexample | |
7835 | |
7836 @noindent | |
7837 @code{char-table-p} is an hitherto unseen function. It determines | |
7838 whether its argument is a character table. When it is, it sets the | |
7839 character passed to @code{zap-to-char} to one of them, if that | |
7840 character exists, or to the character itself. (This becomes important | |
7841 for certain characters in non-European languages. The @code{aref} | |
7842 function extracts an element from an array. It is an array-specific | |
7843 function that is not described in this document. @xref{Arrays, , | |
7844 Arrays, elisp, The GNU Emacs Lisp Reference Manual}.) | |
7845 | |
7846 @noindent | |
7847 @code{(point)} is the current position of the cursor. | |
7848 | |
7849 The next part of the code is an expression using @code{progn}. The body | |
7850 of the @code{progn} consists of calls to @code{search-forward} and | |
7851 @code{point}. | |
7852 | |
7853 It is easier to understand how @code{progn} works after learning about | |
7854 @code{search-forward}, so we will look at @code{search-forward} and | |
7855 then at @code{progn}. | |
7856 | |
7857 @node search-forward, progn, zap-to-char body, zap-to-char | |
7858 @comment node-name, next, previous, up | |
7859 @subsection The @code{search-forward} Function | |
7860 @findex search-forward | |
7861 | |
7862 The @code{search-forward} function is used to locate the | |
7863 zapped-for-character in @code{zap-to-char}. If the search is | |
7864 successful, @code{search-forward} leaves point immediately after the | |
7865 last character in the target string. (In @code{zap-to-char}, the | |
7866 target string is just one character long. @code{zap-to-char} uses the | |
7867 function @code{char-to-string} to ensure that the computer treats that | |
7868 character as a string.) If the search is backwards, | |
7869 @code{search-forward} leaves point just before the first character in | |
7870 the target. Also, @code{search-forward} returns @code{t} for true. | |
7871 (Moving point is therefore a `side effect'.) | |
7872 | |
7873 @need 1250 | |
7874 In @code{zap-to-char}, the @code{search-forward} function looks like this: | |
7875 | |
7876 @smallexample | |
7877 (search-forward (char-to-string char) nil nil arg) | |
7878 @end smallexample | |
7879 | |
7880 The @code{search-forward} function takes four arguments: | |
7881 | |
7882 @enumerate | |
7883 @item | |
7884 The first argument is the target, what is searched for. This must be a | |
7885 string, such as @samp{"z"}. | |
7886 | |
7887 As it happens, the argument passed to @code{zap-to-char} is a single | |
7888 character. Because of the way computers are built, the Lisp | |
7889 interpreter may treat a single character as being different from a | |
7890 string of characters. Inside the computer, a single character has a | |
7891 different electronic format than a string of one character. (A single | |
7892 character can often be recorded in the computer using exactly one | |
7893 byte; but a string may be longer, and the computer needs to be ready | |
7894 for this.) Since the @code{search-forward} function searches for a | |
7895 string, the character that the @code{zap-to-char} function receives as | |
7896 its argument must be converted inside the computer from one format to | |
7897 the other; otherwise the @code{search-forward} function will fail. | |
7898 The @code{char-to-string} function is used to make this conversion. | |
7899 | |
7900 @item | |
7901 The second argument bounds the search; it is specified as a position in | |
7902 the buffer. In this case, the search can go to the end of the buffer, | |
7903 so no bound is set and the second argument is @code{nil}. | |
7904 | |
7905 @item | |
7906 The third argument tells the function what it should do if the search | |
7907 fails---it can signal an error (and print a message) or it can return | |
7908 @code{nil}. A @code{nil} as the third argument causes the function to | |
7909 signal an error when the search fails. | |
7910 | |
7911 @item | |
7912 The fourth argument to @code{search-forward} is the repeat count---how | |
7913 many occurrences of the string to look for. This argument is optional | |
7914 and if the function is called without a repeat count, this argument is | |
7915 passed the value 1. If this argument is negative, the search goes | |
7916 backwards. | |
7917 @end enumerate | |
7918 | |
7919 @need 800 | |
7920 In template form, a @code{search-forward} expression looks like this: | |
7921 | |
7922 @smallexample | |
7923 @group | |
7924 (search-forward "@var{target-string}" | |
7925 @var{limit-of-search} | |
7926 @var{what-to-do-if-search-fails} | |
7927 @var{repeat-count}) | |
7928 @end group | |
7929 @end smallexample | |
7930 | |
7931 We will look at @code{progn} next. | |
7932 | |
7933 @node progn, Summing up zap-to-char, search-forward, zap-to-char | |
7934 @comment node-name, next, previous, up | |
7935 @subsection The @code{progn} Special Form | |
7936 @findex progn | |
7937 | |
7938 @code{progn} is a special form that causes each of its arguments to be | |
7939 evaluated in sequence and then returns the value of the last one. The | |
7940 preceding expressions are evaluated only for the side effects they | |
7941 perform. The values produced by them are discarded. | |
7942 | |
7943 @need 800 | |
7944 The template for a @code{progn} expression is very simple: | |
7945 | |
7946 @smallexample | |
7947 @group | |
7948 (progn | |
7949 @var{body}@dots{}) | |
7950 @end group | |
7951 @end smallexample | |
7952 | |
7953 In @code{zap-to-char}, the @code{progn} expression has to do two things: | |
7954 put point in exactly the right position; and return the location of | |
7955 point so that @code{kill-region} will know how far to kill to. | |
7956 | |
7957 The first argument to the @code{progn} is @code{search-forward}. When | |
7958 @code{search-forward} finds the string, the function leaves point | |
7959 immediately after the last character in the target string. (In this | |
7960 case the target string is just one character long.) If the search is | |
7961 backwards, @code{search-forward} leaves point just before the first | |
7962 character in the target. The movement of point is a side effect. | |
7963 | |
7964 The second and last argument to @code{progn} is the expression | |
7965 @code{(point)}. This expression returns the value of point, which in | |
7966 this case will be the location to which it has been moved by | |
7967 @code{search-forward}. (In the source, a line that tells the function | |
7968 to go to the previous character, if it is going forward, was commented | |
7969 out in 1999; I don't remember whether that feature or mis-feature was | |
7970 ever a part of the distributed source.) The value of @code{point} is | |
7971 returned by the @code{progn} expression and is passed to | |
7972 @code{kill-region} as @code{kill-region}'s second argument. | |
7973 | |
7974 @node Summing up zap-to-char, , progn, zap-to-char | |
7975 @comment node-name, next, previous, up | |
7976 @subsection Summing up @code{zap-to-char} | |
7977 | |
7978 Now that we have seen how @code{search-forward} and @code{progn} work, | |
7979 we can see how the @code{zap-to-char} function works as a whole. | |
7980 | |
7981 The first argument to @code{kill-region} is the position of the cursor | |
7982 when the @code{zap-to-char} command is given---the value of point at | |
7983 that time. Within the @code{progn}, the search function then moves | |
7984 point to just after the zapped-to-character and @code{point} returns the | |
7985 value of this location. The @code{kill-region} function puts together | |
7986 these two values of point, the first one as the beginning of the region | |
7987 and the second one as the end of the region, and removes the region. | |
7988 | |
7989 The @code{progn} special form is necessary because the | |
7990 @code{kill-region} command takes two arguments; and it would fail if | |
7991 @code{search-forward} and @code{point} expressions were written in | |
7992 sequence as two additional arguments. The @code{progn} expression is | |
7993 a single argument to @code{kill-region} and returns the one value that | |
7994 @code{kill-region} needs for its second argument. | |
7995 | |
7996 @node kill-region, copy-region-as-kill, zap-to-char, Cutting & Storing Text | |
7997 @comment node-name, next, previous, up | |
7998 @section @code{kill-region} | |
7999 @findex kill-region | |
8000 | |
8001 The @code{zap-to-char} function uses the @code{kill-region} function. | |
8002 This function clips text from a region and copies that text to | |
8003 the kill ring, from which it may be retrieved. | |
8004 | |
8005 @ignore | |
8006 GNU Emacs 22: | |
8007 | |
8008 (defun kill-region (beg end &optional yank-handler) | |
8009 "Kill (\"cut\") text between point and mark. | |
8010 This deletes the text from the buffer and saves it in the kill ring. | |
8011 The command \\[yank] can retrieve it from there. | |
8012 \(If you want to kill and then yank immediately, use \\[kill-ring-save].) | |
8013 | |
8014 If you want to append the killed region to the last killed text, | |
8015 use \\[append-next-kill] before \\[kill-region]. | |
8016 | |
8017 If the buffer is read-only, Emacs will beep and refrain from deleting | |
8018 the text, but put the text in the kill ring anyway. This means that | |
8019 you can use the killing commands to copy text from a read-only buffer. | |
8020 | |
8021 This is the primitive for programs to kill text (as opposed to deleting it). | |
8022 Supply two arguments, character positions indicating the stretch of text | |
8023 to be killed. | |
8024 Any command that calls this function is a \"kill command\". | |
8025 If the previous command was also a kill command, | |
8026 the text killed this time appends to the text killed last time | |
8027 to make one entry in the kill ring. | |
8028 | |
8029 In Lisp code, optional third arg YANK-HANDLER, if non-nil, | |
8030 specifies the yank-handler text property to be set on the killed | |
8031 text. See `insert-for-yank'." | |
8032 ;; Pass point first, then mark, because the order matters | |
8033 ;; when calling kill-append. | |
8034 (interactive (list (point) (mark))) | |
8035 (unless (and beg end) | |
8036 (error "The mark is not set now, so there is no region")) | |
8037 (condition-case nil | |
8038 (let ((string (filter-buffer-substring beg end t))) | |
8039 (when string ;STRING is nil if BEG = END | |
8040 ;; Add that string to the kill ring, one way or another. | |
8041 (if (eq last-command 'kill-region) | |
8042 (kill-append string (< end beg) yank-handler) | |
8043 (kill-new string nil yank-handler))) | |
8044 (when (or string (eq last-command 'kill-region)) | |
8045 (setq this-command 'kill-region)) | |
8046 nil) | |
8047 ((buffer-read-only text-read-only) | |
8048 ;; The code above failed because the buffer, or some of the characters | |
8049 ;; in the region, are read-only. | |
8050 ;; We should beep, in case the user just isn't aware of this. | |
8051 ;; However, there's no harm in putting | |
8052 ;; the region's text in the kill ring, anyway. | |
8053 (copy-region-as-kill beg end) | |
8054 ;; Set this-command now, so it will be set even if we get an error. | |
8055 (setq this-command 'kill-region) | |
8056 ;; This should barf, if appropriate, and give us the correct error. | |
8057 (if kill-read-only-ok | |
8058 (progn (message "Read only text copied to kill ring") nil) | |
8059 ;; Signal an error if the buffer is read-only. | |
8060 (barf-if-buffer-read-only) | |
8061 ;; If the buffer isn't read-only, the text is. | |
8062 (signal 'text-read-only (list (current-buffer))))))) | |
8063 @end ignore | |
8064 | |
8065 The Emacs 22 version of that function uses @code{condition-case} and | |
8066 @code{copy-region-as-kill}, both of which we will explain. | |
8067 @code{condition-case} is an important special form. | |
8068 | |
8069 In essence, the @code{kill-region} function calls | |
8070 @code{condition-case}, which takes three arguments. In this function, | |
8071 the first argument does nothing. The second argument contains the | |
8072 code that does the work when all goes well. The third argument | |
8073 contains the code that is called in the event of an error. | |
8074 | |
8075 @menu | |
8076 * Complete kill-region:: The function definition. | |
8077 * condition-case:: Dealing with a problem. | |
8078 * Lisp macro:: | |
8079 @end menu | |
8080 | |
8081 @node Complete kill-region, condition-case, kill-region, kill-region | |
8082 @ifnottex | |
8083 @unnumberedsubsec The Complete @code{kill-region} Definition | |
8084 @end ifnottex | |
8085 | |
8086 @need 1200 | |
8087 We will go through the @code{condition-case} code in a moment. First, | |
8088 let us look at the definition of @code{kill-region}, with comments | |
8089 added: | |
8090 | |
8091 @c GNU Emacs 22: | |
8092 @smallexample | |
8093 @group | |
8094 (defun kill-region (beg end) | |
8095 "Kill (\"cut\") text between point and mark. | |
8096 This deletes the text from the buffer and saves it in the kill ring. | |
8097 The command \\[yank] can retrieve it from there. @dots{} " | |
8098 @end group | |
8099 | |
8100 @group | |
8101 ;; @bullet{} Since order matters, pass point first. | |
8102 (interactive (list (point) (mark))) | |
8103 ;; @bullet{} And tell us if we cannot cut the text. | |
8104 ;; `unless' is an `if' without a then-part. | |
8105 (unless (and beg end) | |
8106 (error "The mark is not set now, so there is no region")) | |
8107 @end group | |
8108 | |
8109 @group | |
8110 ;; @bullet{} `condition-case' takes three arguments. | |
8111 ;; If the first argument is nil, as it is here, | |
8112 ;; information about the error signal is not | |
8113 ;; stored for use by another function. | |
8114 (condition-case nil | |
8115 @end group | |
8116 | |
8117 @group | |
8118 ;; @bullet{} The second argument to `condition-case' tells the | |
8119 ;; Lisp interpreter what to do when all goes well. | |
8120 @end group | |
8121 | |
8122 @group | |
8123 ;; It starts with a `let' function that extracts the string | |
8124 ;; and tests whether it exists. If so (that is what the | |
8125 ;; `when' checks), it calls an `if' function that determines | |
8126 ;; whether the previous command was another call to | |
8127 ;; `kill-region'; if it was, then the new text is appended to | |
8128 ;; the previous text; if not, then a different function, | |
8129 ;; `kill-new', is called. | |
8130 @end group | |
8131 | |
8132 @group | |
8133 ;; The `kill-append' function concatenates the new string and | |
8134 ;; the old. The `kill-new' function inserts text into a new | |
8135 ;; item in the kill ring. | |
8136 @end group | |
8137 | |
8138 @group | |
8139 ;; `when' is an `if' without an else-part. The second `when' | |
8140 ;; again checks whether the current string exists; in | |
8141 ;; addition, it checks whether the previous command was | |
8142 ;; another call to `kill-region'. If one or the other | |
8143 ;; condition is true, then it sets the current command to | |
8144 ;; be `kill-region'. | |
8145 @end group | |
8146 @group | |
8147 (let ((string (filter-buffer-substring beg end t))) | |
8148 (when string ;STRING is nil if BEG = END | |
8149 ;; Add that string to the kill ring, one way or another. | |
8150 (if (eq last-command 'kill-region) | |
8151 @end group | |
8152 @group | |
8153 ;; @minus{} `yank-handler' is an optional argument to | |
8154 ;; `kill-region' that tells the `kill-append' and | |
8155 ;; `kill-new' functions how deal with properties | |
8156 ;; added to the text, such as `bold' or `italics'. | |
8157 (kill-append string (< end beg) yank-handler) | |
8158 (kill-new string nil yank-handler))) | |
8159 (when (or string (eq last-command 'kill-region)) | |
8160 (setq this-command 'kill-region)) | |
8161 nil) | |
8162 @end group | |
8163 | |
8164 @group | |
8165 ;; @bullet{} The third argument to `condition-case' tells the interpreter | |
8166 ;; what to do with an error. | |
8167 @end group | |
8168 @group | |
8169 ;; The third argument has a conditions part and a body part. | |
8170 ;; If the conditions are met (in this case, | |
8171 ;; if text or buffer are read-only) | |
8172 ;; then the body is executed. | |
8173 @end group | |
8174 @group | |
8175 ;; The first part of the third argument is the following: | |
8176 ((buffer-read-only text-read-only) ;; the if-part | |
8177 ;; @dots{} the then-part | |
8178 (copy-region-as-kill beg end) | |
8179 @end group | |
8180 @group | |
8181 ;; Next, also as part of the then-part, set this-command, so | |
8182 ;; it will be set in an error | |
8183 (setq this-command 'kill-region) | |
8184 ;; Finally, in the then-part, send a message if you may copy | |
8185 ;; the text to the kill ring without signally an error, but | |
8186 ;; don't if you may not. | |
8187 @end group | |
8188 @group | |
8189 (if kill-read-only-ok | |
8190 (progn (message "Read only text copied to kill ring") nil) | |
8191 (barf-if-buffer-read-only) | |
8192 ;; If the buffer isn't read-only, the text is. | |
8193 (signal 'text-read-only (list (current-buffer))))) | |
8194 @end group | |
8195 @end smallexample | |
8196 | |
8197 @ignore | |
8198 @c v 21 | |
8199 @smallexample | |
8200 @group | |
8201 (defun kill-region (beg end) | |
8202 "Kill between point and mark. | |
8203 The text is deleted but saved in the kill ring." | |
8204 (interactive "r") | |
8205 @end group | |
8206 | |
8207 @group | |
8208 ;; 1. `condition-case' takes three arguments. | |
8209 ;; If the first argument is nil, as it is here, | |
8210 ;; information about the error signal is not | |
8211 ;; stored for use by another function. | |
8212 (condition-case nil | |
8213 @end group | |
8214 | |
8215 @group | |
8216 ;; 2. The second argument to `condition-case' | |
8217 ;; tells the Lisp interpreter what to do when all goes well. | |
8218 @end group | |
8219 | |
8220 @group | |
8221 ;; The `delete-and-extract-region' function usually does the | |
8222 ;; work. If the beginning and ending of the region are both | |
8223 ;; the same, then the variable `string' will be empty, or nil | |
8224 (let ((string (delete-and-extract-region beg end))) | |
8225 @end group | |
8226 | |
8227 @group | |
8228 ;; `when' is an `if' clause that cannot take an `else-part'. | |
8229 ;; Emacs normally sets the value of `last-command' to the | |
8230 ;; previous command. | |
8231 @end group | |
8232 @group | |
8233 ;; `kill-append' concatenates the new string and the old. | |
8234 ;; `kill-new' inserts text into a new item in the kill ring. | |
8235 (when string | |
8236 (if (eq last-command 'kill-region) | |
8237 ;; if true, prepend string | |
8238 (kill-append string (< end beg)) | |
8239 (kill-new string))) | |
8240 (setq this-command 'kill-region)) | |
8241 @end group | |
8242 | |
8243 @group | |
8244 ;; 3. The third argument to `condition-case' tells the interpreter | |
8245 ;; what to do with an error. | |
8246 @end group | |
8247 @group | |
8248 ;; The third argument has a conditions part and a body part. | |
8249 ;; If the conditions are met (in this case, | |
8250 ;; if text or buffer are read-only) | |
8251 ;; then the body is executed. | |
8252 @end group | |
8253 @group | |
8254 ((buffer-read-only text-read-only) ;; this is the if-part | |
8255 ;; then... | |
8256 (copy-region-as-kill beg end) | |
8257 @end group | |
8258 @group | |
8259 (if kill-read-only-ok ;; usually this variable is nil | |
8260 (message "Read only text copied to kill ring") | |
8261 ;; or else, signal an error if the buffer is read-only; | |
8262 (barf-if-buffer-read-only) | |
8263 ;; and, in any case, signal that the text is read-only. | |
8264 (signal 'text-read-only (list (current-buffer))))))) | |
8265 @end group | |
8266 @end smallexample | |
8267 @end ignore | |
8268 | |
8269 @node condition-case, Lisp macro, Complete kill-region, kill-region | |
8270 @comment node-name, next, previous, up | |
8271 @subsection @code{condition-case} | |
8272 @findex condition-case | |
8273 | |
8274 As we have seen earlier (@pxref{Making Errors, , Generate an Error | |
8275 Message}), when the Emacs Lisp interpreter has trouble evaluating an | |
8276 expression, it provides you with help; in the jargon, this is called | |
8277 ``signaling an error''. Usually, the computer stops the program and | |
8278 shows you a message. | |
8279 | |
8280 However, some programs undertake complicated actions. They should not | |
8281 simply stop on an error. In the @code{kill-region} function, the most | |
8282 likely error is that you will try to kill text that is read-only and | |
8283 cannot be removed. So the @code{kill-region} function contains code | |
8284 to handle this circumstance. This code, which makes up the body of | |
8285 the @code{kill-region} function, is inside of a @code{condition-case} | |
8286 special form. | |
8287 | |
8288 @need 800 | |
8289 The template for @code{condition-case} looks like this: | |
8290 | |
8291 @smallexample | |
8292 @group | |
8293 (condition-case | |
8294 @var{var} | |
8295 @var{bodyform} | |
8296 @var{error-handler}@dots{}) | |
8297 @end group | |
8298 @end smallexample | |
8299 | |
8300 The second argument, @var{bodyform}, is straightforward. The | |
8301 @code{condition-case} special form causes the Lisp interpreter to | |
8302 evaluate the code in @var{bodyform}. If no error occurs, the special | |
8303 form returns the code's value and produces the side-effects, if any. | |
8304 | |
8305 In short, the @var{bodyform} part of a @code{condition-case} | |
8306 expression determines what should happen when everything works | |
8307 correctly. | |
8308 | |
8309 However, if an error occurs, among its other actions, the function | |
8310 generating the error signal will define one or more error condition | |
8311 names. | |
8312 | |
8313 An error handler is the third argument to @code{condition case}. | |
8314 An error handler has two parts, a @var{condition-name} and a | |
8315 @var{body}. If the @var{condition-name} part of an error handler | |
8316 matches a condition name generated by an error, then the @var{body} | |
8317 part of the error handler is run. | |
8318 | |
8319 As you will expect, the @var{condition-name} part of an error handler | |
8320 may be either a single condition name or a list of condition names. | |
8321 | |
8322 Also, a complete @code{condition-case} expression may contain more | |
8323 than one error handler. When an error occurs, the first applicable | |
8324 handler is run. | |
8325 | |
8326 Lastly, the first argument to the @code{condition-case} expression, | |
8327 the @var{var} argument, is sometimes bound to a variable that | |
8328 contains information about the error. However, if that argument is | |
8329 nil, as is the case in @code{kill-region}, that information is | |
8330 discarded. | |
8331 | |
8332 @need 1200 | |
8333 In brief, in the @code{kill-region} function, the code | |
8334 @code{condition-case} works like this: | |
8335 | |
8336 @smallexample | |
8337 @group | |
8338 @var{If no errors}, @var{run only this code} | |
8339 @var{but}, @var{if errors}, @var{run this other code}. | |
8340 @end group | |
8341 @end smallexample | |
8342 | |
8343 @ignore | |
8344 2006 Oct 24 | |
8345 In Emacs 22, | |
8346 copy-region-as-kill is short, 12 lines, and uses | |
8347 filter-buffer-substring, which is longer, 39 lines | |
8348 and has delete-and-extract-region in it. | |
8349 delete-and-extract-region is written in C. | |
8350 | |
8351 see Initializing a Variable with @code{defvar} | |
8352 this is line 8054 | |
8353 Initializing a Variable with @code{defvar} includes line 8350 | |
8354 @end ignore | |
8355 | |
8356 @node Lisp macro, , condition-case, kill-region | |
8357 @comment node-name, next, previous, up | |
8358 @subsection Lisp macro | |
8359 @cindex Macro, lisp | |
8360 @cindex Lisp macro | |
8361 | |
8362 The part of the @code{condition-case} expression that is evaluated in | |
8363 the expectation that all goes well has a @code{when}. The code uses | |
8364 @code{when} to determine whether the @code{string} variable points to | |
8365 text that exists. | |
8366 | |
8367 A @code{when} expression is simply a programmers' convenience. It is | |
8368 an @code{if} without the possibility of an else clause. In your mind, | |
8369 you can replace @code{when} with @code{if} and understand what goes | |
8370 on. That is what the Lisp interpreter does. | |
8371 | |
8372 Technically speaking, @code{when} is a Lisp macro. A Lisp @dfn{macro} | |
8373 enables you to define new control constructs and other language | |
8374 features. It tells the interpreter how to compute another Lisp | |
8375 expression which will in turn compute the value. In this case, the | |
8376 `other expression' is an @code{if} expression. | |
8377 | |
8378 The @code{kill-region} function definition also has an @code{unless} | |
8379 macro; it is the converse of @code{when}. The @code{unless} macro is | |
8380 an @code{if} without a then clause | |
8381 | |
8382 For more about Lisp macros, see @ref{Macros, , Macros, elisp, The GNU | |
8383 Emacs Lisp Reference Manual}. The C programming language also | |
8384 provides macros. These are different, but also useful. | |
8385 | |
8386 @ignore | |
8387 We will briefly look at C macros in | |
8388 @ref{Digression into C}. | |
8389 @end ignore | |
8390 | |
8391 @need 1200 | |
8392 Regarding the @code{when} macro, in the @code{condition-case} | |
8393 expression, when the string has content, then another conditional | |
8394 expression is executed. This is an @code{if} with both a then-part | |
8395 and an else-part. | |
8396 | |
8397 @smallexample | |
8398 @group | |
8399 (if (eq last-command 'kill-region) | |
8400 (kill-append string (< end beg) yank-handler) | |
8401 (kill-new string nil yank-handler)) | |
8402 @end group | |
8403 @end smallexample | |
8404 | |
8405 The then-part is evaluated if the previous command was another call to | |
8406 @code{kill-region}; if not, the else-part is evaluated. | |
8407 | |
8408 @code{yank-handler} is an optional argument to @code{kill-region} that | |
8409 tells the @code{kill-append} and @code{kill-new} functions how deal | |
8410 with properties added to the text, such as `bold' or `italics'. | |
8411 | |
8412 @code{last-command} is a variable that comes with Emacs that we have | |
8413 not seen before. Normally, whenever a function is executed, Emacs | |
8414 sets the value of @code{last-command} to the previous command. | |
8415 | |
8416 @need 1200 | |
8417 In this segment of the definition, the @code{if} expression checks | |
8418 whether the previous command was @code{kill-region}. If it was, | |
8419 | |
8420 @smallexample | |
8421 (kill-append string (< end beg) yank-handler) | |
8422 @end smallexample | |
8423 | |
8424 @noindent | |
8425 concatenates a copy of the newly clipped text to the just previously | |
8426 clipped text in the kill ring. | |
8427 | |
8428 @node copy-region-as-kill, Digression into C, kill-region, Cutting & Storing Text | |
8429 @comment node-name, next, previous, up | |
8430 @section @code{copy-region-as-kill} | |
8431 @findex copy-region-as-kill | |
8432 @findex nthcdr | |
8433 | |
8434 The @code{copy-region-as-kill} function copies a region of text from a | |
8435 buffer and (via either @code{kill-append} or @code{kill-new}) saves it | |
8436 in the @code{kill-ring}. | |
8437 | |
8438 If you call @code{copy-region-as-kill} immediately after a | |
8439 @code{kill-region} command, Emacs appends the newly copied text to the | |
8440 previously copied text. This means that if you yank back the text, you | |
8441 get it all, from both this and the previous operation. On the other | |
8442 hand, if some other command precedes the @code{copy-region-as-kill}, | |
8443 the function copies the text into a separate entry in the kill ring. | |
8444 | |
8445 @menu | |
8446 * Complete copy-region-as-kill:: The complete function definition. | |
8447 * copy-region-as-kill body:: The body of @code{copy-region-as-kill}. | |
8448 @end menu | |
8449 | |
8450 @node Complete copy-region-as-kill, copy-region-as-kill body, copy-region-as-kill, copy-region-as-kill | |
8451 @ifnottex | |
8452 @unnumberedsubsec The complete @code{copy-region-as-kill} function definition | |
8453 @end ifnottex | |
8454 | |
8455 @need 1200 | |
8456 Here is the complete text of the version 22 @code{copy-region-as-kill} | |
8457 function: | |
8458 | |
8459 @smallexample | |
8460 @group | |
8461 (defun copy-region-as-kill (beg end) | |
8462 "Save the region as if killed, but don't kill it. | |
8463 In Transient Mark mode, deactivate the mark. | |
8464 If `interprogram-cut-function' is non-nil, also save the text for a window | |
8465 system cut and paste." | |
8466 (interactive "r") | |
8467 @end group | |
8468 @group | |
8469 (if (eq last-command 'kill-region) | |
8470 (kill-append (filter-buffer-substring beg end) (< end beg)) | |
8471 (kill-new (filter-buffer-substring beg end))) | |
8472 @end group | |
8473 @group | |
8474 (if transient-mark-mode | |
8475 (setq deactivate-mark t)) | |
8476 nil) | |
8477 @end group | |
8478 @end smallexample | |
8479 | |
8480 @need 800 | |
8481 As usual, this function can be divided into its component parts: | |
8482 | |
8483 @smallexample | |
8484 @group | |
8485 (defun copy-region-as-kill (@var{argument-list}) | |
8486 "@var{documentation}@dots{}" | |
8487 (interactive "r") | |
8488 @var{body}@dots{}) | |
8489 @end group | |
8490 @end smallexample | |
8491 | |
8492 The arguments are @code{beg} and @code{end} and the function is | |
8493 interactive with @code{"r"}, so the two arguments must refer to the | |
8494 beginning and end of the region. If you have been reading though this | |
8495 document from the beginning, understanding these parts of a function is | |
8496 almost becoming routine. | |
8497 | |
8498 The documentation is somewhat confusing unless you remember that the | |
8499 word `kill' has a meaning different from usual. The `Transient Mark' | |
8500 and @code{interprogram-cut-function} comments explain certain | |
8501 side-effects. | |
8502 | |
8503 After you once set a mark, a buffer always contains a region. If you | |
8504 wish, you can use Transient Mark mode to highlight the region | |
8505 temporarily. (No one wants to highlight the region all the time, so | |
8506 Transient Mark mode highlights it only at appropriate times. Many | |
8507 people turn off Transient Mark mode, so the region is never | |
8508 highlighted.) | |
8509 | |
8510 Also, a windowing system allows you to copy, cut, and paste among | |
8511 different programs. In the X windowing system, for example, the | |
8512 @code{interprogram-cut-function} function is @code{x-select-text}, | |
8513 which works with the windowing system's equivalent of the Emacs kill | |
8514 ring. | |
8515 | |
8516 The body of the @code{copy-region-as-kill} function starts with an | |
8517 @code{if} clause. What this clause does is distinguish between two | |
8518 different situations: whether or not this command is executed | |
8519 immediately after a previous @code{kill-region} command. In the first | |
8520 case, the new region is appended to the previously copied text. | |
8521 Otherwise, it is inserted into the beginning of the kill ring as a | |
8522 separate piece of text from the previous piece. | |
8523 | |
8524 The last two lines of the function prevent the region from lighting up | |
8525 if Transient Mark mode is turned on. | |
8526 | |
8527 The body of @code{copy-region-as-kill} merits discussion in detail. | |
8528 | |
8529 @node copy-region-as-kill body, , Complete copy-region-as-kill, copy-region-as-kill | |
8530 @comment node-name, next, previous, up | |
8531 @subsection The Body of @code{copy-region-as-kill} | |
8532 | |
8533 The @code{copy-region-as-kill} function works in much the same way as | |
8534 the @code{kill-region} function. Both are written so that two or more | |
8535 kills in a row combine their text into a single entry. If you yank | |
8536 back the text from the kill ring, you get it all in one piece. | |
8537 Moreover, kills that kill forward from the current position of the | |
8538 cursor are added to the end of the previously copied text and commands | |
8539 that copy text backwards add it to the beginning of the previously | |
8540 copied text. This way, the words in the text stay in the proper | |
8541 order. | |
8542 | |
8543 Like @code{kill-region}, the @code{copy-region-as-kill} function makes | |
8544 use of the @code{last-command} variable that keeps track of the | |
8545 previous Emacs command. | |
8546 | |
8547 @menu | |
8548 * last-command & this-command:: | |
8549 * kill-append function:: | |
8550 * kill-new function:: | |
8551 @end menu | |
8552 | |
8553 @node last-command & this-command, kill-append function, copy-region-as-kill body, copy-region-as-kill body | |
8554 @ifnottex | |
8555 @unnumberedsubsubsec @code{last-command} and @code{this-command} | |
8556 @end ifnottex | |
8557 | |
8558 Normally, whenever a function is executed, Emacs sets the value of | |
8559 @code{this-command} to the function being executed (which in this case | |
8560 would be @code{copy-region-as-kill}). At the same time, Emacs sets | |
8561 the value of @code{last-command} to the previous value of | |
8562 @code{this-command}. | |
8563 | |
8564 In the first part of the body of the @code{copy-region-as-kill} | |
8565 function, an @code{if} expression determines whether the value of | |
8566 @code{last-command} is @code{kill-region}. If so, the then-part of | |
8567 the @code{if} expression is evaluated; it uses the @code{kill-append} | |
8568 function to concatenate the text copied at this call to the function | |
8569 with the text already in the first element (the @sc{car}) of the kill | |
8570 ring. On the other hand, if the value of @code{last-command} is not | |
8571 @code{kill-region}, then the @code{copy-region-as-kill} function | |
8572 attaches a new element to the kill ring using the @code{kill-new} | |
8573 function. | |
8574 | |
8575 @need 1250 | |
8576 The @code{if} expression reads as follows; it uses @code{eq}: | |
8577 | |
8578 @smallexample | |
8579 @group | |
8580 (if (eq last-command 'kill-region) | |
8581 ;; @r{then-part} | |
8582 (kill-append (filter-buffer-substring beg end) (< end beg)) | |
8583 ;; @r{else-part} | |
8584 (kill-new (filter-buffer-substring beg end))) | |
8585 @end group | |
8586 @end smallexample | |
8587 | |
8588 @findex filter-buffer-substring | |
8589 (The @code{filter-buffer-substring} function returns a filtered | |
8590 substring of the buffer, if any. Optionally---the arguments are not | |
8591 here, so neither is done---the function may delete the initial text or | |
8592 return the text without its properties; this function is a replacement | |
8593 for the older @code{buffer-substring} function, which came before text | |
8594 properties were implemented.) | |
8595 | |
8596 @findex eq @r{(example of use)} | |
8597 @noindent | |
8598 The @code{eq} function tests whether its first argument is the same Lisp | |
8599 object as its second argument. The @code{eq} function is similar to the | |
8600 @code{equal} function in that it is used to test for equality, but | |
8601 differs in that it determines whether two representations are actually | |
8602 the same object inside the computer, but with different names. | |
8603 @code{equal} determines whether the structure and contents of two | |
8604 expressions are the same. | |
8605 | |
8606 If the previous command was @code{kill-region}, then the Emacs Lisp | |
8607 interpreter calls the @code{kill-append} function | |
8608 | |
8609 @node kill-append function, kill-new function, last-command & this-command, copy-region-as-kill body | |
8610 @unnumberedsubsubsec The @code{kill-append} function | |
8611 @findex kill-append | |
8612 | |
8613 @need 800 | |
8614 The @code{kill-append} function looks like this: | |
8615 | |
8616 @c in GNU Emacs 22 | |
8617 @smallexample | |
8618 @group | |
8619 (defun kill-append (string before-p &optional yank-handler) | |
8620 "Append STRING to the end of the latest kill in the kill ring. | |
8621 If BEFORE-P is non-nil, prepend STRING to the kill. | |
8622 @dots{} " | |
8623 (let* ((cur (car kill-ring))) | |
8624 (kill-new (if before-p (concat string cur) (concat cur string)) | |
8625 (or (= (length cur) 0) | |
8626 (equal yank-handler | |
8627 (get-text-property 0 'yank-handler cur))) | |
8628 yank-handler))) | |
8629 @end group | |
8630 @end smallexample | |
8631 | |
8632 @ignore | |
8633 was: | |
8634 (defun kill-append (string before-p) | |
8635 "Append STRING to the end of the latest kill in the kill ring. | |
8636 If BEFORE-P is non-nil, prepend STRING to the kill. | |
8637 If `interprogram-cut-function' is set, pass the resulting kill to | |
8638 it." | |
8639 (kill-new (if before-p | |
8640 (concat string (car kill-ring)) | |
8641 (concat (car kill-ring) string)) | |
8642 t)) | |
8643 @end ignore | |
8644 | |
8645 @noindent | |
8646 The @code{kill-append} function is fairly straightforward. It uses | |
8647 the @code{kill-new} function, which we will discuss in more detail in | |
8648 a moment. | |
8649 | |
8650 (Also, the function provides an optional argument called | |
8651 @code{yank-handler}; when invoked, this argument tells the function | |
8652 how to deal with properties added to the text, such as `bold' or | |
8653 `italics'.) | |
8654 | |
8655 @c !!! bug in GNU Emacs 22 version of kill-append ? | |
8656 It has a @code{let*} function to set the value of the first element of | |
8657 the kill ring to @code{cur}. (I do not know why the function does not | |
8658 use @code{let} instead; only one value is set in the expression. | |
8659 Perhaps this is a bug that produces no problems?) | |
8660 | |
8661 Consider the conditional that is one of the two arguments to | |
8662 @code{kill-new}. It uses @code{concat} to concatenate the new text to | |
8663 the @sc{car} of the kill ring. Whether it prepends or appends the | |
8664 text depends on the results of an @code{if} expression: | |
8665 | |
8666 @smallexample | |
8667 @group | |
8668 (if before-p ; @r{if-part} | |
8669 (concat string cur) ; @r{then-part} | |
8670 (concat cur string)) ; @r{else-part} | |
8671 @end group | |
8672 @end smallexample | |
8673 | |
8674 @noindent | |
8675 If the region being killed is before the region that was killed in the | |
8676 last command, then it should be prepended before the material that was | |
8677 saved in the previous kill; and conversely, if the killed text follows | |
8678 what was just killed, it should be appended after the previous text. | |
8679 The @code{if} expression depends on the predicate @code{before-p} to | |
8680 decide whether the newly saved text should be put before or after the | |
8681 previously saved text. | |
8682 | |
8683 The symbol @code{before-p} is the name of one of the arguments to | |
8684 @code{kill-append}. When the @code{kill-append} function is | |
8685 evaluated, it is bound to the value returned by evaluating the actual | |
8686 argument. In this case, this is the expression @code{(< end beg)}. | |
8687 This expression does not directly determine whether the killed text in | |
8688 this command is located before or after the kill text of the last | |
8689 command; what it does is determine whether the value of the variable | |
8690 @code{end} is less than the value of the variable @code{beg}. If it | |
8691 is, it means that the user is most likely heading towards the | |
8692 beginning of the buffer. Also, the result of evaluating the predicate | |
8693 expression, @code{(< end beg)}, will be true and the text will be | |
8694 prepended before the previous text. On the other hand, if the value of | |
8695 the variable @code{end} is greater than the value of the variable | |
8696 @code{beg}, the text will be appended after the previous text. | |
8697 | |
8698 @need 800 | |
8699 When the newly saved text will be prepended, then the string with the new | |
8700 text will be concatenated before the old text: | |
8701 | |
8702 @smallexample | |
8703 (concat string cur) | |
8704 @end smallexample | |
8705 | |
8706 @need 1200 | |
8707 @noindent | |
8708 But if the text will be appended, it will be concatenated | |
8709 after the old text: | |
8710 | |
8711 @smallexample | |
8712 (concat cur string)) | |
8713 @end smallexample | |
8714 | |
8715 To understand how this works, we first need to review the | |
8716 @code{concat} function. The @code{concat} function links together or | |
8717 unites two strings of text. The result is a string. For example: | |
8718 | |
8719 @smallexample | |
8720 @group | |
8721 (concat "abc" "def") | |
8722 @result{} "abcdef" | |
8723 @end group | |
8724 | |
8725 @group | |
8726 (concat "new " | |
8727 (car '("first element" "second element"))) | |
8728 @result{} "new first element" | |
8729 | |
8730 (concat (car | |
8731 '("first element" "second element")) " modified") | |
8732 @result{} "first element modified" | |
8733 @end group | |
8734 @end smallexample | |
8735 | |
8736 We can now make sense of @code{kill-append}: it modifies the contents | |
8737 of the kill ring. The kill ring is a list, each element of which is | |
8738 saved text. The @code{kill-append} function uses the @code{kill-new} | |
8739 function which in turn uses the @code{setcar} function. | |
8740 | |
8741 @node kill-new function, , kill-append function, copy-region-as-kill body | |
8742 @unnumberedsubsubsec The @code{kill-new} function | |
8743 @findex kill-new | |
8744 | |
8745 @c in GNU Emacs 22, additional documentation to kill-new: | |
8746 @ignore | |
8747 Optional third arguments YANK-HANDLER controls how the STRING is later | |
8748 inserted into a buffer; see `insert-for-yank' for details. | |
8749 When a yank handler is specified, STRING must be non-empty (the yank | |
8750 handler, if non-nil, is stored as a `yank-handler' text property on STRING). | |
8751 | |
8752 When the yank handler has a non-nil PARAM element, the original STRING | |
8753 argument is not used by `insert-for-yank'. However, since Lisp code | |
8754 may access and use elements from the kill ring directly, the STRING | |
8755 argument should still be a \"useful\" string for such uses." | |
8756 @end ignore | |
8757 @need 1200 | |
8758 The @code{kill-new} function looks like this: | |
8759 | |
8760 @smallexample | |
8761 @group | |
8762 (defun kill-new (string &optional replace yank-handler) | |
8763 "Make STRING the latest kill in the kill ring. | |
8764 Set `kill-ring-yank-pointer' to point to it. | |
8765 | |
8766 If `interprogram-cut-function' is non-nil, apply it to STRING. | |
8767 Optional second argument REPLACE non-nil means that STRING will replace | |
8768 the front of the kill ring, rather than being added to the list. | |
8769 @dots{}" | |
8770 @end group | |
8771 @group | |
8772 (if (> (length string) 0) | |
8773 (if yank-handler | |
8774 (put-text-property 0 (length string) | |
8775 'yank-handler yank-handler string)) | |
8776 (if yank-handler | |
8777 (signal 'args-out-of-range | |
8778 (list string "yank-handler specified for empty string")))) | |
8779 @end group | |
8780 @group | |
8781 (if (fboundp 'menu-bar-update-yank-menu) | |
8782 (menu-bar-update-yank-menu string (and replace (car kill-ring)))) | |
8783 @end group | |
8784 @group | |
8785 (if (and replace kill-ring) | |
8786 (setcar kill-ring string) | |
8787 (push string kill-ring) | |
8788 (if (> (length kill-ring) kill-ring-max) | |
8789 (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil))) | |
8790 @end group | |
8791 @group | |
8792 (setq kill-ring-yank-pointer kill-ring) | |
8793 (if interprogram-cut-function | |
8794 (funcall interprogram-cut-function string (not replace)))) | |
8795 @end group | |
8796 @end smallexample | |
8797 @ignore | |
8798 was: | |
8799 (defun kill-new (string &optional replace) | |
8800 "Make STRING the latest kill in the kill ring. | |
8801 Set the kill-ring-yank pointer to point to it. | |
8802 If `interprogram-cut-function' is non-nil, apply it to STRING. | |
8803 Optional second argument REPLACE non-nil means that STRING will replace | |
8804 the front of the kill ring, rather than being added to the list." | |
8805 (and (fboundp 'menu-bar-update-yank-menu) | |
8806 (menu-bar-update-yank-menu string (and replace (car kill-ring)))) | |
8807 (if (and replace kill-ring) | |
8808 (setcar kill-ring string) | |
8809 (setq kill-ring (cons string kill-ring)) | |
8810 (if (> (length kill-ring) kill-ring-max) | |
8811 (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil))) | |
8812 (setq kill-ring-yank-pointer kill-ring) | |
8813 (if interprogram-cut-function | |
8814 (funcall interprogram-cut-function string (not replace)))) | |
8815 @end ignore | |
8816 | |
8817 (Notice that the function is not interactive.) | |
8818 | |
8819 As usual, we can look at this function in parts. | |
8820 | |
8821 The function definition has an optional @code{yank-handler} argument, | |
8822 which when invoked tells the function how to deal with properties | |
8823 added to the text, such as `bold' or `italics'. We will skip that. | |
8824 | |
8825 @need 1200 | |
8826 The first line of the documentation makes sense: | |
8827 | |
8828 @smallexample | |
8829 Make STRING the latest kill in the kill ring. | |
8830 @end smallexample | |
8831 | |
8832 @noindent | |
8833 Let's skip over the rest of the documentation for the moment. | |
8834 | |
8835 @noindent | |
8836 Also, let's skip over the initial @code{if} expression and those lines | |
8837 of code involving @code{menu-bar-update-yank-menu}. We will explain | |
8838 them below. | |
8839 | |
8840 @need 1200 | |
8841 The critical lines are these: | |
8842 | |
8843 @smallexample | |
8844 @group | |
8845 (if (and replace kill-ring) | |
8846 ;; @r{then} | |
8847 (setcar kill-ring string) | |
8848 @end group | |
8849 @group | |
8850 ;; @r{else} | |
8851 (push string kill-ring) | |
8852 @end group | |
8853 @group | |
8854 (setq kill-ring (cons string kill-ring)) | |
8855 (if (> (length kill-ring) kill-ring-max) | |
8856 ;; @r{avoid overly long kill ring} | |
8857 (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil))) | |
8858 @end group | |
8859 @group | |
8860 (setq kill-ring-yank-pointer kill-ring) | |
8861 (if interprogram-cut-function | |
8862 (funcall interprogram-cut-function string (not replace)))) | |
8863 @end group | |
8864 @end smallexample | |
8865 | |
8866 The conditional test is @w{@code{(and replace kill-ring)}}. | |
8867 This will be true when two conditions are met: the kill ring has | |
8868 something in it, and the @code{replace} variable is true. | |
8869 | |
8870 @need 1250 | |
8871 When the @code{kill-append} function sets @code{replace} to be true | |
8872 and when the kill ring has at least one item in it, the @code{setcar} | |
8873 expression is executed: | |
8874 | |
8875 @smallexample | |
8876 (setcar kill-ring string) | |
8877 @end smallexample | |
8878 | |
8879 The @code{setcar} function actually changes the first element of the | |
8880 @code{kill-ring} list to the value of @code{string}. It replaces the | |
8881 first element. | |
8882 | |
8883 @need 1250 | |
8884 On the other hand, if the kill ring is empty, or replace is false, the | |
8885 else-part of the condition is executed: | |
8886 | |
8887 @smallexample | |
8888 (push string kill-ring) | |
8889 @end smallexample | |
8890 | |
8891 @noindent | |
8892 @need 1250 | |
8893 @code{push} puts its first argument onto the second. It is similar to | |
8894 the older | |
8895 | |
8896 @smallexample | |
8897 (setq kill-ring (cons string kill-ring)) | |
8898 @end smallexample | |
8899 | |
8900 @noindent | |
8901 @need 1250 | |
8902 or the newer | |
8903 | |
8904 @smallexample | |
8905 (add-to-list kill-ring string) | |
8906 @end smallexample | |
8907 | |
8908 @noindent | |
8909 When it is false, the expression first constructs a new version of the | |
8910 kill ring by prepending @code{string} to the existing kill ring as a | |
8911 new element (that is what the @code{push} does). Then it executes a | |
8912 second @code{if} clause. This second @code{if} clause keeps the kill | |
8913 ring from growing too long. | |
8914 | |
8915 Let's look at these two expressions in order. | |
8916 | |
8917 The @code{push} line of the else-part sets the new value of the kill | |
8918 ring to what results from adding the string being killed to the old | |
8919 kill ring. | |
8920 | |
8921 We can see how this works with an example. | |
8922 | |
8923 @need 800 | |
8924 First, | |
8925 | |
8926 @smallexample | |
8927 (setq example-list '("here is a clause" "another clause")) | |
8928 @end smallexample | |
8929 | |
8930 @need 1200 | |
8931 @noindent | |
8932 After evaluating this expression with @kbd{C-x C-e}, you can evaluate | |
8933 @code{example-list} and see what it returns: | |
8934 | |
8935 @smallexample | |
8936 @group | |
8937 example-list | |
8938 @result{} ("here is a clause" "another clause") | |
8939 @end group | |
8940 @end smallexample | |
8941 | |
8942 @need 1200 | |
8943 @noindent | |
8944 Now, we can add a new element on to this list by evaluating the | |
8945 following expression: | |
8946 @findex push, @r{example} | |
8947 | |
8948 @smallexample | |
8949 (push "a third clause" example-list) | |
8950 @end smallexample | |
8951 | |
8952 @need 800 | |
8953 @noindent | |
8954 When we evaluate @code{example-list}, we find its value is: | |
8955 | |
8956 @smallexample | |
8957 @group | |
8958 example-list | |
8959 @result{} ("a third clause" "here is a clause" "another clause") | |
8960 @end group | |
8961 @end smallexample | |
8962 | |
8963 @noindent | |
8964 Thus, the third clause is added to the list by @code{push}. | |
8965 | |
8966 @need 1200 | |
8967 Now for the second part of the @code{if} clause. This expression | |
8968 keeps the kill ring from growing too long. It looks like this: | |
8969 | |
8970 @smallexample | |
8971 @group | |
8972 (if (> (length kill-ring) kill-ring-max) | |
8973 (setcdr (nthcdr (1- kill-ring-max) kill-ring) nil)) | |
8974 @end group | |
8975 @end smallexample | |
8976 | |
8977 The code checks whether the length of the kill ring is greater than | |
8978 the maximum permitted length. This is the value of | |
8979 @code{kill-ring-max} (which is 60, by default). If the length of the | |
8980 kill ring is too long, then this code sets the last element of the | |
8981 kill ring to @code{nil}. It does this by using two functions, | |
8982 @code{nthcdr} and @code{setcdr}. | |
8983 | |
8984 We looked at @code{setcdr} earlier (@pxref{setcdr, , @code{setcdr}}). | |
8985 It sets the @sc{cdr} of a list, just as @code{setcar} sets the | |
8986 @sc{car} of a list. In this case, however, @code{setcdr} will not be | |
8987 setting the @sc{cdr} of the whole kill ring; the @code{nthcdr} | |
8988 function is used to cause it to set the @sc{cdr} of the next to last | |
8989 element of the kill ring---this means that since the @sc{cdr} of the | |
8990 next to last element is the last element of the kill ring, it will set | |
8991 the last element of the kill ring. | |
8992 | |
8993 @findex nthcdr, @r{example} | |
8994 The @code{nthcdr} function works by repeatedly taking the @sc{cdr} of a | |
8995 list---it takes the @sc{cdr} of the @sc{cdr} of the @sc{cdr} | |
8996 @dots{} It does this @var{N} times and returns the results. | |
8997 (@xref{nthcdr, , @code{nthcdr}}.) | |
8998 | |
8999 @findex setcdr, @r{example} | |
9000 Thus, if we had a four element list that was supposed to be three | |
9001 elements long, we could set the @sc{cdr} of the next to last element | |
9002 to @code{nil}, and thereby shorten the list. (If you set the last | |
9003 element to some other value than @code{nil}, which you could do, then | |
9004 you would not have shortened the list. @xref{setcdr, , | |
9005 @code{setcdr}}.) | |
9006 | |
9007 You can see shortening by evaluating the following three expressions | |
9008 in turn. First set the value of @code{trees} to @code{(maple oak pine | |
9009 birch)}, then set the @sc{cdr} of its second @sc{cdr} to @code{nil} | |
9010 and then find the value of @code{trees}: | |
9011 | |
9012 @smallexample | |
9013 @group | |
9014 (setq trees '(maple oak pine birch)) | |
9015 @result{} (maple oak pine birch) | |
9016 @end group | |
9017 | |
9018 @group | |
9019 (setcdr (nthcdr 2 trees) nil) | |
9020 @result{} nil | |
9021 | |
9022 trees | |
9023 @result{} (maple oak pine) | |
9024 @end group | |
9025 @end smallexample | |
9026 | |
9027 @noindent | |
9028 (The value returned by the @code{setcdr} expression is @code{nil} since | |
9029 that is what the @sc{cdr} is set to.) | |
9030 | |
9031 To repeat, in @code{kill-new}, the @code{nthcdr} function takes the | |
9032 @sc{cdr} a number of times that is one less than the maximum permitted | |
9033 size of the kill ring and @code{setcdr} sets the @sc{cdr} of that | |
9034 element (which will be the rest of the elements in the kill ring) to | |
9035 @code{nil}. This prevents the kill ring from growing too long. | |
9036 | |
9037 @need 800 | |
9038 The next to last expression in the @code{kill-new} function is | |
9039 | |
9040 @smallexample | |
9041 (setq kill-ring-yank-pointer kill-ring) | |
9042 @end smallexample | |
9043 | |
9044 The @code{kill-ring-yank-pointer} is a global variable that is set to be | |
9045 the @code{kill-ring}. | |
9046 | |
9047 Even though the @code{kill-ring-yank-pointer} is called a | |
9048 @samp{pointer}, it is a variable just like the kill ring. However, the | |
9049 name has been chosen to help humans understand how the variable is used. | |
9050 | |
9051 @need 1200 | |
9052 Now, to return to an early expression in the body of the function: | |
9053 | |
9054 @smallexample | |
9055 @group | |
9056 (if (fboundp 'menu-bar-update-yank-menu) | |
9057 (menu-bar-update-yank-menu string (and replace (car kill-ring)))) | |
9058 @end group | |
9059 @end smallexample | |
9060 | |
9061 @noindent | |
9062 It starts with an @code{if} expression | |
9063 | |
9064 In this case, the expression tests first to see whether | |
9065 @code{menu-bar-update-yank-menu} exists as a function, and if so, | |
9066 calls it. The @code{fboundp} function returns true if the symbol it | |
9067 is testing has a function definition that `is not void'. If the | |
9068 symbol's function definition were void, we would receive an error | |
9069 message, as we did when we created errors intentionally (@pxref{Making | |
9070 Errors, , Generate an Error Message}). | |
9071 | |
9072 @noindent | |
9073 The then-part contains an expression whose first element is the | |
9074 function @code{and}. | |
9075 | |
9076 @findex and | |
9077 The @code{and} special form evaluates each of its arguments until one | |
9078 of the arguments returns a value of @code{nil}, in which case the | |
9079 @code{and} expression returns @code{nil}; however, if none of the | |
9080 arguments returns a value of @code{nil}, the value resulting from | |
9081 evaluating the last argument is returned. (Since such a value is not | |
9082 @code{nil}, it is considered true in Emacs Lisp.) In other words, an | |
9083 @code{and} expression returns a true value only if all its arguments | |
9084 are true. (@xref{Second Buffer Related Review}.) | |
9085 | |
9086 The expression determines whether the second argument to | |
9087 @code{menu-bar-update-yank-menu} is true or not. | |
9088 @ignore | |
9089 ;; If we're supposed to be extending an existing string, and that | |
9090 ;; string really is at the front of the menu, then update it in place. | |
9091 @end ignore | |
9092 | |
9093 @code{menu-bar-update-yank-menu} is one of the functions that make it | |
9094 possible to use the `Select and Paste' menu in the Edit item of a menu | |
9095 bar; using a mouse, you can look at the various pieces of text you | |
9096 have saved and select one piece to paste. | |
9097 | |
9098 The last expression in the @code{kill-new} function adds the newly | |
9099 copied string to whatever facility exists for copying and pasting | |
9100 among different programs running in a windowing system. In the X | |
9101 Windowing system, for example, the @code{x-select-text} function takes | |
9102 the string and stores it in memory operated by X. You can paste the | |
9103 string in another program, such as an Xterm. | |
9104 | |
9105 @need 1200 | |
9106 The expression looks like this: | |
9107 | |
9108 @smallexample | |
9109 @group | |
9110 (if interprogram-cut-function | |
9111 (funcall interprogram-cut-function string (not replace)))) | |
9112 @end group | |
9113 @end smallexample | |
9114 | |
9115 If an @code{interprogram-cut-function} exists, then Emacs executes | |
9116 @code{funcall}, which in turn calls its first argument as a function | |
9117 and passes the remaining arguments to it. (Incidentally, as far as I | |
9118 can see, this @code{if} expression could be replaced by an @code{and} | |
9119 expression similar to the one in the first part of the function.) | |
9120 | |
9121 We are not going to discuss windowing systems and other programs | |
9122 further, but merely note that this is a mechanism that enables GNU | |
9123 Emacs to work easily and well with other programs. | |
9124 | |
9125 This code for placing text in the kill ring, either concatenated with | |
9126 an existing element or as a new element, leads us to the code for | |
9127 bringing back text that has been cut out of the buffer---the yank | |
9128 commands. However, before discussing the yank commands, it is better | |
9129 to learn how lists are implemented in a computer. This will make | |
9130 clear such mysteries as the use of the term `pointer'. But before | |
9131 that, we will digress into C. | |
9132 | |
9133 @ignore | |
9134 @c is this true in Emacs 22? Does not seems to be | |
9135 | |
9136 (If the @w{@code{(< end beg))}} | |
9137 expression is true, @code{kill-append} prepends the string to the just | |
9138 previously clipped text. For a detailed discussion, see | |
9139 @ref{kill-append function, , The @code{kill-append} function}.) | |
9140 | |
9141 If you then yank back the text, i.e., `paste' it, you get both | |
9142 pieces of text at once. That way, if you delete two words in a row, | |
9143 and then yank them back, you get both words, in their proper order, | |
9144 with one yank. (The @w{@code{(< end beg))}} expression makes sure the | |
9145 order is correct.) | |
9146 | |
9147 On the other hand, if the previous command is not @code{kill-region}, | |
9148 then the @code{kill-new} function is called, which adds the text to | |
9149 the kill ring as the latest item, and sets the | |
9150 @code{kill-ring-yank-pointer} variable to point to it. | |
9151 @end ignore | |
9152 @ignore | |
9153 | |
9154 @c Evidently, changed for Emacs 22. The zap-to-char command does not | |
9155 @c use the delete-and-extract-region function | |
9156 | |
9157 2006 Oct 26, the Digression into C is now OK but should come after | |
9158 copy-region-as-kill and filter-buffer-substring | |
9159 | |
9160 2006 Oct 24 | |
9161 In Emacs 22, | |
9162 copy-region-as-kill is short, 12 lines, and uses | |
9163 filter-buffer-substring, which is longer, 39 lines | |
9164 and has delete-and-extract-region in it. | |
9165 delete-and-extract-region is written in C. | |
9166 | |
9167 see Initializing a Variable with @code{defvar} | |
9168 @end ignore | |
9169 | |
9170 @node Digression into C, defvar, copy-region-as-kill, Cutting & Storing Text | |
9171 @comment node-name, next, previous, up | |
9172 @section Digression into C | |
9173 @findex delete-and-extract-region | |
9174 @cindex C, a digression into | |
9175 @cindex Digression into C | |
9176 | |
9177 The @code{copy-region-as-kill} function (@pxref{copy-region-as-kill, , | |
9178 @code{copy-region-as-kill}}) uses the @code{filter-buffer-substring} | |
9179 function, which in turn uses the @code{delete-and-extract-region} | |
9180 function. It removes the contents of a region and you cannot get them | |
9181 back. | |
9182 | |
9183 Unlike the other code discussed here, the | |
9184 @code{delete-and-extract-region} function is not written in Emacs | |
9185 Lisp; it is written in C and is one of the primitives of the GNU Emacs | |
9186 system. Since it is very simple, I will digress briefly from Lisp and | |
9187 describe it here. | |
9188 | |
9189 @c GNU Emacs 22 in /usr/local/src/emacs/src/editfns.c | |
9190 @c the DEFUN for buffer-substring-no-properties | |
9191 | |
9192 @need 1500 | |
9193 Like many of the other Emacs primitives, | |
9194 @code{delete-and-extract-region} is written as an instance of a C | |
9195 macro, a macro being a template for code. The complete macro looks | |
9196 like this: | |
9197 | |
9198 @smallexample | |
9199 @group | |
9200 DEFUN ("buffer-substring-no-properties", Fbuffer_substring_no_properties, | |
9201 Sbuffer_substring_no_properties, 2, 2, 0, | |
9202 doc: /* Return the characters of part of the buffer, | |
9203 without the text properties. | |
9204 The two arguments START and END are character positions; | |
9205 they can be in either order. */) | |
9206 (start, end) | |
9207 Lisp_Object start, end; | |
9208 @{ | |
9209 register int b, e; | |
9210 | |
9211 validate_region (&start, &end); | |
9212 b = XINT (start); | |
9213 e = XINT (end); | |
9214 | |
9215 return make_buffer_string (b, e, 0); | |
9216 @} | |
9217 @end group | |
9218 @end smallexample | |
9219 | |
9220 Without going into the details of the macro writing process, let me | |
9221 point out that this macro starts with the word @code{DEFUN}. The word | |
9222 @code{DEFUN} was chosen since the code serves the same purpose as | |
9223 @code{defun} does in Lisp. (The @code{DEFUN} C macro is defined in | |
9224 @file{emacs/src/lisp.h}.) | |
9225 | |
9226 The word @code{DEFUN} is followed by seven parts inside of | |
9227 parentheses: | |
9228 | |
9229 @itemize @bullet | |
9230 @item | |
9231 The first part is the name given to the function in Lisp, | |
9232 @code{delete-and-extract-region}. | |
9233 | |
9234 @item | |
9235 The second part is the name of the function in C, | |
9236 @code{Fdelete_and_extract_region}. By convention, it starts with | |
9237 @samp{F}. Since C does not use hyphens in names, underscores are used | |
9238 instead. | |
9239 | |
9240 @item | |
9241 The third part is the name for the C constant structure that records | |
9242 information on this function for internal use. It is the name of the | |
9243 function in C but begins with an @samp{S} instead of an @samp{F}. | |
9244 | |
9245 @item | |
9246 The fourth and fifth parts specify the minimum and maximum number of | |
9247 arguments the function can have. This function demands exactly 2 | |
9248 arguments. | |
9249 | |
9250 @item | |
9251 The sixth part is nearly like the argument that follows the | |
9252 @code{interactive} declaration in a function written in Lisp: a letter | |
9253 followed, perhaps, by a prompt. The only difference from the Lisp is | |
9254 when the macro is called with no arguments. Then you write a @code{0} | |
9255 (which is a `null string'), as in this macro. | |
9256 | |
9257 If you were to specify arguments, you would place them between | |
9258 quotation marks. The C macro for @code{goto-char} includes | |
9259 @code{"NGoto char: "} in this position to indicate that the function | |
9260 expects a raw prefix, in this case, a numerical location in a buffer, | |
9261 and provides a prompt. | |
9262 | |
9263 @item | |
9264 The seventh part is a documentation string, just like the one for a | |
9265 function written in Emacs Lisp, except that every newline must be | |
9266 written explicitly as @samp{\n} followed by a backslash and carriage | |
9267 return. | |
9268 | |
9269 @need 1000 | |
9270 Thus, the first two lines of documentation for @code{goto-char} are | |
9271 written like this: | |
9272 | |
9273 @smallexample | |
9274 @group | |
9275 "Set point to POSITION, a number or marker.\n\ | |
9276 Beginning of buffer is position (point-min), end is (point-max)." | |
9277 @end group | |
9278 @end smallexample | |
9279 @end itemize | |
9280 | |
9281 @need 1200 | |
9282 In a C macro, the formal parameters come next, with a statement of | |
9283 what kind of object they are, followed by what might be called the `body' | |
9284 of the macro. For @code{delete-and-extract-region} the `body' | |
9285 consists of the following four lines: | |
9286 | |
9287 @smallexample | |
9288 @group | |
9289 validate_region (&start, &end); | |
9290 if (XINT (start) == XINT (end)) | |
9291 return build_string (""); | |
9292 return del_range_1 (XINT (start), XINT (end), 1, 1); | |
9293 @end group | |
9294 @end smallexample | |
9295 | |
9296 The @code{validate_region} function checks whether the values | |
9297 passed as the beginning and end of the region are the proper type and | |
9298 are within range. If the beginning and end positions are the same, | |
9299 then return and empty string. | |
9300 | |
9301 The @code{del_range_1} function actually deletes the text. It is a | |
9302 complex function we will not look into. It updates the buffer and | |
9303 does other things. However, it is worth looking at the two arguments | |
9304 passed to @code{del_range}. These are @w{@code{XINT (start)}} and | |
9305 @w{@code{XINT (end)}}. | |
9306 | |
9307 As far as the C language is concerned, @code{start} and @code{end} are | |
9308 two integers that mark the beginning and end of the region to be | |
9309 deleted@footnote{More precisely, and requiring more expert knowledge | |
9310 to understand, the two integers are of type `Lisp_Object', which can | |
9311 also be a C union instead of an integer type.}. | |
9312 | |
9313 In early versions of Emacs, these two numbers were thirty-two bits | |
9314 long, but the code is slowly being generalized to handle other | |
9315 lengths. Three of the available bits are used to specify the type of | |
9316 information; the remaining bits are used as `content'. | |
9317 | |
9318 @samp{XINT} is a C macro that extracts the relevant number from the | |
9319 longer collection of bits; the three other bits are discarded. | |
9320 | |
9321 @need 800 | |
9322 The command in @code{delete-and-extract-region} looks like this: | |
9323 | |
9324 @smallexample | |
9325 del_range_1 (XINT (start), XINT (end), 1, 1); | |
9326 @end smallexample | |
9327 | |
9328 @noindent | |
9329 It deletes the region between the beginning position, @code{start}, | |
9330 and the ending position, @code{end}. | |
9331 | |
9332 From the point of view of the person writing Lisp, Emacs is all very | |
9333 simple; but hidden underneath is a great deal of complexity to make it | |
9334 all work. | |
9335 | |
9336 @node defvar, cons & search-fwd Review, Digression into C, Cutting & Storing Text | |
9337 @comment node-name, next, previous, up | |
9338 @section Initializing a Variable with @code{defvar} | |
9339 @findex defvar | |
9340 @cindex Initializing a variable | |
9341 @cindex Variable initialization | |
9342 | |
9343 @ignore | |
9344 2006 Oct 24 | |
9345 In Emacs 22, | |
9346 copy-region-as-kill is short, 12 lines, and uses | |
9347 filter-buffer-substring, which is longer, 39 lines | |
9348 and has delete-and-extract-region in it. | |
9349 delete-and-extract-region is written in C. | |
9350 | |
9351 see Initializing a Variable with @code{defvar} | |
9352 | |
9353 @end ignore | |
9354 | |
9355 The @code{copy-region-as-kill} function is written in Emacs Lisp. Two | |
9356 functions within it, @code{kill-append} and @code{kill-new}, copy a | |
9357 region in a buffer and save it in a variable called the | |
9358 @code{kill-ring}. This section describes how the @code{kill-ring} | |
9359 variable is created and initialized using the @code{defvar} special | |
9360 form. | |
9361 | |
9362 (Again we note that the term @code{kill-ring} is a misnomer. The text | |
9363 that is clipped out of the buffer can be brought back; it is not a ring | |
9364 of corpses, but a ring of resurrectable text.) | |
9365 | |
9366 In Emacs Lisp, a variable such as the @code{kill-ring} is created and | |
9367 given an initial value by using the @code{defvar} special form. The | |
9368 name comes from ``define variable''. | |
9369 | |
9370 The @code{defvar} special form is similar to @code{setq} in that it sets | |
9371 the value of a variable. It is unlike @code{setq} in two ways: first, | |
9372 it only sets the value of the variable if the variable does not already | |
9373 have a value. If the variable already has a value, @code{defvar} does | |
9374 not override the existing value. Second, @code{defvar} has a | |
9375 documentation string. | |
9376 | |
9377 (Another special form, @code{defcustom}, is designed for variables | |
9378 that people customize. It has more features than @code{defvar}. | |
9379 (@xref{defcustom, , Setting Variables with @code{defcustom}}.) | |
9380 | |
9381 @menu | |
9382 * See variable current value:: | |
9383 * defvar and asterisk:: | |
9384 @end menu | |
9385 | |
9386 @node See variable current value, defvar and asterisk, defvar, defvar | |
9387 @ifnottex | |
9388 @unnumberedsubsec Seeing the Current Value of a Variable | |
9389 @end ifnottex | |
9390 | |
9391 You can see the current value of a variable, any variable, by using | |
9392 the @code{describe-variable} function, which is usually invoked by | |
9393 typing @kbd{C-h v}. If you type @kbd{C-h v} and then @code{kill-ring} | |
9394 (followed by @key{RET}) when prompted, you will see what is in your | |
9395 current kill ring---this may be quite a lot! Conversely, if you have | |
9396 been doing nothing this Emacs session except read this document, you | |
9397 may have nothing in it. Also, you will see the documentation for | |
9398 @code{kill-ring}: | |
9399 | |
9400 @smallexample | |
9401 @group | |
9402 Documentation: | |
9403 List of killed text sequences. | |
9404 Since the kill ring is supposed to interact nicely with cut-and-paste | |
9405 facilities offered by window systems, use of this variable should | |
9406 @end group | |
9407 @group | |
9408 interact nicely with `interprogram-cut-function' and | |
9409 `interprogram-paste-function'. The functions `kill-new', | |
9410 `kill-append', and `current-kill' are supposed to implement this | |
9411 interaction; you may want to use them instead of manipulating the kill | |
9412 ring directly. | |
9413 @end group | |
9414 @end smallexample | |
9415 | |
9416 @need 800 | |
9417 The kill ring is defined by a @code{defvar} in the following way: | |
9418 | |
9419 @smallexample | |
9420 @group | |
9421 (defvar kill-ring nil | |
9422 "List of killed text sequences. | |
9423 @dots{}") | |
9424 @end group | |
9425 @end smallexample | |
9426 | |
9427 @noindent | |
9428 In this variable definition, the variable is given an initial value of | |
9429 @code{nil}, which makes sense, since if you have saved nothing, you want | |
9430 nothing back if you give a @code{yank} command. The documentation | |
9431 string is written just like the documentation string of a @code{defun}. | |
9432 As with the documentation string of the @code{defun}, the first line of | |
9433 the documentation should be a complete sentence, since some commands, | |
9434 like @code{apropos}, print only the first line of documentation. | |
9435 Succeeding lines should not be indented; otherwise they look odd when | |
9436 you use @kbd{C-h v} (@code{describe-variable}). | |
9437 | |
9438 @node defvar and asterisk, , See variable current value, defvar | |
9439 @subsection @code{defvar} and an asterisk | |
9440 @findex defvar @r{for a user customizable variable} | |
9441 @findex defvar @r{with an asterisk} | |
9442 | |
9443 In the past, Emacs used the @code{defvar} special form both for | |
9444 internal variables that you would not expect a user to change and for | |
9445 variables that you do expect a user to change. Although you can still | |
9446 use @code{defvar} for user customizable variables, please use | |
9447 @code{defcustom} instead, since that special form provides a path into | |
9448 the Customization commands. (@xref{defcustom, , Specifying Variables | |
9449 using @code{defcustom}}.) | |
9450 | |
9451 When you specified a variable using the @code{defvar} special form, | |
9452 you could distinguish a readily settable variable from others by | |
9453 typing an asterisk, @samp{*}, in the first column of its documentation | |
9454 string. For example: | |
9455 | |
9456 @smallexample | |
9457 @group | |
9458 (defvar shell-command-default-error-buffer nil | |
9459 "*Buffer name for `shell-command' @dots{} error output. | |
9460 @dots{} ") | |
9461 @end group | |
9462 @end smallexample | |
9463 | |
9464 @findex set-variable | |
9465 @noindent | |
9466 You could (and still can) use the @code{set-variable} command to | |
9467 change the value of @code{shell-command-default-error-buffer} | |
9468 temporarily. However, options set using @code{set-variable} are set | |
9469 only for the duration of your editing session. The new values are not | |
9470 saved between sessions. Each time Emacs starts, it reads the original | |
9471 value, unless you change the value within your @file{.emacs} file, | |
9472 either by setting it manually or by using @code{customize}. | |
9473 @xref{Emacs Initialization, , Your @file{.emacs} File}. | |
9474 | |
9475 For me, the major use of the @code{set-variable} command is to suggest | |
9476 variables that I might want to set in my @file{.emacs} file. There | |
9477 are now more than 700 such variables --- far too many to remember | |
9478 readily. Fortunately, you can press @key{TAB} after calling the | |
9479 @code{M-x set-variable} command to see the list of variables. | |
9480 (@xref{Examining, , Examining and Setting Variables, emacs, | |
9481 The GNU Emacs Manual}.) | |
9482 | |
9483 @need 1250 | |
9484 @node cons & search-fwd Review, search Exercises, defvar, Cutting & Storing Text | |
9485 @comment node-name, next, previous, up | |
9486 @section Review | |
9487 | |
9488 Here is a brief summary of some recently introduced functions. | |
9489 | |
9490 @table @code | |
9491 @item car | |
9492 @itemx cdr | |
9493 @code{car} returns the first element of a list; @code{cdr} returns the | |
9494 second and subsequent elements of a list. | |
9495 | |
9496 @need 1250 | |
9497 For example: | |
9498 | |
9499 @smallexample | |
9500 @group | |
9501 (car '(1 2 3 4 5 6 7)) | |
9502 @result{} 1 | |
9503 (cdr '(1 2 3 4 5 6 7)) | |
9504 @result{} (2 3 4 5 6 7) | |
9505 @end group | |
9506 @end smallexample | |
9507 | |
9508 @item cons | |
9509 @code{cons} constructs a list by prepending its first argument to its | |
9510 second argument. | |
9511 | |
9512 @need 1250 | |
9513 For example: | |
9514 | |
9515 @smallexample | |
9516 @group | |
9517 (cons 1 '(2 3 4)) | |
9518 @result{} (1 2 3 4) | |
9519 @end group | |
9520 @end smallexample | |
9521 | |
9522 @item funcall | |
9523 @code{funcall} evaluates its first argument as a function. It passes | |
9524 its remaining arguments to its first argument. | |
9525 | |
9526 @item nthcdr | |
9527 Return the result of taking @sc{cdr} `n' times on a list. | |
9528 @iftex | |
9529 The | |
9530 @tex | |
9531 $n^{th}$ | |
9532 @end tex | |
9533 @code{cdr}. | |
9534 @end iftex | |
9535 The `rest of the rest', as it were. | |
9536 | |
9537 @need 1250 | |
9538 For example: | |
9539 | |
9540 @smallexample | |
9541 @group | |
9542 (nthcdr 3 '(1 2 3 4 5 6 7)) | |
9543 @result{} (4 5 6 7) | |
9544 @end group | |
9545 @end smallexample | |
9546 | |
9547 @item setcar | |
9548 @itemx setcdr | |
9549 @code{setcar} changes the first element of a list; @code{setcdr} | |
9550 changes the second and subsequent elements of a list. | |
9551 | |
9552 @need 1250 | |
9553 For example: | |
9554 | |
9555 @smallexample | |
9556 @group | |
9557 (setq triple '(1 2 3)) | |
9558 | |
9559 (setcar triple '37) | |
9560 | |
9561 triple | |
9562 @result{} (37 2 3) | |
9563 | |
9564 (setcdr triple '("foo" "bar")) | |
9565 | |
9566 triple | |
9567 @result{} (37 "foo" "bar") | |
9568 @end group | |
9569 @end smallexample | |
9570 | |
9571 @item progn | |
9572 Evaluate each argument in sequence and then return the value of the | |
9573 last. | |
9574 | |
9575 @need 1250 | |
9576 For example: | |
9577 | |
9578 @smallexample | |
9579 @group | |
9580 (progn 1 2 3 4) | |
9581 @result{} 4 | |
9582 @end group | |
9583 @end smallexample | |
9584 | |
9585 @item save-restriction | |
9586 Record whatever narrowing is in effect in the current buffer, if any, | |
9587 and restore that narrowing after evaluating the arguments. | |
9588 | |
9589 @item search-forward | |
9590 Search for a string, and if the string is found, move point. With a | |
9591 regular expression, use the similar @code{re-search-forward}. | |
9592 (@xref{Regexp Search, , Regular Expression Searches}, for an | |
9593 explanation of regular expression patterns and searches.) | |
9594 | |
9595 @need 1250 | |
9596 @noindent | |
9597 @code{search-forward} and @code{re-search-forward} take four | |
9598 arguments: | |
9599 | |
9600 @enumerate | |
9601 @item | |
9602 The string or regular expression to search for. | |
9603 | |
9604 @item | |
9605 Optionally, the limit of the search. | |
9606 | |
9607 @item | |
9608 Optionally, what to do if the search fails, return @code{nil} or an | |
9609 error message. | |
9610 | |
9611 @item | |
9612 Optionally, how many times to repeat the search; if negative, the | |
9613 search goes backwards. | |
9614 @end enumerate | |
9615 | |
9616 @item kill-region | |
9617 @itemx delete-and-extract-region | |
9618 @itemx copy-region-as-kill | |
9619 | |
9620 @code{kill-region} cuts the text between point and mark from the | |
9621 buffer and stores that text in the kill ring, so you can get it back | |
9622 by yanking. | |
9623 | |
9624 @code{copy-region-as-kill} copies the text between point and mark into | |
9625 the kill ring, from which you can get it by yanking. The function | |
9626 does not cut or remove the text from the buffer. | |
9627 @end table | |
9628 | |
9629 @code{delete-and-extract-region} removes the text between point and | |
9630 mark from the buffer and throws it away. You cannot get it back. | |
9631 (This is not an interactive command.) | |
9632 | |
9633 @need 1500 | |
9634 @node search Exercises, , cons & search-fwd Review, Cutting & Storing Text | |
9635 @section Searching Exercises | |
9636 | |
9637 @itemize @bullet | |
9638 @item | |
9639 Write an interactive function that searches for a string. If the | |
9640 search finds the string, leave point after it and display a message | |
9641 that says ``Found!''. (Do not use @code{search-forward} for the name | |
9642 of this function; if you do, you will overwrite the existing version of | |
9643 @code{search-forward} that comes with Emacs. Use a name such as | |
9644 @code{test-search} instead.) | |
9645 | |
9646 @item | |
9647 Write a function that prints the third element of the kill ring in the | |
9648 echo area, if any; if the kill ring does not contain a third element, | |
9649 print an appropriate message. | |
9650 @end itemize | |
9651 | |
9652 @node List Implementation, Yanking, Cutting & Storing Text, Top | |
9653 @comment node-name, next, previous, up | |
9654 @chapter How Lists are Implemented | |
9655 @cindex Lists in a computer | |
9656 | |
9657 In Lisp, atoms are recorded in a straightforward fashion; if the | |
9658 implementation is not straightforward in practice, it is, nonetheless, | |
9659 straightforward in theory. The atom @samp{rose}, for example, is | |
9660 recorded as the four contiguous letters @samp{r}, @samp{o}, @samp{s}, | |
9661 @samp{e}. A list, on the other hand, is kept differently. The mechanism | |
9662 is equally simple, but it takes a moment to get used to the idea. A | |
9663 list is kept using a series of pairs of pointers. In the series, the | |
9664 first pointer in each pair points to an atom or to another list, and the | |
9665 second pointer in each pair points to the next pair, or to the symbol | |
9666 @code{nil}, which marks the end of the list. | |
9667 | |
9668 A pointer itself is quite simply the electronic address of what is | |
9669 pointed to. Hence, a list is kept as a series of electronic addresses. | |
9670 | |
9671 @menu | |
9672 * Lists diagrammed:: | |
9673 * Symbols as Chest:: Exploring a powerful metaphor. | |
9674 * List Exercise:: | |
9675 @end menu | |
9676 | |
9677 @node Lists diagrammed, Symbols as Chest, List Implementation, List Implementation | |
9678 @ifnottex | |
9679 @unnumberedsec Lists diagrammed | |
9680 @end ifnottex | |
9681 | |
9682 For example, the list @code{(rose violet buttercup)} has three elements, | |
9683 @samp{rose}, @samp{violet}, and @samp{buttercup}. In the computer, the | |
9684 electronic address of @samp{rose} is recorded in a segment of computer | |
9685 memory along with the address that gives the electronic address of where | |
9686 the atom @samp{violet} is located; and that address (the one that tells | |
9687 where @samp{violet} is located) is kept along with an address that tells | |
9688 where the address for the atom @samp{buttercup} is located. | |
9689 | |
9690 @need 1200 | |
9691 This sounds more complicated than it is and is easier seen in a diagram: | |
9692 | |
9693 @c clear print-postscript-figures | |
9694 @c !!! cons-cell-diagram #1 | |
9695 @ifnottex | |
9696 @smallexample | |
9697 @group | |
9698 ___ ___ ___ ___ ___ ___ | |
9699 |___|___|--> |___|___|--> |___|___|--> nil | |
9700 | | | | |
9701 | | | | |
9702 --> rose --> violet --> buttercup | |
9703 @end group | |
9704 @end smallexample | |
9705 @end ifnottex | |
9706 @ifset print-postscript-figures | |
9707 @sp 1 | |
9708 @tex | |
9709 @center @image{cons-1} | |
9710 %%%% old method of including an image | |
9711 % \input /usr/local/lib/tex/inputs/psfig.tex | |
9712 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-1.eps}} | |
9713 % \catcode`\@=0 % | |
9714 @end tex | |
9715 @sp 1 | |
9716 @end ifset | |
9717 @ifclear print-postscript-figures | |
9718 @iftex | |
9719 @smallexample | |
9720 @group | |
9721 ___ ___ ___ ___ ___ ___ | |
9722 |___|___|--> |___|___|--> |___|___|--> nil | |
9723 | | | | |
9724 | | | | |
9725 --> rose --> violet --> buttercup | |
9726 @end group | |
9727 @end smallexample | |
9728 @end iftex | |
9729 @end ifclear | |
9730 | |
9731 @noindent | |
9732 In the diagram, each box represents a word of computer memory that | |
9733 holds a Lisp object, usually in the form of a memory address. The boxes, | |
9734 i.e.@: the addresses, are in pairs. Each arrow points to what the address | |
9735 is the address of, either an atom or another pair of addresses. The | |
9736 first box is the electronic address of @samp{rose} and the arrow points | |
9737 to @samp{rose}; the second box is the address of the next pair of boxes, | |
9738 the first part of which is the address of @samp{violet} and the second | |
9739 part of which is the address of the next pair. The very last box | |
9740 points to the symbol @code{nil}, which marks the end of the list. | |
9741 | |
9742 @need 1200 | |
9743 When a variable is set to a list with a function such as @code{setq}, | |
9744 it stores the address of the first box in the variable. Thus, | |
9745 evaluation of the expression | |
9746 | |
9747 @smallexample | |
9748 (setq bouquet '(rose violet buttercup)) | |
9749 @end smallexample | |
9750 | |
9751 @need 1250 | |
9752 @noindent | |
9753 creates a situation like this: | |
9754 | |
9755 @c cons-cell-diagram #2 | |
9756 @ifnottex | |
9757 @smallexample | |
9758 @group | |
9759 bouquet | |
9760 | | |
9761 | ___ ___ ___ ___ ___ ___ | |
9762 --> |___|___|--> |___|___|--> |___|___|--> nil | |
9763 | | | | |
9764 | | | | |
9765 --> rose --> violet --> buttercup | |
9766 @end group | |
9767 @end smallexample | |
9768 @end ifnottex | |
9769 @ifset print-postscript-figures | |
9770 @sp 1 | |
9771 @tex | |
9772 @center @image{cons-2} | |
9773 %%%% old method of including an image | |
9774 % \input /usr/local/lib/tex/inputs/psfig.tex | |
9775 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-2.eps}} | |
9776 % \catcode`\@=0 % | |
9777 @end tex | |
9778 @sp 1 | |
9779 @end ifset | |
9780 @ifclear print-postscript-figures | |
9781 @iftex | |
9782 @smallexample | |
9783 @group | |
9784 bouquet | |
9785 | | |
9786 | ___ ___ ___ ___ ___ ___ | |
9787 --> |___|___|--> |___|___|--> |___|___|--> nil | |
9788 | | | | |
9789 | | | | |
9790 --> rose --> violet --> buttercup | |
9791 @end group | |
9792 @end smallexample | |
9793 @end iftex | |
9794 @end ifclear | |
9795 | |
9796 @noindent | |
9797 In this example, the symbol @code{bouquet} holds the address of the first | |
9798 pair of boxes. | |
9799 | |
9800 @need 1200 | |
9801 This same list can be illustrated in a different sort of box notation | |
9802 like this: | |
9803 | |
9804 @c cons-cell-diagram #2a | |
9805 @ifnottex | |
9806 @smallexample | |
9807 @group | |
9808 bouquet | |
9809 | | |
9810 | -------------- --------------- ---------------- | |
9811 | | car | cdr | | car | cdr | | car | cdr | | |
9812 -->| rose | o------->| violet | o------->| butter- | nil | | |
9813 | | | | | | | cup | | | |
9814 -------------- --------------- ---------------- | |
9815 @end group | |
9816 @end smallexample | |
9817 @end ifnottex | |
9818 @ifset print-postscript-figures | |
9819 @sp 1 | |
9820 @tex | |
9821 @center @image{cons-2a} | |
9822 %%%% old method of including an image | |
9823 % \input /usr/local/lib/tex/inputs/psfig.tex | |
9824 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-2a.eps}} | |
9825 % \catcode`\@=0 % | |
9826 @end tex | |
9827 @sp 1 | |
9828 @end ifset | |
9829 @ifclear print-postscript-figures | |
9830 @iftex | |
9831 @smallexample | |
9832 @group | |
9833 bouquet | |
9834 | | |
9835 | -------------- --------------- ---------------- | |
9836 | | car | cdr | | car | cdr | | car | cdr | | |
9837 -->| rose | o------->| violet | o------->| butter- | nil | | |
9838 | | | | | | | cup | | | |
9839 -------------- --------------- ---------------- | |
9840 @end group | |
9841 @end smallexample | |
9842 @end iftex | |
9843 @end ifclear | |
9844 | |
9845 (Symbols consist of more than pairs of addresses, but the structure of | |
9846 a symbol is made up of addresses. Indeed, the symbol @code{bouquet} | |
9847 consists of a group of address-boxes, one of which is the address of | |
9848 the printed word @samp{bouquet}, a second of which is the address of a | |
9849 function definition attached to the symbol, if any, a third of which | |
9850 is the address of the first pair of address-boxes for the list | |
9851 @code{(rose violet buttercup)}, and so on. Here we are showing that | |
9852 the symbol's third address-box points to the first pair of | |
9853 address-boxes for the list.) | |
9854 | |
9855 If a symbol is set to the @sc{cdr} of a list, the list itself is not | |
9856 changed; the symbol simply has an address further down the list. (In | |
9857 the jargon, @sc{car} and @sc{cdr} are `non-destructive'.) Thus, | |
9858 evaluation of the following expression | |
9859 | |
9860 @smallexample | |
9861 (setq flowers (cdr bouquet)) | |
9862 @end smallexample | |
9863 | |
9864 @need 800 | |
9865 @noindent | |
9866 produces this: | |
9867 | |
9868 @c cons-cell-diagram #3 | |
9869 @ifnottex | |
9870 @sp 1 | |
9871 @smallexample | |
9872 @group | |
9873 bouquet flowers | |
9874 | | | |
9875 | ___ ___ | ___ ___ ___ ___ | |
9876 --> | | | --> | | | | | | | |
9877 |___|___|----> |___|___|--> |___|___|--> nil | |
9878 | | | | |
9879 | | | | |
9880 --> rose --> violet --> buttercup | |
9881 @end group | |
9882 @end smallexample | |
9883 @sp 1 | |
9884 @end ifnottex | |
9885 @ifset print-postscript-figures | |
9886 @sp 1 | |
9887 @tex | |
9888 @center @image{cons-3} | |
9889 %%%% old method of including an image | |
9890 % \input /usr/local/lib/tex/inputs/psfig.tex | |
9891 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-3.eps}} | |
9892 % \catcode`\@=0 % | |
9893 @end tex | |
9894 @sp 1 | |
9895 @end ifset | |
9896 @ifclear print-postscript-figures | |
9897 @iftex | |
9898 @sp 1 | |
9899 @smallexample | |
9900 @group | |
9901 bouquet flowers | |
9902 | | | |
9903 | ___ ___ | ___ ___ ___ ___ | |
9904 --> | | | --> | | | | | | | |
9905 |___|___|----> |___|___|--> |___|___|--> nil | |
9906 | | | | |
9907 | | | | |
9908 --> rose --> violet --> buttercup | |
9909 @end group | |
9910 @end smallexample | |
9911 @sp 1 | |
9912 @end iftex | |
9913 @end ifclear | |
9914 | |
9915 @noindent | |
9916 The value of @code{flowers} is @code{(violet buttercup)}, which is | |
9917 to say, the symbol @code{flowers} holds the address of the pair of | |
9918 address-boxes, the first of which holds the address of @code{violet}, | |
9919 and the second of which holds the address of @code{buttercup}. | |
9920 | |
9921 A pair of address-boxes is called a @dfn{cons cell} or @dfn{dotted | |
9922 pair}. @xref{Cons Cell Type, , Cons Cell and List Types, elisp, The GNU Emacs Lisp | |
9923 Reference Manual}, and @ref{Dotted Pair Notation, , Dotted Pair | |
9924 Notation, elisp, The GNU Emacs Lisp Reference Manual}, for more | |
9925 information about cons cells and dotted pairs. | |
9926 | |
9927 @need 1200 | |
9928 The function @code{cons} adds a new pair of addresses to the front of | |
9929 a series of addresses like that shown above. For example, evaluating | |
9930 the expression | |
9931 | |
9932 @smallexample | |
9933 (setq bouquet (cons 'lily bouquet)) | |
9934 @end smallexample | |
9935 | |
9936 @need 1500 | |
9937 @noindent | |
9938 produces: | |
9939 | |
9940 @c cons-cell-diagram #4 | |
9941 @ifnottex | |
9942 @sp 1 | |
9943 @smallexample | |
9944 @group | |
9945 bouquet flowers | |
9946 | | | |
9947 | ___ ___ ___ ___ | ___ ___ ___ ___ | |
9948 --> | | | | | | --> | | | | | | | |
9949 |___|___|----> |___|___|----> |___|___|---->|___|___|--> nil | |
9950 | | | | | |
9951 | | | | | |
9952 --> lily --> rose --> violet --> buttercup | |
9953 @end group | |
9954 @end smallexample | |
9955 @sp 1 | |
9956 @end ifnottex | |
9957 @ifset print-postscript-figures | |
9958 @sp 1 | |
9959 @tex | |
9960 @center @image{cons-4} | |
9961 %%%% old method of including an image | |
9962 % \input /usr/local/lib/tex/inputs/psfig.tex | |
9963 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-4.eps}} | |
9964 % \catcode`\@=0 % | |
9965 @end tex | |
9966 @sp 1 | |
9967 @end ifset | |
9968 @ifclear print-postscript-figures | |
9969 @iftex | |
9970 @sp 1 | |
9971 @smallexample | |
9972 @group | |
9973 bouquet flowers | |
9974 | | | |
9975 | ___ ___ ___ ___ | ___ ___ ___ ___ | |
9976 --> | | | | | | --> | | | | | | | |
9977 |___|___|----> |___|___|----> |___|___|---->|___|___|--> nil | |
9978 | | | | | |
9979 | | | | | |
9980 --> lily --> rose --> violet --> buttercup | |
9981 @end group | |
9982 @end smallexample | |
9983 @sp 1 | |
9984 @end iftex | |
9985 @end ifclear | |
9986 | |
9987 @need 1200 | |
9988 @noindent | |
9989 However, this does not change the value of the symbol | |
9990 @code{flowers}, as you can see by evaluating the following, | |
9991 | |
9992 @smallexample | |
9993 (eq (cdr (cdr bouquet)) flowers) | |
9994 @end smallexample | |
9995 | |
9996 @noindent | |
9997 which returns @code{t} for true. | |
9998 | |
9999 Until it is reset, @code{flowers} still has the value | |
10000 @code{(violet buttercup)}; that is, it has the address of the cons | |
10001 cell whose first address is of @code{violet}. Also, this does not | |
10002 alter any of the pre-existing cons cells; they are all still there. | |
10003 | |
10004 Thus, in Lisp, to get the @sc{cdr} of a list, you just get the address | |
10005 of the next cons cell in the series; to get the @sc{car} of a list, | |
10006 you get the address of the first element of the list; to @code{cons} a | |
10007 new element on a list, you add a new cons cell to the front of the list. | |
10008 That is all there is to it! The underlying structure of Lisp is | |
10009 brilliantly simple! | |
10010 | |
10011 And what does the last address in a series of cons cells refer to? It | |
10012 is the address of the empty list, of @code{nil}. | |
10013 | |
10014 In summary, when a Lisp variable is set to a value, it is provided with | |
10015 the address of the list to which the variable refers. | |
10016 | |
10017 @node Symbols as Chest, List Exercise, Lists diagrammed, List Implementation | |
10018 @section Symbols as a Chest of Drawers | |
10019 @cindex Symbols as a Chest of Drawers | |
10020 @cindex Chest of Drawers, metaphor for a symbol | |
10021 @cindex Drawers, Chest of, metaphor for a symbol | |
10022 | |
10023 In an earlier section, I suggested that you might imagine a symbol as | |
10024 being a chest of drawers. The function definition is put in one | |
10025 drawer, the value in another, and so on. What is put in the drawer | |
10026 holding the value can be changed without affecting the contents of the | |
10027 drawer holding the function definition, and vice-verse. | |
10028 | |
10029 Actually, what is put in each drawer is the address of the value or | |
10030 function definition. It is as if you found an old chest in the attic, | |
10031 and in one of its drawers you found a map giving you directions to | |
10032 where the buried treasure lies. | |
10033 | |
10034 (In addition to its name, symbol definition, and variable value, a | |
10035 symbol has a `drawer' for a @dfn{property list} which can be used to | |
10036 record other information. Property lists are not discussed here; see | |
10037 @ref{Property Lists, , Property Lists, elisp, The GNU Emacs Lisp | |
10038 Reference Manual}.) | |
10039 | |
10040 @need 1500 | |
10041 Here is a fanciful representation: | |
10042 | |
10043 @c chest-of-drawers diagram | |
10044 @ifnottex | |
10045 @sp 1 | |
10046 @smallexample | |
10047 @group | |
10048 Chest of Drawers Contents of Drawers | |
10049 | |
10050 __ o0O0o __ | |
10051 / \ | |
10052 --------------------- | |
10053 | directions to | [map to] | |
10054 | symbol name | bouquet | |
10055 | | | |
10056 +---------------------+ | |
10057 | directions to | | |
10058 | symbol definition | [none] | |
10059 | | | |
10060 +---------------------+ | |
10061 | directions to | [map to] | |
10062 | variable value | (rose violet buttercup) | |
10063 | | | |
10064 +---------------------+ | |
10065 | directions to | | |
10066 | property list | [not described here] | |
10067 | | | |
10068 +---------------------+ | |
10069 |/ \| | |
10070 @end group | |
10071 @end smallexample | |
10072 @sp 1 | |
10073 @end ifnottex | |
10074 @ifset print-postscript-figures | |
10075 @sp 1 | |
10076 @tex | |
10077 @center @image{drawers} | |
10078 %%%% old method of including an image | |
10079 % \input /usr/local/lib/tex/inputs/psfig.tex | |
10080 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/drawers.eps}} | |
10081 % \catcode`\@=0 % | |
10082 @end tex | |
10083 @sp 1 | |
10084 @end ifset | |
10085 @ifclear print-postscript-figures | |
10086 @iftex | |
10087 @sp 1 | |
10088 @smallexample | |
10089 @group | |
10090 Chest of Drawers Contents of Drawers | |
10091 | |
10092 __ o0O0o __ | |
10093 / \ | |
10094 --------------------- | |
10095 | directions to | [map to] | |
10096 | symbol name | bouquet | |
10097 | | | |
10098 +---------------------+ | |
10099 | directions to | | |
10100 | symbol definition | [none] | |
10101 | | | |
10102 +---------------------+ | |
10103 | directions to | [map to] | |
10104 | variable value | (rose violet buttercup) | |
10105 | | | |
10106 +---------------------+ | |
10107 | directions to | | |
10108 | property list | [not described here] | |
10109 | | | |
10110 +---------------------+ | |
10111 |/ \| | |
10112 @end group | |
10113 @end smallexample | |
10114 @sp 1 | |
10115 @end iftex | |
10116 @end ifclear | |
10117 | |
10118 @node List Exercise, , Symbols as Chest, List Implementation | |
10119 @section Exercise | |
10120 | |
10121 Set @code{flowers} to @code{violet} and @code{buttercup}. Cons two | |
10122 more flowers on to this list and set this new list to | |
10123 @code{more-flowers}. Set the @sc{car} of @code{flowers} to a fish. | |
10124 What does the @code{more-flowers} list now contain? | |
10125 | |
10126 @node Yanking, Loops & Recursion, List Implementation, Top | |
10127 @comment node-name, next, previous, up | |
10128 @chapter Yanking Text Back | |
10129 @findex yank | |
10130 @cindex Text retrieval | |
10131 @cindex Retrieving text | |
10132 @cindex Pasting text | |
10133 | |
10134 Whenever you cut text out of a buffer with a `kill' command in GNU Emacs, | |
10135 you can bring it back with a `yank' command. The text that is cut out of | |
10136 the buffer is put in the kill ring and the yank commands insert the | |
10137 appropriate contents of the kill ring back into a buffer (not necessarily | |
10138 the original buffer). | |
10139 | |
10140 A simple @kbd{C-y} (@code{yank}) command inserts the first item from | |
10141 the kill ring into the current buffer. If the @kbd{C-y} command is | |
10142 followed immediately by @kbd{M-y}, the first element is replaced by | |
10143 the second element. Successive @kbd{M-y} commands replace the second | |
10144 element with the third, fourth, or fifth element, and so on. When the | |
10145 last element in the kill ring is reached, it is replaced by the first | |
10146 element and the cycle is repeated. (Thus the kill ring is called a | |
10147 `ring' rather than just a `list'. However, the actual data structure | |
10148 that holds the text is a list. | |
10149 @xref{Kill Ring, , Handling the Kill Ring}, for the details of how the | |
10150 list is handled as a ring.) | |
10151 | |
10152 @menu | |
10153 * Kill Ring Overview:: | |
10154 * kill-ring-yank-pointer:: The kill ring is a list. | |
10155 * yank nthcdr Exercises:: The @code{kill-ring-yank-pointer} variable. | |
10156 @end menu | |
10157 | |
10158 @node Kill Ring Overview, kill-ring-yank-pointer, Yanking, Yanking | |
10159 @comment node-name, next, previous, up | |
10160 @section Kill Ring Overview | |
10161 @cindex Kill ring overview | |
10162 | |
10163 The kill ring is a list of textual strings. This is what it looks like: | |
10164 | |
10165 @smallexample | |
10166 ("some text" "a different piece of text" "yet more text") | |
10167 @end smallexample | |
10168 | |
10169 If this were the contents of my kill ring and I pressed @kbd{C-y}, the | |
10170 string of characters saying @samp{some text} would be inserted in this | |
10171 buffer where my cursor is located. | |
10172 | |
10173 The @code{yank} command is also used for duplicating text by copying it. | |
10174 The copied text is not cut from the buffer, but a copy of it is put on the | |
10175 kill ring and is inserted by yanking it back. | |
10176 | |
10177 Three functions are used for bringing text back from the kill ring: | |
10178 @code{yank}, which is usually bound to @kbd{C-y}; @code{yank-pop}, | |
10179 which is usually bound to @kbd{M-y}; and @code{rotate-yank-pointer}, | |
10180 which is used by the two other functions. | |
10181 | |
10182 These functions refer to the kill ring through a variable called the | |
10183 @code{kill-ring-yank-pointer}. Indeed, the insertion code for both the | |
10184 @code{yank} and @code{yank-pop} functions is: | |
10185 | |
10186 @smallexample | |
10187 (insert (car kill-ring-yank-pointer)) | |
10188 @end smallexample | |
10189 | |
10190 @noindent | |
10191 (Well, no more. In GNU Emacs 22, the function has been replaced by | |
10192 @code{insert-for-yank} which calls @code{insert-for-yank-1} | |
10193 repetitively for each @code{yank-handler} segment. In turn, | |
10194 @code{insert-for-yank-1} strips text properties from the inserted text | |
10195 according to @code{yank-excluded-properties}. Otherwise, it is just | |
10196 like @code{insert}. We will stick with plain @code{insert} since it | |
10197 is easier to understand.) | |
10198 | |
10199 To begin to understand how @code{yank} and @code{yank-pop} work, it is | |
10200 first necessary to look at the @code{kill-ring-yank-pointer} variable. | |
10201 | |
10202 @node kill-ring-yank-pointer, yank nthcdr Exercises, Kill Ring Overview, Yanking | |
10203 @comment node-name, next, previous, up | |
10204 @section The @code{kill-ring-yank-pointer} Variable | |
10205 | |
10206 @code{kill-ring-yank-pointer} is a variable, just as @code{kill-ring} is | |
10207 a variable. It points to something by being bound to the value of what | |
10208 it points to, like any other Lisp variable. | |
10209 | |
10210 @need 1000 | |
10211 Thus, if the value of the kill ring is: | |
10212 | |
10213 @smallexample | |
10214 ("some text" "a different piece of text" "yet more text") | |
10215 @end smallexample | |
10216 | |
10217 @need 1250 | |
10218 @noindent | |
10219 and the @code{kill-ring-yank-pointer} points to the second clause, the | |
10220 value of @code{kill-ring-yank-pointer} is: | |
10221 | |
10222 @smallexample | |
10223 ("a different piece of text" "yet more text") | |
10224 @end smallexample | |
10225 | |
10226 As explained in the previous chapter (@pxref{List Implementation}), the | |
10227 computer does not keep two different copies of the text being pointed to | |
10228 by both the @code{kill-ring} and the @code{kill-ring-yank-pointer}. The | |
10229 words ``a different piece of text'' and ``yet more text'' are not | |
10230 duplicated. Instead, the two Lisp variables point to the same pieces of | |
10231 text. Here is a diagram: | |
10232 | |
10233 @c cons-cell-diagram #5 | |
10234 @ifnottex | |
10235 @smallexample | |
10236 @group | |
10237 kill-ring kill-ring-yank-pointer | |
10238 | | | |
10239 | ___ ___ | ___ ___ ___ ___ | |
10240 ---> | | | --> | | | | | | | |
10241 |___|___|----> |___|___|--> |___|___|--> nil | |
10242 | | | | |
10243 | | | | |
10244 | | --> "yet more text" | |
10245 | | | |
10246 | --> "a different piece of text" | |
10247 | | |
10248 --> "some text" | |
10249 @end group | |
10250 @end smallexample | |
10251 @sp 1 | |
10252 @end ifnottex | |
10253 @ifset print-postscript-figures | |
10254 @sp 1 | |
10255 @tex | |
10256 @center @image{cons-5} | |
10257 %%%% old method of including an image | |
10258 % \input /usr/local/lib/tex/inputs/psfig.tex | |
10259 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/cons-5.eps}} | |
10260 % \catcode`\@=0 % | |
10261 @end tex | |
10262 @sp 1 | |
10263 @end ifset | |
10264 @ifclear print-postscript-figures | |
10265 @iftex | |
10266 @smallexample | |
10267 @group | |
10268 kill-ring kill-ring-yank-pointer | |
10269 | | | |
10270 | ___ ___ | ___ ___ ___ ___ | |
10271 ---> | | | --> | | | | | | | |
10272 |___|___|----> |___|___|--> |___|___|--> nil | |
10273 | | | | |
10274 | | | | |
10275 | | --> "yet more text" | |
10276 | | | |
10277 | --> "a different piece of text | |
10278 | | |
10279 --> "some text" | |
10280 @end group | |
10281 @end smallexample | |
10282 @sp 1 | |
10283 @end iftex | |
10284 @end ifclear | |
10285 | |
10286 Both the variable @code{kill-ring} and the variable | |
10287 @code{kill-ring-yank-pointer} are pointers. But the kill ring itself is | |
10288 usually described as if it were actually what it is composed of. The | |
10289 @code{kill-ring} is spoken of as if it were the list rather than that it | |
10290 points to the list. Conversely, the @code{kill-ring-yank-pointer} is | |
10291 spoken of as pointing to a list. | |
10292 | |
10293 These two ways of talking about the same thing sound confusing at first but | |
10294 make sense on reflection. The kill ring is generally thought of as the | |
10295 complete structure of data that holds the information of what has recently | |
10296 been cut out of the Emacs buffers. The @code{kill-ring-yank-pointer} | |
10297 on the other hand, serves to indicate---that is, to `point to'---that part | |
10298 of the kill ring of which the first element (the @sc{car}) will be | |
10299 inserted. | |
10300 | |
10301 @ignore | |
10302 In GNU Emacs 22, the @code{kill-new} function calls | |
10303 | |
10304 @code{(setq kill-ring-yank-pointer kill-ring)} | |
10305 | |
10306 (defun rotate-yank-pointer (arg) | |
10307 "Rotate the yanking point in the kill ring. | |
10308 With argument, rotate that many kills forward (or backward, if negative)." | |
10309 (interactive "p") | |
10310 (current-kill arg)) | |
10311 | |
10312 (defun current-kill (n &optional do-not-move) | |
10313 "Rotate the yanking point by N places, and then return that kill. | |
10314 If N is zero, `interprogram-paste-function' is set, and calling it | |
10315 returns a string, then that string is added to the front of the | |
10316 kill ring and returned as the latest kill. | |
10317 If optional arg DO-NOT-MOVE is non-nil, then don't actually move the | |
10318 yanking point; just return the Nth kill forward." | |
10319 (let ((interprogram-paste (and (= n 0) | |
10320 interprogram-paste-function | |
10321 (funcall interprogram-paste-function)))) | |
10322 (if interprogram-paste | |
10323 (progn | |
10324 ;; Disable the interprogram cut function when we add the new | |
10325 ;; text to the kill ring, so Emacs doesn't try to own the | |
10326 ;; selection, with identical text. | |
10327 (let ((interprogram-cut-function nil)) | |
10328 (kill-new interprogram-paste)) | |
10329 interprogram-paste) | |
10330 (or kill-ring (error "Kill ring is empty")) | |
10331 (let ((ARGth-kill-element | |
10332 (nthcdr (mod (- n (length kill-ring-yank-pointer)) | |
10333 (length kill-ring)) | |
10334 kill-ring))) | |
10335 (or do-not-move | |
10336 (setq kill-ring-yank-pointer ARGth-kill-element)) | |
10337 (car ARGth-kill-element))))) | |
10338 | |
10339 @end ignore | |
10340 | |
10341 @need 1500 | |
10342 @node yank nthcdr Exercises, , kill-ring-yank-pointer, Yanking | |
10343 @section Exercises with @code{yank} and @code{nthcdr} | |
10344 | |
10345 @itemize @bullet | |
10346 @item | |
10347 Using @kbd{C-h v} (@code{describe-variable}), look at the value of | |
10348 your kill ring. Add several items to your kill ring; look at its | |
10349 value again. Using @kbd{M-y} (@code{yank-pop)}, move all the way | |
10350 around the kill ring. How many items were in your kill ring? Find | |
10351 the value of @code{kill-ring-max}. Was your kill ring full, or could | |
10352 you have kept more blocks of text within it? | |
10353 | |
10354 @item | |
10355 Using @code{nthcdr} and @code{car}, construct a series of expressions | |
10356 to return the first, second, third, and fourth elements of a list. | |
10357 @end itemize | |
10358 | |
10359 @node Loops & Recursion, Regexp Search, Yanking, Top | |
10360 @comment node-name, next, previous, up | |
10361 @chapter Loops and Recursion | |
10362 @cindex Loops and recursion | |
10363 @cindex Recursion and loops | |
10364 @cindex Repetition (loops) | |
10365 | |
10366 Emacs Lisp has two primary ways to cause an expression, or a series of | |
10367 expressions, to be evaluated repeatedly: one uses a @code{while} | |
10368 loop, and the other uses @dfn{recursion}. | |
10369 | |
10370 Repetition can be very valuable. For example, to move forward four | |
10371 sentences, you need only write a program that will move forward one | |
10372 sentence and then repeat the process four times. Since a computer does | |
10373 not get bored or tired, such repetitive action does not have the | |
10374 deleterious effects that excessive or the wrong kinds of repetition can | |
10375 have on humans. | |
10376 | |
10377 People mostly write Emacs Lisp functions using @code{while} loops and | |
10378 their kin; but you can use recursion, which provides a very powerful | |
10379 way to think about and then to solve problems@footnote{You can write | |
10380 recursive functions to be frugal or wasteful of mental or computer | |
10381 resources; as it happens, methods that people find easy---that are | |
10382 frugal of `mental resources'---sometimes use considerable computer | |
10383 resources. Emacs was designed to run on machines that we now consider | |
10384 limited and its default settings are conservative. You may want to | |
10385 increase the values of @code{max-specpdl-size} and | |
10386 @code{max-lisp-eval-depth}. In my @file{.emacs} file, I set them to | |
10387 15 and 30 times their default value.}. | |
10388 | |
10389 @menu | |
10390 * while:: Causing a stretch of code to repeat. | |
10391 * dolist dotimes:: | |
10392 * Recursion:: Causing a function to call itself. | |
10393 * Looping exercise:: | |
10394 @end menu | |
10395 | |
10396 @node while, dolist dotimes, Loops & Recursion, Loops & Recursion | |
10397 @comment node-name, next, previous, up | |
10398 @section @code{while} | |
10399 @cindex Loops | |
10400 @findex while | |
10401 | |
10402 The @code{while} special form tests whether the value returned by | |
10403 evaluating its first argument is true or false. This is similar to what | |
10404 the Lisp interpreter does with an @code{if}; what the interpreter does | |
10405 next, however, is different. | |
10406 | |
10407 In a @code{while} expression, if the value returned by evaluating the | |
10408 first argument is false, the Lisp interpreter skips the rest of the | |
10409 expression (the @dfn{body} of the expression) and does not evaluate it. | |
10410 However, if the value is true, the Lisp interpreter evaluates the body | |
10411 of the expression and then again tests whether the first argument to | |
10412 @code{while} is true or false. If the value returned by evaluating the | |
10413 first argument is again true, the Lisp interpreter again evaluates the | |
10414 body of the expression. | |
10415 | |
10416 @need 1200 | |
10417 The template for a @code{while} expression looks like this: | |
10418 | |
10419 @smallexample | |
10420 @group | |
10421 (while @var{true-or-false-test} | |
10422 @var{body}@dots{}) | |
10423 @end group | |
10424 @end smallexample | |
10425 | |
10426 @menu | |
10427 * Looping with while:: Repeat so long as test returns true. | |
10428 * Loop Example:: A @code{while} loop that uses a list. | |
10429 * print-elements-of-list:: Uses @code{while}, @code{car}, @code{cdr}. | |
10430 * Incrementing Loop:: A loop with an incrementing counter. | |
10431 * Incrementing Loop Details:: | |
10432 * Decrementing Loop:: A loop with a decrementing counter. | |
10433 @end menu | |
10434 | |
10435 @node Looping with while, Loop Example, while, while | |
10436 @ifnottex | |
10437 @unnumberedsubsec Looping with @code{while} | |
10438 @end ifnottex | |
10439 | |
10440 So long as the true-or-false-test of the @code{while} expression | |
10441 returns a true value when it is evaluated, the body is repeatedly | |
10442 evaluated. This process is called a loop since the Lisp interpreter | |
10443 repeats the same thing again and again, like an airplane doing a loop. | |
10444 When the result of evaluating the true-or-false-test is false, the | |
10445 Lisp interpreter does not evaluate the rest of the @code{while} | |
10446 expression and `exits the loop'. | |
10447 | |
10448 Clearly, if the value returned by evaluating the first argument to | |
10449 @code{while} is always true, the body following will be evaluated | |
10450 again and again @dots{} and again @dots{} forever. Conversely, if the | |
10451 value returned is never true, the expressions in the body will never | |
10452 be evaluated. The craft of writing a @code{while} loop consists of | |
10453 choosing a mechanism such that the true-or-false-test returns true | |
10454 just the number of times that you want the subsequent expressions to | |
10455 be evaluated, and then have the test return false. | |
10456 | |
10457 The value returned by evaluating a @code{while} is the value of the | |
10458 true-or-false-test. An interesting consequence of this is that a | |
10459 @code{while} loop that evaluates without error will return @code{nil} | |
10460 or false regardless of whether it has looped 1 or 100 times or none at | |
10461 all. A @code{while} expression that evaluates successfully never | |
10462 returns a true value! What this means is that @code{while} is always | |
10463 evaluated for its side effects, which is to say, the consequences of | |
10464 evaluating the expressions within the body of the @code{while} loop. | |
10465 This makes sense. It is not the mere act of looping that is desired, | |
10466 but the consequences of what happens when the expressions in the loop | |
10467 are repeatedly evaluated. | |
10468 | |
10469 @node Loop Example, print-elements-of-list, Looping with while, while | |
10470 @comment node-name, next, previous, up | |
10471 @subsection A @code{while} Loop and a List | |
10472 | |
10473 A common way to control a @code{while} loop is to test whether a list | |
10474 has any elements. If it does, the loop is repeated; but if it does not, | |
10475 the repetition is ended. Since this is an important technique, we will | |
10476 create a short example to illustrate it. | |
10477 | |
10478 A simple way to test whether a list has elements is to evaluate the | |
10479 list: if it has no elements, it is an empty list and will return the | |
10480 empty list, @code{()}, which is a synonym for @code{nil} or false. On | |
10481 the other hand, a list with elements will return those elements when it | |
10482 is evaluated. Since Emacs Lisp considers as true any value that is not | |
10483 @code{nil}, a list that returns elements will test true in a | |
10484 @code{while} loop. | |
10485 | |
10486 @need 1200 | |
10487 For example, you can set the variable @code{empty-list} to @code{nil} by | |
10488 evaluating the following @code{setq} expression: | |
10489 | |
10490 @smallexample | |
10491 (setq empty-list ()) | |
10492 @end smallexample | |
10493 | |
10494 @noindent | |
10495 After evaluating the @code{setq} expression, you can evaluate the | |
10496 variable @code{empty-list} in the usual way, by placing the cursor after | |
10497 the symbol and typing @kbd{C-x C-e}; @code{nil} will appear in your | |
10498 echo area: | |
10499 | |
10500 @smallexample | |
10501 empty-list | |
10502 @end smallexample | |
10503 | |
10504 On the other hand, if you set a variable to be a list with elements, the | |
10505 list will appear when you evaluate the variable, as you can see by | |
10506 evaluating the following two expressions: | |
10507 | |
10508 @smallexample | |
10509 @group | |
10510 (setq animals '(gazelle giraffe lion tiger)) | |
10511 | |
10512 animals | |
10513 @end group | |
10514 @end smallexample | |
10515 | |
10516 Thus, to create a @code{while} loop that tests whether there are any | |
10517 items in the list @code{animals}, the first part of the loop will be | |
10518 written like this: | |
10519 | |
10520 @smallexample | |
10521 @group | |
10522 (while animals | |
10523 @dots{} | |
10524 @end group | |
10525 @end smallexample | |
10526 | |
10527 @noindent | |
10528 When the @code{while} tests its first argument, the variable | |
10529 @code{animals} is evaluated. It returns a list. So long as the list | |
10530 has elements, the @code{while} considers the results of the test to be | |
10531 true; but when the list is empty, it considers the results of the test | |
10532 to be false. | |
10533 | |
10534 To prevent the @code{while} loop from running forever, some mechanism | |
10535 needs to be provided to empty the list eventually. An oft-used | |
10536 technique is to have one of the subsequent forms in the @code{while} | |
10537 expression set the value of the list to be the @sc{cdr} of the list. | |
10538 Each time the @code{cdr} function is evaluated, the list will be made | |
10539 shorter, until eventually only the empty list will be left. At this | |
10540 point, the test of the @code{while} loop will return false, and the | |
10541 arguments to the @code{while} will no longer be evaluated. | |
10542 | |
10543 For example, the list of animals bound to the variable @code{animals} | |
10544 can be set to be the @sc{cdr} of the original list with the | |
10545 following expression: | |
10546 | |
10547 @smallexample | |
10548 (setq animals (cdr animals)) | |
10549 @end smallexample | |
10550 | |
10551 @noindent | |
10552 If you have evaluated the previous expressions and then evaluate this | |
10553 expression, you will see @code{(giraffe lion tiger)} appear in the echo | |
10554 area. If you evaluate the expression again, @code{(lion tiger)} will | |
10555 appear in the echo area. If you evaluate it again and yet again, | |
10556 @code{(tiger)} appears and then the empty list, shown by @code{nil}. | |
10557 | |
10558 A template for a @code{while} loop that uses the @code{cdr} function | |
10559 repeatedly to cause the true-or-false-test eventually to test false | |
10560 looks like this: | |
10561 | |
10562 @smallexample | |
10563 @group | |
10564 (while @var{test-whether-list-is-empty} | |
10565 @var{body}@dots{} | |
10566 @var{set-list-to-cdr-of-list}) | |
10567 @end group | |
10568 @end smallexample | |
10569 | |
10570 This test and use of @code{cdr} can be put together in a function that | |
10571 goes through a list and prints each element of the list on a line of its | |
10572 own. | |
10573 | |
10574 @node print-elements-of-list, Incrementing Loop, Loop Example, while | |
10575 @subsection An Example: @code{print-elements-of-list} | |
10576 @findex print-elements-of-list | |
10577 | |
10578 The @code{print-elements-of-list} function illustrates a @code{while} | |
10579 loop with a list. | |
10580 | |
10581 @cindex @file{*scratch*} buffer | |
10582 The function requires several lines for its output. If you are | |
10583 reading this in a recent instance of GNU Emacs, | |
10584 @c GNU Emacs 21, GNU Emacs 22, or a later version, | |
10585 you can evaluate the following expression inside of Info, as usual. | |
10586 | |
10587 If you are using an earlier version of Emacs, you need to copy the | |
10588 necessary expressions to your @file{*scratch*} buffer and evaluate | |
10589 them there. This is because the echo area had only one line in the | |
10590 earlier versions. | |
10591 | |
10592 You can copy the expressions by marking the beginning of the region | |
10593 with @kbd{C-@key{SPC}} (@code{set-mark-command}), moving the cursor to | |
10594 the end of the region and then copying the region using @kbd{M-w} | |
10595 (@code{kill-ring-save}, which calls @code{copy-region-as-kill} and | |
10596 then provides visual feedback). In the @file{*scratch*} | |
10597 buffer, you can yank the expressions back by typing @kbd{C-y} | |
10598 (@code{yank}). | |
10599 | |
10600 After you have copied the expressions to the @file{*scratch*} buffer, | |
10601 evaluate each expression in turn. Be sure to evaluate the last | |
10602 expression, @code{(print-elements-of-list animals)}, by typing | |
10603 @kbd{C-u C-x C-e}, that is, by giving an argument to | |
10604 @code{eval-last-sexp}. This will cause the result of the evaluation | |
10605 to be printed in the @file{*scratch*} buffer instead of being printed | |
10606 in the echo area. (Otherwise you will see something like this in your | |
10607 echo area: @code{^Jgazelle^J^Jgiraffe^J^Jlion^J^Jtiger^Jnil}, in which | |
10608 each @samp{^J} stands for a `newline'.) | |
10609 | |
10610 @need 1500 | |
10611 In a recent instance of GNU Emacs, you can evaluate these expressions | |
10612 directly in the Info buffer, and the echo area will grow to show the | |
10613 results. | |
10614 | |
10615 @smallexample | |
10616 @group | |
10617 (setq animals '(gazelle giraffe lion tiger)) | |
10618 | |
10619 (defun print-elements-of-list (list) | |
10620 "Print each element of LIST on a line of its own." | |
10621 (while list | |
10622 (print (car list)) | |
10623 (setq list (cdr list)))) | |
10624 | |
10625 (print-elements-of-list animals) | |
10626 @end group | |
10627 @end smallexample | |
10628 | |
10629 @need 1200 | |
10630 @noindent | |
10631 When you evaluate the three expressions in sequence, you will see | |
10632 this: | |
10633 | |
10634 @smallexample | |
10635 @group | |
10636 gazelle | |
10637 | |
10638 giraffe | |
10639 | |
10640 lion | |
10641 | |
10642 tiger | |
10643 nil | |
10644 @end group | |
10645 @end smallexample | |
10646 | |
10647 Each element of the list is printed on a line of its own (that is what | |
10648 the function @code{print} does) and then the value returned by the | |
10649 function is printed. Since the last expression in the function is the | |
10650 @code{while} loop, and since @code{while} loops always return | |
10651 @code{nil}, a @code{nil} is printed after the last element of the list. | |
10652 | |
10653 @node Incrementing Loop, Incrementing Loop Details, print-elements-of-list, while | |
10654 @comment node-name, next, previous, up | |
10655 @subsection A Loop with an Incrementing Counter | |
10656 | |
10657 A loop is not useful unless it stops when it ought. Besides | |
10658 controlling a loop with a list, a common way of stopping a loop is to | |
10659 write the first argument as a test that returns false when the correct | |
10660 number of repetitions are complete. This means that the loop must | |
10661 have a counter---an expression that counts how many times the loop | |
10662 repeats itself. | |
10663 | |
10664 @node Incrementing Loop Details, Decrementing Loop, Incrementing Loop, while | |
10665 @ifnottex | |
10666 @unnumberedsubsec Details of an Incrementing Loop | |
10667 @end ifnottex | |
10668 | |
10669 The test for a loop with an incrementing counter can be an expression | |
10670 such as @code{(< count desired-number)} which returns @code{t} for | |
10671 true if the value of @code{count} is less than the | |
10672 @code{desired-number} of repetitions and @code{nil} for false if the | |
10673 value of @code{count} is equal to or is greater than the | |
10674 @code{desired-number}. The expression that increments the count can | |
10675 be a simple @code{setq} such as @code{(setq count (1+ count))}, where | |
10676 @code{1+} is a built-in function in Emacs Lisp that adds 1 to its | |
10677 argument. (The expression @w{@code{(1+ count)}} has the same result | |
10678 as @w{@code{(+ count 1)}}, but is easier for a human to read.) | |
10679 | |
10680 @need 1250 | |
10681 The template for a @code{while} loop controlled by an incrementing | |
10682 counter looks like this: | |
10683 | |
10684 @smallexample | |
10685 @group | |
10686 @var{set-count-to-initial-value} | |
10687 (while (< count desired-number) ; @r{true-or-false-test} | |
10688 @var{body}@dots{} | |
10689 (setq count (1+ count))) ; @r{incrementer} | |
10690 @end group | |
10691 @end smallexample | |
10692 | |
10693 @noindent | |
10694 Note that you need to set the initial value of @code{count}; usually it | |
10695 is set to 1. | |
10696 | |
10697 @menu | |
10698 * Incrementing Example:: Counting pebbles in a triangle. | |
10699 * Inc Example parts:: The parts of the function definition. | |
10700 * Inc Example altogether:: Putting the function definition together. | |
10701 @end menu | |
10702 | |
10703 @node Incrementing Example, Inc Example parts, Incrementing Loop Details, Incrementing Loop Details | |
10704 @unnumberedsubsubsec Example with incrementing counter | |
10705 | |
10706 Suppose you are playing on the beach and decide to make a triangle of | |
10707 pebbles, putting one pebble in the first row, two in the second row, | |
10708 three in the third row and so on, like this: | |
10709 | |
10710 @sp 1 | |
10711 @c pebble diagram | |
10712 @ifnottex | |
10713 @smallexample | |
10714 @group | |
10715 * | |
10716 * * | |
10717 * * * | |
10718 * * * * | |
10719 @end group | |
10720 @end smallexample | |
10721 @end ifnottex | |
10722 @iftex | |
10723 @smallexample | |
10724 @group | |
10725 @bullet{} | |
10726 @bullet{} @bullet{} | |
10727 @bullet{} @bullet{} @bullet{} | |
10728 @bullet{} @bullet{} @bullet{} @bullet{} | |
10729 @end group | |
10730 @end smallexample | |
10731 @end iftex | |
10732 @sp 1 | |
10733 | |
10734 @noindent | |
10735 (About 2500 years ago, Pythagoras and others developed the beginnings of | |
10736 number theory by considering questions such as this.) | |
10737 | |
10738 Suppose you want to know how many pebbles you will need to make a | |
10739 triangle with 7 rows? | |
10740 | |
10741 Clearly, what you need to do is add up the numbers from 1 to 7. There | |
10742 are two ways to do this; start with the smallest number, one, and add up | |
10743 the list in sequence, 1, 2, 3, 4 and so on; or start with the largest | |
10744 number and add the list going down: 7, 6, 5, 4 and so on. Because both | |
10745 mechanisms illustrate common ways of writing @code{while} loops, we will | |
10746 create two examples, one counting up and the other counting down. In | |
10747 this first example, we will start with 1 and add 2, 3, 4 and so on. | |
10748 | |
10749 If you are just adding up a short list of numbers, the easiest way to do | |
10750 it is to add up all the numbers at once. However, if you do not know | |
10751 ahead of time how many numbers your list will have, or if you want to be | |
10752 prepared for a very long list, then you need to design your addition so | |
10753 that what you do is repeat a simple process many times instead of doing | |
10754 a more complex process once. | |
10755 | |
10756 For example, instead of adding up all the pebbles all at once, what you | |
10757 can do is add the number of pebbles in the first row, 1, to the number | |
10758 in the second row, 2, and then add the total of those two rows to the | |
10759 third row, 3. Then you can add the number in the fourth row, 4, to the | |
10760 total of the first three rows; and so on. | |
10761 | |
10762 The critical characteristic of the process is that each repetitive | |
10763 action is simple. In this case, at each step we add only two numbers, | |
10764 the number of pebbles in the row and the total already found. This | |
10765 process of adding two numbers is repeated again and again until the last | |
10766 row has been added to the total of all the preceding rows. In a more | |
10767 complex loop the repetitive action might not be so simple, but it will | |
10768 be simpler than doing everything all at once. | |
10769 | |
10770 @node Inc Example parts, Inc Example altogether, Incrementing Example, Incrementing Loop Details | |
10771 @unnumberedsubsubsec The parts of the function definition | |
10772 | |
10773 The preceding analysis gives us the bones of our function definition: | |
10774 first, we will need a variable that we can call @code{total} that will | |
10775 be the total number of pebbles. This will be the value returned by | |
10776 the function. | |
10777 | |
10778 Second, we know that the function will require an argument: this | |
10779 argument will be the total number of rows in the triangle. It can be | |
10780 called @code{number-of-rows}. | |
10781 | |
10782 Finally, we need a variable to use as a counter. We could call this | |
10783 variable @code{counter}, but a better name is @code{row-number}. That | |
10784 is because what the counter does in this function is count rows, and a | |
10785 program should be written to be as understandable as possible. | |
10786 | |
10787 When the Lisp interpreter first starts evaluating the expressions in the | |
10788 function, the value of @code{total} should be set to zero, since we have | |
10789 not added anything to it. Then the function should add the number of | |
10790 pebbles in the first row to the total, and then add the number of | |
10791 pebbles in the second to the total, and then add the number of | |
10792 pebbles in the third row to the total, and so on, until there are no | |
10793 more rows left to add. | |
10794 | |
10795 Both @code{total} and @code{row-number} are used only inside the | |
10796 function, so they can be declared as local variables with @code{let} | |
10797 and given initial values. Clearly, the initial value for @code{total} | |
10798 should be 0. The initial value of @code{row-number} should be 1, | |
10799 since we start with the first row. This means that the @code{let} | |
10800 statement will look like this: | |
10801 | |
10802 @smallexample | |
10803 @group | |
10804 (let ((total 0) | |
10805 (row-number 1)) | |
10806 @var{body}@dots{}) | |
10807 @end group | |
10808 @end smallexample | |
10809 | |
10810 After the internal variables are declared and bound to their initial | |
10811 values, we can begin the @code{while} loop. The expression that serves | |
10812 as the test should return a value of @code{t} for true so long as the | |
10813 @code{row-number} is less than or equal to the @code{number-of-rows}. | |
10814 (If the expression tests true only so long as the row number is less | |
10815 than the number of rows in the triangle, the last row will never be | |
10816 added to the total; hence the row number has to be either less than or | |
10817 equal to the number of rows.) | |
10818 | |
10819 @need 1500 | |
10820 @findex <= @r{(less than or equal)} | |
10821 Lisp provides the @code{<=} function that returns true if the value of | |
10822 its first argument is less than or equal to the value of its second | |
10823 argument and false otherwise. So the expression that the @code{while} | |
10824 will evaluate as its test should look like this: | |
10825 | |
10826 @smallexample | |
10827 (<= row-number number-of-rows) | |
10828 @end smallexample | |
10829 | |
10830 The total number of pebbles can be found by repeatedly adding the number | |
10831 of pebbles in a row to the total already found. Since the number of | |
10832 pebbles in the row is equal to the row number, the total can be found by | |
10833 adding the row number to the total. (Clearly, in a more complex | |
10834 situation, the number of pebbles in the row might be related to the row | |
10835 number in a more complicated way; if this were the case, the row number | |
10836 would be replaced by the appropriate expression.) | |
10837 | |
10838 @smallexample | |
10839 (setq total (+ total row-number)) | |
10840 @end smallexample | |
10841 | |
10842 @noindent | |
10843 What this does is set the new value of @code{total} to be equal to the | |
10844 sum of adding the number of pebbles in the row to the previous total. | |
10845 | |
10846 After setting the value of @code{total}, the conditions need to be | |
10847 established for the next repetition of the loop, if there is one. This | |
10848 is done by incrementing the value of the @code{row-number} variable, | |
10849 which serves as a counter. After the @code{row-number} variable has | |
10850 been incremented, the true-or-false-test at the beginning of the | |
10851 @code{while} loop tests whether its value is still less than or equal to | |
10852 the value of the @code{number-of-rows} and if it is, adds the new value | |
10853 of the @code{row-number} variable to the @code{total} of the previous | |
10854 repetition of the loop. | |
10855 | |
10856 @need 1200 | |
10857 The built-in Emacs Lisp function @code{1+} adds 1 to a number, so the | |
10858 @code{row-number} variable can be incremented with this expression: | |
10859 | |
10860 @smallexample | |
10861 (setq row-number (1+ row-number)) | |
10862 @end smallexample | |
10863 | |
10864 @node Inc Example altogether, , Inc Example parts, Incrementing Loop Details | |
10865 @unnumberedsubsubsec Putting the function definition together | |
10866 | |
10867 We have created the parts for the function definition; now we need to | |
10868 put them together. | |
10869 | |
10870 @need 800 | |
10871 First, the contents of the @code{while} expression: | |
10872 | |
10873 @smallexample | |
10874 @group | |
10875 (while (<= row-number number-of-rows) ; @r{true-or-false-test} | |
10876 (setq total (+ total row-number)) | |
10877 (setq row-number (1+ row-number))) ; @r{incrementer} | |
10878 @end group | |
10879 @end smallexample | |
10880 | |
10881 Along with the @code{let} expression varlist, this very nearly | |
10882 completes the body of the function definition. However, it requires | |
10883 one final element, the need for which is somewhat subtle. | |
10884 | |
10885 The final touch is to place the variable @code{total} on a line by | |
10886 itself after the @code{while} expression. Otherwise, the value returned | |
10887 by the whole function is the value of the last expression that is | |
10888 evaluated in the body of the @code{let}, and this is the value | |
10889 returned by the @code{while}, which is always @code{nil}. | |
10890 | |
10891 This may not be evident at first sight. It almost looks as if the | |
10892 incrementing expression is the last expression of the whole function. | |
10893 But that expression is part of the body of the @code{while}; it is the | |
10894 last element of the list that starts with the symbol @code{while}. | |
10895 Moreover, the whole of the @code{while} loop is a list within the body | |
10896 of the @code{let}. | |
10897 | |
10898 @need 1250 | |
10899 In outline, the function will look like this: | |
10900 | |
10901 @smallexample | |
10902 @group | |
10903 (defun @var{name-of-function} (@var{argument-list}) | |
10904 "@var{documentation}@dots{}" | |
10905 (let (@var{varlist}) | |
10906 (while (@var{true-or-false-test}) | |
10907 @var{body-of-while}@dots{} ) | |
10908 @dots{} )) ; @r{Need final expression here.} | |
10909 @end group | |
10910 @end smallexample | |
10911 | |
10912 The result of evaluating the @code{let} is what is going to be returned | |
10913 by the @code{defun} since the @code{let} is not embedded within any | |
10914 containing list, except for the @code{defun} as a whole. However, if | |
10915 the @code{while} is the last element of the @code{let} expression, the | |
10916 function will always return @code{nil}. This is not what we want! | |
10917 Instead, what we want is the value of the variable @code{total}. This | |
10918 is returned by simply placing the symbol as the last element of the list | |
10919 starting with @code{let}. It gets evaluated after the preceding | |
10920 elements of the list are evaluated, which means it gets evaluated after | |
10921 it has been assigned the correct value for the total. | |
10922 | |
10923 It may be easier to see this by printing the list starting with | |
10924 @code{let} all on one line. This format makes it evident that the | |
10925 @var{varlist} and @code{while} expressions are the second and third | |
10926 elements of the list starting with @code{let}, and the @code{total} is | |
10927 the last element: | |
10928 | |
10929 @smallexample | |
10930 @group | |
10931 (let (@var{varlist}) (while (@var{true-or-false-test}) @var{body-of-while}@dots{} ) total) | |
10932 @end group | |
10933 @end smallexample | |
10934 | |
10935 @need 1200 | |
10936 Putting everything together, the @code{triangle} function definition | |
10937 looks like this: | |
10938 | |
10939 @smallexample | |
10940 @group | |
10941 (defun triangle (number-of-rows) ; @r{Version with} | |
10942 ; @r{ incrementing counter.} | |
10943 "Add up the number of pebbles in a triangle. | |
10944 The first row has one pebble, the second row two pebbles, | |
10945 the third row three pebbles, and so on. | |
10946 The argument is NUMBER-OF-ROWS." | |
10947 @end group | |
10948 @group | |
10949 (let ((total 0) | |
10950 (row-number 1)) | |
10951 (while (<= row-number number-of-rows) | |
10952 (setq total (+ total row-number)) | |
10953 (setq row-number (1+ row-number))) | |
10954 total)) | |
10955 @end group | |
10956 @end smallexample | |
10957 | |
10958 @need 1200 | |
10959 After you have installed @code{triangle} by evaluating the function, you | |
10960 can try it out. Here are two examples: | |
10961 | |
10962 @smallexample | |
10963 @group | |
10964 (triangle 4) | |
10965 | |
10966 (triangle 7) | |
10967 @end group | |
10968 @end smallexample | |
10969 | |
10970 @noindent | |
10971 The sum of the first four numbers is 10 and the sum of the first seven | |
10972 numbers is 28. | |
10973 | |
10974 @node Decrementing Loop, , Incrementing Loop Details, while | |
10975 @comment node-name, next, previous, up | |
10976 @subsection Loop with a Decrementing Counter | |
10977 | |
10978 Another common way to write a @code{while} loop is to write the test | |
10979 so that it determines whether a counter is greater than zero. So long | |
10980 as the counter is greater than zero, the loop is repeated. But when | |
10981 the counter is equal to or less than zero, the loop is stopped. For | |
10982 this to work, the counter has to start out greater than zero and then | |
10983 be made smaller and smaller by a form that is evaluated | |
10984 repeatedly. | |
10985 | |
10986 The test will be an expression such as @code{(> counter 0)} which | |
10987 returns @code{t} for true if the value of @code{counter} is greater | |
10988 than zero, and @code{nil} for false if the value of @code{counter} is | |
10989 equal to or less than zero. The expression that makes the number | |
10990 smaller and smaller can be a simple @code{setq} such as @code{(setq | |
10991 counter (1- counter))}, where @code{1-} is a built-in function in | |
10992 Emacs Lisp that subtracts 1 from its argument. | |
10993 | |
10994 @need 1250 | |
10995 The template for a decrementing @code{while} loop looks like this: | |
10996 | |
10997 @smallexample | |
10998 @group | |
10999 (while (> counter 0) ; @r{true-or-false-test} | |
11000 @var{body}@dots{} | |
11001 (setq counter (1- counter))) ; @r{decrementer} | |
11002 @end group | |
11003 @end smallexample | |
11004 | |
11005 @menu | |
11006 * Decrementing Example:: More pebbles on the beach. | |
11007 * Dec Example parts:: The parts of the function definition. | |
11008 * Dec Example altogether:: Putting the function definition together. | |
11009 @end menu | |
11010 | |
11011 @node Decrementing Example, Dec Example parts, Decrementing Loop, Decrementing Loop | |
11012 @unnumberedsubsubsec Example with decrementing counter | |
11013 | |
11014 To illustrate a loop with a decrementing counter, we will rewrite the | |
11015 @code{triangle} function so the counter decreases to zero. | |
11016 | |
11017 This is the reverse of the earlier version of the function. In this | |
11018 case, to find out how many pebbles are needed to make a triangle with | |
11019 3 rows, add the number of pebbles in the third row, 3, to the number | |
11020 in the preceding row, 2, and then add the total of those two rows to | |
11021 the row that precedes them, which is 1. | |
11022 | |
11023 Likewise, to find the number of pebbles in a triangle with 7 rows, add | |
11024 the number of pebbles in the seventh row, 7, to the number in the | |
11025 preceding row, which is 6, and then add the total of those two rows to | |
11026 the row that precedes them, which is 5, and so on. As in the previous | |
11027 example, each addition only involves adding two numbers, the total of | |
11028 the rows already added up and the number of pebbles in the row that is | |
11029 being added to the total. This process of adding two numbers is | |
11030 repeated again and again until there are no more pebbles to add. | |
11031 | |
11032 We know how many pebbles to start with: the number of pebbles in the | |
11033 last row is equal to the number of rows. If the triangle has seven | |
11034 rows, the number of pebbles in the last row is 7. Likewise, we know how | |
11035 many pebbles are in the preceding row: it is one less than the number in | |
11036 the row. | |
11037 | |
11038 @node Dec Example parts, Dec Example altogether, Decrementing Example, Decrementing Loop | |
11039 @unnumberedsubsubsec The parts of the function definition | |
11040 | |
11041 We start with three variables: the total number of rows in the | |
11042 triangle; the number of pebbles in a row; and the total number of | |
11043 pebbles, which is what we want to calculate. These variables can be | |
11044 named @code{number-of-rows}, @code{number-of-pebbles-in-row}, and | |
11045 @code{total}, respectively. | |
11046 | |
11047 Both @code{total} and @code{number-of-pebbles-in-row} are used only | |
11048 inside the function and are declared with @code{let}. The initial | |
11049 value of @code{total} should, of course, be zero. However, the | |
11050 initial value of @code{number-of-pebbles-in-row} should be equal to | |
11051 the number of rows in the triangle, since the addition will start with | |
11052 the longest row. | |
11053 | |
11054 @need 1250 | |
11055 This means that the beginning of the @code{let} expression will look | |
11056 like this: | |
11057 | |
11058 @smallexample | |
11059 @group | |
11060 (let ((total 0) | |
11061 (number-of-pebbles-in-row number-of-rows)) | |
11062 @var{body}@dots{}) | |
11063 @end group | |
11064 @end smallexample | |
11065 | |
11066 The total number of pebbles can be found by repeatedly adding the number | |
11067 of pebbles in a row to the total already found, that is, by repeatedly | |
11068 evaluating the following expression: | |
11069 | |
11070 @smallexample | |
11071 (setq total (+ total number-of-pebbles-in-row)) | |
11072 @end smallexample | |
11073 | |
11074 @noindent | |
11075 After the @code{number-of-pebbles-in-row} is added to the @code{total}, | |
11076 the @code{number-of-pebbles-in-row} should be decremented by one, since | |
11077 the next time the loop repeats, the preceding row will be | |
11078 added to the total. | |
11079 | |
11080 The number of pebbles in a preceding row is one less than the number of | |
11081 pebbles in a row, so the built-in Emacs Lisp function @code{1-} can be | |
11082 used to compute the number of pebbles in the preceding row. This can be | |
11083 done with the following expression: | |
11084 | |
11085 @smallexample | |
11086 @group | |
11087 (setq number-of-pebbles-in-row | |
11088 (1- number-of-pebbles-in-row)) | |
11089 @end group | |
11090 @end smallexample | |
11091 | |
11092 Finally, we know that the @code{while} loop should stop making repeated | |
11093 additions when there are no pebbles in a row. So the test for | |
11094 the @code{while} loop is simply: | |
11095 | |
11096 @smallexample | |
11097 (while (> number-of-pebbles-in-row 0) | |
11098 @end smallexample | |
11099 | |
11100 @node Dec Example altogether, , Dec Example parts, Decrementing Loop | |
11101 @unnumberedsubsubsec Putting the function definition together | |
11102 | |
11103 We can put these expressions together to create a function definition | |
11104 that works. However, on examination, we find that one of the local | |
11105 variables is unneeded! | |
11106 | |
11107 @need 1250 | |
11108 The function definition looks like this: | |
11109 | |
11110 @smallexample | |
11111 @group | |
11112 ;;; @r{First subtractive version.} | |
11113 (defun triangle (number-of-rows) | |
11114 "Add up the number of pebbles in a triangle." | |
11115 (let ((total 0) | |
11116 (number-of-pebbles-in-row number-of-rows)) | |
11117 (while (> number-of-pebbles-in-row 0) | |
11118 (setq total (+ total number-of-pebbles-in-row)) | |
11119 (setq number-of-pebbles-in-row | |
11120 (1- number-of-pebbles-in-row))) | |
11121 total)) | |
11122 @end group | |
11123 @end smallexample | |
11124 | |
11125 As written, this function works. | |
11126 | |
11127 However, we do not need @code{number-of-pebbles-in-row}. | |
11128 | |
11129 @cindex Argument as local variable | |
11130 When the @code{triangle} function is evaluated, the symbol | |
11131 @code{number-of-rows} will be bound to a number, giving it an initial | |
11132 value. That number can be changed in the body of the function as if | |
11133 it were a local variable, without any fear that such a change will | |
11134 effect the value of the variable outside of the function. This is a | |
11135 very useful characteristic of Lisp; it means that the variable | |
11136 @code{number-of-rows} can be used anywhere in the function where | |
11137 @code{number-of-pebbles-in-row} is used. | |
11138 | |
11139 @need 800 | |
11140 Here is a second version of the function written a bit more cleanly: | |
11141 | |
11142 @smallexample | |
11143 @group | |
11144 (defun triangle (number) ; @r{Second version.} | |
11145 "Return sum of numbers 1 through NUMBER inclusive." | |
11146 (let ((total 0)) | |
11147 (while (> number 0) | |
11148 (setq total (+ total number)) | |
11149 (setq number (1- number))) | |
11150 total)) | |
11151 @end group | |
11152 @end smallexample | |
11153 | |
11154 In brief, a properly written @code{while} loop will consist of three parts: | |
11155 | |
11156 @enumerate | |
11157 @item | |
11158 A test that will return false after the loop has repeated itself the | |
11159 correct number of times. | |
11160 | |
11161 @item | |
11162 An expression the evaluation of which will return the value desired | |
11163 after being repeatedly evaluated. | |
11164 | |
11165 @item | |
11166 An expression to change the value passed to the true-or-false-test so | |
11167 that the test returns false after the loop has repeated itself the right | |
11168 number of times. | |
11169 @end enumerate | |
11170 | |
11171 @node dolist dotimes, Recursion, while, Loops & Recursion | |
11172 @comment node-name, next, previous, up | |
11173 @section Save your time: @code{dolist} and @code{dotimes} | |
11174 | |
11175 In addition to @code{while}, both @code{dolist} and @code{dotimes} | |
11176 provide for looping. Sometimes these are quicker to write than the | |
11177 equivalent @code{while} loop. Both are Lisp macros. (@xref{Macros, , | |
11178 Macros, elisp, The GNU Emacs Lisp Reference Manual}. ) | |
11179 | |
11180 @code{dolist} works like a @code{while} loop that `@sc{cdr}s down a | |
11181 list': @code{dolist} automatically shortens the list each time it | |
11182 loops---takes the @sc{cdr} of the list---and binds the @sc{car} of | |
11183 each shorter version of the list to the first of its arguments. | |
11184 | |
11185 @code{dotimes} loops a specific number of times: you specify the number. | |
11186 | |
11187 @menu | |
11188 * dolist:: | |
11189 * dotimes:: | |
11190 @end menu | |
11191 | |
11192 @node dolist, dotimes, dolist dotimes, dolist dotimes | |
11193 @unnumberedsubsubsec The @code{dolist} Macro | |
11194 @findex dolist | |
11195 | |
11196 Suppose, for example, you want to reverse a list, so that | |
11197 ``first'' ``second'' ``third'' becomes ``third'' ``second'' ``first''. | |
11198 | |
11199 @need 1250 | |
11200 In practice, you would use the @code{reverse} function, like this: | |
11201 | |
11202 @smallexample | |
11203 @group | |
11204 (setq animals '(gazelle giraffe lion tiger)) | |
11205 | |
11206 (reverse animals) | |
11207 @end group | |
11208 @end smallexample | |
11209 | |
11210 @need 800 | |
11211 @noindent | |
11212 Here is how you could reverse the list using a @code{while} loop: | |
11213 | |
11214 @smallexample | |
11215 @group | |
11216 (setq animals '(gazelle giraffe lion tiger)) | |
11217 | |
11218 (defun reverse-list-with-while (list) | |
11219 "Using while, reverse the order of LIST." | |
11220 (let (value) ; make sure list starts empty | |
11221 (while list | |
11222 (setq value (cons (car list) value)) | |
11223 (setq list (cdr list))) | |
11224 value)) | |
11225 | |
11226 (reverse-list-with-while animals) | |
11227 @end group | |
11228 @end smallexample | |
11229 | |
11230 @need 800 | |
11231 @noindent | |
11232 And here is how you could use the @code{dolist} macro: | |
11233 | |
11234 @smallexample | |
11235 @group | |
11236 (setq animals '(gazelle giraffe lion tiger)) | |
11237 | |
11238 (defun reverse-list-with-dolist (list) | |
11239 "Using dolist, reverse the order of LIST." | |
11240 (let (value) ; make sure list starts empty | |
11241 (dolist (element list value) | |
11242 (setq value (cons element value))))) | |
11243 | |
11244 (reverse-list-with-dolist animals) | |
11245 @end group | |
11246 @end smallexample | |
11247 | |
11248 @need 1250 | |
11249 @noindent | |
11250 In Info, you can place your cursor after the closing parenthesis of | |
11251 each expression and type @kbd{C-x C-e}; in each case, you should see | |
11252 | |
11253 @smallexample | |
11254 (tiger lion giraffe gazelle) | |
11255 @end smallexample | |
11256 | |
11257 @noindent | |
11258 in the echo area. | |
11259 | |
11260 For this example, the existing @code{reverse} function is obviously best. | |
11261 The @code{while} loop is just like our first example (@pxref{Loop | |
11262 Example, , A @code{while} Loop and a List}). The @code{while} first | |
11263 checks whether the list has elements; if so, it constructs a new list | |
11264 by adding the first element of the list to the existing list (which in | |
11265 the first iteration of the loop is @code{nil}). Since the second | |
11266 element is prepended in front of the first element, and the third | |
11267 element is prepended in front of the second element, the list is reversed. | |
11268 | |
11269 In the expression using a @code{while} loop, | |
11270 the @w{@code{(setq list (cdr list))}} | |
11271 expression shortens the list, so the @code{while} loop eventually | |
11272 stops. In addition, it provides the @code{cons} expression with a new | |
11273 first element by creating a new and shorter list at each repetition of | |
11274 the loop. | |
11275 | |
11276 The @code{dolist} expression does very much the same as the | |
11277 @code{while} expression, except that the @code{dolist} macro does some | |
11278 of the work you have to do when writing a @code{while} expression. | |
11279 | |
11280 Like a @code{while} loop, a @code{dolist} loops. What is different is | |
11281 that it automatically shortens the list each time it loops --- it | |
11282 `@sc{cdr}s down the list' on its own --- and it automatically binds | |
11283 the @sc{car} of each shorter version of the list to the first of its | |
11284 arguments. | |
11285 | |
11286 In the example, the @sc{car} of each shorter version of the list is | |
11287 referred to using the symbol @samp{element}, the list itself is called | |
11288 @samp{list}, and the value returned is called @samp{value}. The | |
11289 remainder of the @code{dolist} expression is the body. | |
11290 | |
11291 The @code{dolist} expression binds the @sc{car} of each shorter | |
11292 version of the list to @code{element} and then evaluates the body of | |
11293 the expression; and repeats the loop. The result is returned in | |
11294 @code{value}. | |
11295 | |
11296 @node dotimes, , dolist, dolist dotimes | |
11297 @unnumberedsubsubsec The @code{dotimes} Macro | |
11298 @findex dotimes | |
11299 | |
11300 The @code{dotimes} macro is similar to @code{dolist}, except that it | |
11301 loops a specific number of times. | |
11302 | |
11303 The first argument to @code{dotimes} is assigned the numbers 0, 1, 2 | |
11304 and so forth each time around the loop, and the value of the third | |
11305 argument is returned. You need to provide the value of the second | |
11306 argument, which is how many times the macro loops. | |
11307 | |
11308 @need 1250 | |
11309 For example, the following binds the numbers from 0 up to, but not | |
11310 including, the number 3 to the first argument, @var{number}, and then | |
11311 constructs a list of the three numbers. (The first number is 0, the | |
11312 second number is 1, and the third number is 2; this makes a total of | |
11313 three numbers in all, starting with zero as the first number.) | |
11314 | |
11315 @smallexample | |
11316 @group | |
11317 (let (value) ; otherwise a value is a void variable | |
11318 (dotimes (number 3 value) | |
11319 (setq value (cons number value)))) | |
11320 | |
11321 @result{} (2 1 0) | |
11322 @end group | |
11323 @end smallexample | |
11324 | |
11325 @noindent | |
11326 @code{dotimes} returns @code{value}, so the way to use | |
11327 @code{dotimes} is to operate on some expression @var{number} number of | |
11328 times and then return the result, either as a list or an atom. | |
11329 | |
11330 @need 1250 | |
11331 Here is an example of a @code{defun} that uses @code{dotimes} to add | |
11332 up the number of pebbles in a triangle. | |
11333 | |
11334 @smallexample | |
11335 @group | |
11336 (defun triangle-using-dotimes (number-of-rows) | |
11337 "Using dotimes, add up the number of pebbles in a triangle." | |
11338 (let ((total 0)) ; otherwise a total is a void variable | |
11339 (dotimes (number number-of-rows total) | |
11340 (setq total (+ total (1+ number)))))) | |
11341 | |
11342 (triangle-using-dotimes 4) | |
11343 @end group | |
11344 @end smallexample | |
11345 | |
11346 @node Recursion, Looping exercise, dolist dotimes, Loops & Recursion | |
11347 @comment node-name, next, previous, up | |
11348 @section Recursion | |
11349 @cindex Recursion | |
11350 | |
11351 A recursive function contains code that tells the Lisp interpreter to | |
11352 call a program that runs exactly like itself, but with slightly | |
11353 different arguments. The code runs exactly the same because it has | |
11354 the same name. However, even though the program has the same name, it | |
11355 is not the same entity. It is different. In the jargon, it is a | |
11356 different `instance'. | |
11357 | |
11358 Eventually, if the program is written correctly, the `slightly | |
11359 different arguments' will become sufficiently different from the first | |
11360 arguments that the final instance will stop. | |
11361 | |
11362 @menu | |
11363 * Building Robots:: Same model, different serial number ... | |
11364 * Recursive Definition Parts:: Walk until you stop ... | |
11365 * Recursion with list:: Using a list as the test whether to recurse. | |
11366 * Recursive triangle function:: | |
11367 * Recursion with cond:: | |
11368 * Recursive Patterns:: Often used templates. | |
11369 * No Deferment:: Don't store up work ... | |
11370 * No deferment solution:: | |
11371 @end menu | |
11372 | |
11373 @node Building Robots, Recursive Definition Parts, Recursion, Recursion | |
11374 @comment node-name, next, previous, up | |
11375 @subsection Building Robots: Extending the Metaphor | |
11376 @cindex Building robots | |
11377 @cindex Robots, building | |
11378 | |
11379 It is sometimes helpful to think of a running program as a robot that | |
11380 does a job. In doing its job, a recursive function calls on a second | |
11381 robot to help it. The second robot is identical to the first in every | |
11382 way, except that the second robot helps the first and has been | |
11383 passed different arguments than the first. | |
11384 | |
11385 In a recursive function, the second robot may call a third; and the | |
11386 third may call a fourth, and so on. Each of these is a different | |
11387 entity; but all are clones. | |
11388 | |
11389 Since each robot has slightly different instructions---the arguments | |
11390 will differ from one robot to the next---the last robot should know | |
11391 when to stop. | |
11392 | |
11393 Let's expand on the metaphor in which a computer program is a robot. | |
11394 | |
11395 A function definition provides the blueprints for a robot. When you | |
11396 install a function definition, that is, when you evaluate a | |
11397 @code{defun} special form, you install the necessary equipment to | |
11398 build robots. It is as if you were in a factory, setting up an | |
11399 assembly line. Robots with the same name are built according to the | |
11400 same blueprints. So they have, as it were, the same `model number', | |
11401 but a different `serial number'. | |
11402 | |
11403 We often say that a recursive function `calls itself'. What we mean | |
11404 is that the instructions in a recursive function cause the Lisp | |
11405 interpreter to run a different function that has the same name and | |
11406 does the same job as the first, but with different arguments. | |
11407 | |
11408 It is important that the arguments differ from one instance to the | |
11409 next; otherwise, the process will never stop. | |
11410 | |
11411 @node Recursive Definition Parts, Recursion with list, Building Robots, Recursion | |
11412 @comment node-name, next, previous, up | |
11413 @subsection The Parts of a Recursive Definition | |
11414 @cindex Parts of a Recursive Definition | |
11415 @cindex Recursive Definition Parts | |
11416 | |
11417 A recursive function typically contains a conditional expression which | |
11418 has three parts: | |
11419 | |
11420 @enumerate | |
11421 @item | |
11422 A true-or-false-test that determines whether the function is called | |
11423 again, here called the @dfn{do-again-test}. | |
11424 | |
11425 @item | |
11426 The name of the function. When this name is called, a new instance of | |
11427 the function---a new robot, as it were---is created and told what to do. | |
11428 | |
11429 @item | |
11430 An expression that returns a different value each time the function is | |
11431 called, here called the @dfn{next-step-expression}. Consequently, the | |
11432 argument (or arguments) passed to the new instance of the function | |
11433 will be different from that passed to the previous instance. This | |
11434 causes the conditional expression, the @dfn{do-again-test}, to test | |
11435 false after the correct number of repetitions. | |
11436 @end enumerate | |
11437 | |
11438 Recursive functions can be much simpler than any other kind of | |
11439 function. Indeed, when people first start to use them, they often look | |
11440 so mysteriously simple as to be incomprehensible. Like riding a | |
11441 bicycle, reading a recursive function definition takes a certain knack | |
11442 which is hard at first but then seems simple. | |
11443 | |
11444 @need 1200 | |
11445 There are several different common recursive patterns. A very simple | |
11446 pattern looks like this: | |
11447 | |
11448 @smallexample | |
11449 @group | |
11450 (defun @var{name-of-recursive-function} (@var{argument-list}) | |
11451 "@var{documentation}@dots{}" | |
11452 (if @var{do-again-test} | |
11453 @var{body}@dots{} | |
11454 (@var{name-of-recursive-function} | |
11455 @var{next-step-expression}))) | |
11456 @end group | |
11457 @end smallexample | |
11458 | |
11459 Each time a recursive function is evaluated, a new instance of it is | |
11460 created and told what to do. The arguments tell the instance what to do. | |
11461 | |
11462 An argument is bound to the value of the next-step-expression. Each | |
11463 instance runs with a different value of the next-step-expression. | |
11464 | |
11465 The value in the next-step-expression is used in the do-again-test. | |
11466 | |
11467 The value returned by the next-step-expression is passed to the new | |
11468 instance of the function, which evaluates it (or some | |
11469 transmogrification of it) to determine whether to continue or stop. | |
11470 The next-step-expression is designed so that the do-again-test returns | |
11471 false when the function should no longer be repeated. | |
11472 | |
11473 The do-again-test is sometimes called the @dfn{stop condition}, | |
11474 since it stops the repetitions when it tests false. | |
11475 | |
11476 @node Recursion with list, Recursive triangle function, Recursive Definition Parts, Recursion | |
11477 @comment node-name, next, previous, up | |
11478 @subsection Recursion with a List | |
11479 | |
11480 The example of a @code{while} loop that printed the elements of a list | |
11481 of numbers can be written recursively. Here is the code, including | |
11482 an expression to set the value of the variable @code{animals} to a list. | |
11483 | |
11484 If you are using GNU Emacs 20 or before, this example must be copied | |
11485 to the @file{*scratch*} buffer and each expression must be evaluated | |
11486 there. Use @kbd{C-u C-x C-e} to evaluate the | |
11487 @code{(print-elements-recursively animals)} expression so that the | |
11488 results are printed in the buffer; otherwise the Lisp interpreter will | |
11489 try to squeeze the results into the one line of the echo area. | |
11490 | |
11491 Also, place your cursor immediately after the last closing parenthesis | |
11492 of the @code{print-elements-recursively} function, before the comment. | |
11493 Otherwise, the Lisp interpreter will try to evaluate the comment. | |
11494 | |
11495 If you are using a more recent version of Emacs, you can evaluate this | |
11496 expression directly in Info. | |
11497 | |
11498 @findex print-elements-recursively | |
11499 @smallexample | |
11500 @group | |
11501 (setq animals '(gazelle giraffe lion tiger)) | |
11502 | |
11503 (defun print-elements-recursively (list) | |
11504 "Print each element of LIST on a line of its own. | |
11505 Uses recursion." | |
11506 (when list ; @r{do-again-test} | |
11507 (print (car list)) ; @r{body} | |
11508 (print-elements-recursively ; @r{recursive call} | |
11509 (cdr list)))) ; @r{next-step-expression} | |
11510 | |
11511 (print-elements-recursively animals) | |
11512 @end group | |
11513 @end smallexample | |
11514 | |
11515 The @code{print-elements-recursively} function first tests whether | |
11516 there is any content in the list; if there is, the function prints the | |
11517 first element of the list, the @sc{car} of the list. Then the | |
11518 function `invokes itself', but gives itself as its argument, not the | |
11519 whole list, but the second and subsequent elements of the list, the | |
11520 @sc{cdr} of the list. | |
11521 | |
11522 Put another way, if the list is not empty, the function invokes | |
11523 another instance of code that is similar to the initial code, but is a | |
11524 different thread of execution, with different arguments than the first | |
11525 instance. | |
11526 | |
11527 Put in yet another way, if the list is not empty, the first robot | |
11528 assemblies a second robot and tells it what to do; the second robot is | |
11529 a different individual from the first, but is the same model. | |
11530 | |
11531 When the second evaluation occurs, the @code{when} expression is | |
11532 evaluated and if true, prints the first element of the list it | |
11533 receives as its argument (which is the second element of the original | |
11534 list). Then the function `calls itself' with the @sc{cdr} of the list | |
11535 it is invoked with, which (the second time around) is the @sc{cdr} of | |
11536 the @sc{cdr} of the original list. | |
11537 | |
11538 Note that although we say that the function `calls itself', what we | |
11539 mean is that the Lisp interpreter assembles and instructs a new | |
11540 instance of the program. The new instance is a clone of the first, | |
11541 but is a separate individual. | |
11542 | |
11543 Each time the function `invokes itself', it invokes itself on a | |
11544 shorter version of the original list. It creates a new instance that | |
11545 works on a shorter list. | |
11546 | |
11547 Eventually, the function invokes itself on an empty list. It creates | |
11548 a new instance whose argument is @code{nil}. The conditional expression | |
11549 tests the value of @code{list}. Since the value of @code{list} is | |
11550 @code{nil}, the @code{when} expression tests false so the then-part is | |
11551 not evaluated. The function as a whole then returns @code{nil}. | |
11552 | |
11553 @need 1200 | |
11554 When you evaluate @code{(print-elements-recursively animals)} in the | |
11555 @file{*scratch*} buffer, you see this result: | |
11556 | |
11557 @smallexample | |
11558 @group | |
11559 gazelle | |
11560 | |
11561 giraffe | |
11562 | |
11563 lion | |
11564 | |
11565 tiger | |
11566 nil | |
11567 @end group | |
11568 @end smallexample | |
11569 | |
11570 @need 2000 | |
11571 @node Recursive triangle function, Recursion with cond, Recursion with list, Recursion | |
11572 @comment node-name, next, previous, up | |
11573 @subsection Recursion in Place of a Counter | |
11574 @findex triangle-recursively | |
11575 | |
11576 @need 1200 | |
11577 The @code{triangle} function described in a previous section can also | |
11578 be written recursively. It looks like this: | |
11579 | |
11580 @smallexample | |
11581 @group | |
11582 (defun triangle-recursively (number) | |
11583 "Return the sum of the numbers 1 through NUMBER inclusive. | |
11584 Uses recursion." | |
11585 (if (= number 1) ; @r{do-again-test} | |
11586 1 ; @r{then-part} | |
11587 (+ number ; @r{else-part} | |
11588 (triangle-recursively ; @r{recursive call} | |
11589 (1- number))))) ; @r{next-step-expression} | |
11590 | |
11591 (triangle-recursively 7) | |
11592 @end group | |
11593 @end smallexample | |
11594 | |
11595 @noindent | |
11596 You can install this function by evaluating it and then try it by | |
11597 evaluating @code{(triangle-recursively 7)}. (Remember to put your | |
11598 cursor immediately after the last parenthesis of the function | |
11599 definition, before the comment.) The function evaluates to 28. | |
11600 | |
11601 To understand how this function works, let's consider what happens in the | |
11602 various cases when the function is passed 1, 2, 3, or 4 as the value of | |
11603 its argument. | |
11604 | |
11605 @menu | |
11606 * Recursive Example arg of 1 or 2:: | |
11607 * Recursive Example arg of 3 or 4:: | |
11608 @end menu | |
11609 | |
11610 @node Recursive Example arg of 1 or 2, Recursive Example arg of 3 or 4, Recursive triangle function, Recursive triangle function | |
11611 @ifnottex | |
11612 @unnumberedsubsubsec An argument of 1 or 2 | |
11613 @end ifnottex | |
11614 | |
11615 First, what happens if the value of the argument is 1? | |
11616 | |
11617 The function has an @code{if} expression after the documentation | |
11618 string. It tests whether the value of @code{number} is equal to 1; if | |
11619 so, Emacs evaluates the then-part of the @code{if} expression, which | |
11620 returns the number 1 as the value of the function. (A triangle with | |
11621 one row has one pebble in it.) | |
11622 | |
11623 Suppose, however, that the value of the argument is 2. In this case, | |
11624 Emacs evaluates the else-part of the @code{if} expression. | |
11625 | |
11626 @need 1200 | |
11627 The else-part consists of an addition, the recursive call to | |
11628 @code{triangle-recursively} and a decrementing action; and it looks like | |
11629 this: | |
11630 | |
11631 @smallexample | |
11632 (+ number (triangle-recursively (1- number))) | |
11633 @end smallexample | |
11634 | |
11635 When Emacs evaluates this expression, the innermost expression is | |
11636 evaluated first; then the other parts in sequence. Here are the steps | |
11637 in detail: | |
11638 | |
11639 @table @i | |
11640 @item Step 1 @w{ } Evaluate the innermost expression. | |
11641 | |
11642 The innermost expression is @code{(1- number)} so Emacs decrements the | |
11643 value of @code{number} from 2 to 1. | |
11644 | |
11645 @item Step 2 @w{ } Evaluate the @code{triangle-recursively} function. | |
11646 | |
11647 The Lisp interpreter creates an individual instance of | |
11648 @code{triangle-recursively}. It does not matter that this function is | |
11649 contained within itself. Emacs passes the result Step 1 as the | |
11650 argument used by this instance of the @code{triangle-recursively} | |
11651 function | |
11652 | |
11653 In this case, Emacs evaluates @code{triangle-recursively} with an | |
11654 argument of 1. This means that this evaluation of | |
11655 @code{triangle-recursively} returns 1. | |
11656 | |
11657 @item Step 3 @w{ } Evaluate the value of @code{number}. | |
11658 | |
11659 The variable @code{number} is the second element of the list that | |
11660 starts with @code{+}; its value is 2. | |
11661 | |
11662 @item Step 4 @w{ } Evaluate the @code{+} expression. | |
11663 | |
11664 The @code{+} expression receives two arguments, the first | |
11665 from the evaluation of @code{number} (Step 3) and the second from the | |
11666 evaluation of @code{triangle-recursively} (Step 2). | |
11667 | |
11668 The result of the addition is the sum of 2 plus 1, and the number 3 is | |
11669 returned, which is correct. A triangle with two rows has three | |
11670 pebbles in it. | |
11671 @end table | |
11672 | |
11673 @node Recursive Example arg of 3 or 4, , Recursive Example arg of 1 or 2, Recursive triangle function | |
11674 @unnumberedsubsubsec An argument of 3 or 4 | |
11675 | |
11676 Suppose that @code{triangle-recursively} is called with an argument of | |
11677 3. | |
11678 | |
11679 @table @i | |
11680 @item Step 1 @w{ } Evaluate the do-again-test. | |
11681 | |
11682 The @code{if} expression is evaluated first. This is the do-again | |
11683 test and returns false, so the else-part of the @code{if} expression | |
11684 is evaluated. (Note that in this example, the do-again-test causes | |
11685 the function to call itself when it tests false, not when it tests | |
11686 true.) | |
11687 | |
11688 @item Step 2 @w{ } Evaluate the innermost expression of the else-part. | |
11689 | |
11690 The innermost expression of the else-part is evaluated, which decrements | |
11691 3 to 2. This is the next-step-expression. | |
11692 | |
11693 @item Step 3 @w{ } Evaluate the @code{triangle-recursively} function. | |
11694 | |
11695 The number 2 is passed to the @code{triangle-recursively} function. | |
11696 | |
11697 We know what happens when Emacs evaluates @code{triangle-recursively} with | |
11698 an argument of 2. After going through the sequence of actions described | |
11699 earlier, it returns a value of 3. So that is what will happen here. | |
11700 | |
11701 @item Step 4 @w{ } Evaluate the addition. | |
11702 | |
11703 3 will be passed as an argument to the addition and will be added to the | |
11704 number with which the function was called, which is 3. | |
11705 @end table | |
11706 | |
11707 @noindent | |
11708 The value returned by the function as a whole will be 6. | |
11709 | |
11710 Now that we know what will happen when @code{triangle-recursively} is | |
11711 called with an argument of 3, it is evident what will happen if it is | |
11712 called with an argument of 4: | |
11713 | |
11714 @quotation | |
11715 @need 800 | |
11716 In the recursive call, the evaluation of | |
11717 | |
11718 @smallexample | |
11719 (triangle-recursively (1- 4)) | |
11720 @end smallexample | |
11721 | |
11722 @need 800 | |
11723 @noindent | |
11724 will return the value of evaluating | |
11725 | |
11726 @smallexample | |
11727 (triangle-recursively 3) | |
11728 @end smallexample | |
11729 | |
11730 @noindent | |
11731 which is 6 and this value will be added to 4 by the addition in the | |
11732 third line. | |
11733 @end quotation | |
11734 | |
11735 @noindent | |
11736 The value returned by the function as a whole will be 10. | |
11737 | |
11738 Each time @code{triangle-recursively} is evaluated, it evaluates a | |
11739 version of itself---a different instance of itself---with a smaller | |
11740 argument, until the argument is small enough so that it does not | |
11741 evaluate itself. | |
11742 | |
11743 Note that this particular design for a recursive function | |
11744 requires that operations be deferred. | |
11745 | |
11746 Before @code{(triangle-recursively 7)} can calculate its answer, it | |
11747 must call @code{(triangle-recursively 6)}; and before | |
11748 @code{(triangle-recursively 6)} can calculate its answer, it must call | |
11749 @code{(triangle-recursively 5)}; and so on. That is to say, the | |
11750 calculation that @code{(triangle-recursively 7)} makes must be | |
11751 deferred until @code{(triangle-recursively 6)} makes its calculation; | |
11752 and @code{(triangle-recursively 6)} must defer until | |
11753 @code{(triangle-recursively 5)} completes; and so on. | |
11754 | |
11755 If each of these instances of @code{triangle-recursively} are thought | |
11756 of as different robots, the first robot must wait for the second to | |
11757 complete its job, which must wait until the third completes, and so | |
11758 on. | |
11759 | |
11760 There is a way around this kind of waiting, which we will discuss in | |
11761 @ref{No Deferment, , Recursion without Deferments}. | |
11762 | |
11763 @node Recursion with cond, Recursive Patterns, Recursive triangle function, Recursion | |
11764 @comment node-name, next, previous, up | |
11765 @subsection Recursion Example Using @code{cond} | |
11766 @findex cond | |
11767 | |
11768 The version of @code{triangle-recursively} described earlier is written | |
11769 with the @code{if} special form. It can also be written using another | |
11770 special form called @code{cond}. The name of the special form | |
11771 @code{cond} is an abbreviation of the word @samp{conditional}. | |
11772 | |
11773 Although the @code{cond} special form is not used as often in the | |
11774 Emacs Lisp sources as @code{if}, it is used often enough to justify | |
11775 explaining it. | |
11776 | |
11777 @need 800 | |
11778 The template for a @code{cond} expression looks like this: | |
11779 | |
11780 @smallexample | |
11781 @group | |
11782 (cond | |
11783 @var{body}@dots{}) | |
11784 @end group | |
11785 @end smallexample | |
11786 | |
11787 @noindent | |
11788 where the @var{body} is a series of lists. | |
11789 | |
11790 @need 800 | |
11791 Written out more fully, the template looks like this: | |
11792 | |
11793 @smallexample | |
11794 @group | |
11795 (cond | |
11796 (@var{first-true-or-false-test} @var{first-consequent}) | |
11797 (@var{second-true-or-false-test} @var{second-consequent}) | |
11798 (@var{third-true-or-false-test} @var{third-consequent}) | |
11799 @dots{}) | |
11800 @end group | |
11801 @end smallexample | |
11802 | |
11803 When the Lisp interpreter evaluates the @code{cond} expression, it | |
11804 evaluates the first element (the @sc{car} or true-or-false-test) of | |
11805 the first expression in a series of expressions within the body of the | |
11806 @code{cond}. | |
11807 | |
11808 If the true-or-false-test returns @code{nil} the rest of that | |
11809 expression, the consequent, is skipped and the true-or-false-test of the | |
11810 next expression is evaluated. When an expression is found whose | |
11811 true-or-false-test returns a value that is not @code{nil}, the | |
11812 consequent of that expression is evaluated. The consequent can be one | |
11813 or more expressions. If the consequent consists of more than one | |
11814 expression, the expressions are evaluated in sequence and the value of | |
11815 the last one is returned. If the expression does not have a consequent, | |
11816 the value of the true-or-false-test is returned. | |
11817 | |
11818 If none of the true-or-false-tests test true, the @code{cond} expression | |
11819 returns @code{nil}. | |
11820 | |
11821 @need 1250 | |
11822 Written using @code{cond}, the @code{triangle} function looks like this: | |
11823 | |
11824 @smallexample | |
11825 @group | |
11826 (defun triangle-using-cond (number) | |
11827 (cond ((<= number 0) 0) | |
11828 ((= number 1) 1) | |
11829 ((> number 1) | |
11830 (+ number (triangle-using-cond (1- number)))))) | |
11831 @end group | |
11832 @end smallexample | |
11833 | |
11834 @noindent | |
11835 In this example, the @code{cond} returns 0 if the number is less than or | |
11836 equal to 0, it returns 1 if the number is 1 and it evaluates @code{(+ | |
11837 number (triangle-using-cond (1- number)))} if the number is greater than | |
11838 1. | |
11839 | |
11840 @node Recursive Patterns, No Deferment, Recursion with cond, Recursion | |
11841 @comment node-name, next, previous, up | |
11842 @subsection Recursive Patterns | |
11843 @cindex Recursive Patterns | |
11844 | |
11845 Here are three common recursive patterns. Each involves a list. | |
11846 Recursion does not need to involve lists, but Lisp is designed for lists | |
11847 and this provides a sense of its primal capabilities. | |
11848 | |
11849 @menu | |
11850 * Every:: | |
11851 * Accumulate:: | |
11852 * Keep:: | |
11853 @end menu | |
11854 | |
11855 @node Every, Accumulate, Recursive Patterns, Recursive Patterns | |
11856 @comment node-name, next, previous, up | |
11857 @unnumberedsubsubsec Recursive Pattern: @emph{every} | |
11858 @cindex Every, type of recursive pattern | |
11859 @cindex Recursive pattern: every | |
11860 | |
11861 In the @code{every} recursive pattern, an action is performed on every | |
11862 element of a list. | |
11863 | |
11864 @need 1500 | |
11865 The basic pattern is: | |
11866 | |
11867 @itemize @bullet | |
11868 @item | |
11869 If a list be empty, return @code{nil}. | |
11870 @item | |
11871 Else, act on the beginning of the list (the @sc{car} of the list) | |
11872 @itemize @minus | |
11873 @item | |
11874 through a recursive call by the function on the rest (the | |
11875 @sc{cdr}) of the list, | |
11876 @item | |
11877 and, optionally, combine the acted-on element, using @code{cons}, | |
11878 with the results of acting on the rest. | |
11879 @end itemize | |
11880 @end itemize | |
11881 | |
11882 @need 1500 | |
11883 Here is example: | |
11884 | |
11885 @smallexample | |
11886 @group | |
11887 (defun square-each (numbers-list) | |
11888 "Square each of a NUMBERS LIST, recursively." | |
11889 (if (not numbers-list) ; do-again-test | |
11890 nil | |
11891 (cons | |
11892 (* (car numbers-list) (car numbers-list)) | |
11893 (square-each (cdr numbers-list))))) ; next-step-expression | |
11894 @end group | |
11895 | |
11896 @group | |
11897 (square-each '(1 2 3)) | |
11898 @result{} (1 4 9) | |
11899 @end group | |
11900 @end smallexample | |
11901 | |
11902 @need 1200 | |
11903 @noindent | |
11904 If @code{numbers-list} is empty, do nothing. But if it has content, | |
11905 construct a list combining the square of the first number in the list | |
11906 with the result of the recursive call. | |
11907 | |
11908 (The example follows the pattern exactly: @code{nil} is returned if | |
11909 the numbers' list is empty. In practice, you would write the | |
11910 conditional so it carries out the action when the numbers' list is not | |
11911 empty.) | |
11912 | |
11913 The @code{print-elements-recursively} function (@pxref{Recursion with | |
11914 list, , Recursion with a List}) is another example of an @code{every} | |
11915 pattern, except in this case, rather than bring the results together | |
11916 using @code{cons}, we print each element of output. | |
11917 | |
11918 @need 1250 | |
11919 The @code{print-elements-recursively} function looks like this: | |
11920 | |
11921 @smallexample | |
11922 @group | |
11923 (setq animals '(gazelle giraffe lion tiger)) | |
11924 @end group | |
11925 | |
11926 @group | |
11927 (defun print-elements-recursively (list) | |
11928 "Print each element of LIST on a line of its own. | |
11929 Uses recursion." | |
11930 (when list ; @r{do-again-test} | |
11931 (print (car list)) ; @r{body} | |
11932 (print-elements-recursively ; @r{recursive call} | |
11933 (cdr list)))) ; @r{next-step-expression} | |
11934 | |
11935 (print-elements-recursively animals) | |
11936 @end group | |
11937 @end smallexample | |
11938 | |
11939 @need 1500 | |
11940 The pattern for @code{print-elements-recursively} is: | |
11941 | |
11942 @itemize @bullet | |
11943 @item | |
11944 When the list is empty, do nothing. | |
11945 @item | |
11946 But when the list has at least one element, | |
11947 @itemize @minus | |
11948 @item | |
11949 act on the beginning of the list (the @sc{car} of the list), | |
11950 @item | |
11951 and make a recursive call on the rest (the @sc{cdr}) of the list. | |
11952 @end itemize | |
11953 @end itemize | |
11954 | |
11955 @node Accumulate, Keep, Every, Recursive Patterns | |
11956 @comment node-name, next, previous, up | |
11957 @unnumberedsubsubsec Recursive Pattern: @emph{accumulate} | |
11958 @cindex Accumulate, type of recursive pattern | |
11959 @cindex Recursive pattern: accumulate | |
11960 | |
11961 Another recursive pattern is called the @code{accumulate} pattern. In | |
11962 the @code{accumulate} recursive pattern, an action is performed on | |
11963 every element of a list and the result of that action is accumulated | |
11964 with the results of performing the action on the other elements. | |
11965 | |
11966 This is very like the `every' pattern using @code{cons}, except that | |
11967 @code{cons} is not used, but some other combiner. | |
11968 | |
11969 @need 1500 | |
11970 The pattern is: | |
11971 | |
11972 @itemize @bullet | |
11973 @item | |
11974 If a list be empty, return zero or some other constant. | |
11975 @item | |
11976 Else, act on the beginning of the list (the @sc{car} of the list), | |
11977 @itemize @minus | |
11978 @item | |
11979 and combine that acted-on element, using @code{+} or | |
11980 some other combining function, with | |
11981 @item | |
11982 a recursive call by the function on the rest (the @sc{cdr}) of the list. | |
11983 @end itemize | |
11984 @end itemize | |
11985 | |
11986 @need 1500 | |
11987 Here is an example: | |
11988 | |
11989 @smallexample | |
11990 @group | |
11991 (defun add-elements (numbers-list) | |
11992 "Add the elements of NUMBERS-LIST together." | |
11993 (if (not numbers-list) | |
11994 0 | |
11995 (+ (car numbers-list) (add-elements (cdr numbers-list))))) | |
11996 @end group | |
11997 | |
11998 @group | |
11999 (add-elements '(1 2 3 4)) | |
12000 @result{} 10 | |
12001 @end group | |
12002 @end smallexample | |
12003 | |
12004 @xref{Files List, , Making a List of Files}, for an example of the | |
12005 accumulate pattern. | |
12006 | |
12007 @node Keep, , Accumulate, Recursive Patterns | |
12008 @comment node-name, next, previous, up | |
12009 @unnumberedsubsubsec Recursive Pattern: @emph{keep} | |
12010 @cindex Keep, type of recursive pattern | |
12011 @cindex Recursive pattern: keep | |
12012 | |
12013 A third recursive pattern is called the @code{keep} pattern. | |
12014 In the @code{keep} recursive pattern, each element of a list is tested; | |
12015 the element is acted on and the results are kept only if the element | |
12016 meets a criterion. | |
12017 | |
12018 Again, this is very like the `every' pattern, except the element is | |
12019 skipped unless it meets a criterion. | |
12020 | |
12021 @need 1500 | |
12022 The pattern has three parts: | |
12023 | |
12024 @itemize @bullet | |
12025 @item | |
12026 If a list be empty, return @code{nil}. | |
12027 @item | |
12028 Else, if the beginning of the list (the @sc{car} of the list) passes | |
12029 a test | |
12030 @itemize @minus | |
12031 @item | |
12032 act on that element and combine it, using @code{cons} with | |
12033 @item | |
12034 a recursive call by the function on the rest (the @sc{cdr}) of the list. | |
12035 @end itemize | |
12036 @item | |
12037 Otherwise, if the beginning of the list (the @sc{car} of the list) fails | |
12038 the test | |
12039 @itemize @minus | |
12040 @item | |
12041 skip on that element, | |
12042 @item | |
12043 and, recursively call the function on the rest (the @sc{cdr}) of the list. | |
12044 @end itemize | |
12045 @end itemize | |
12046 | |
12047 @need 1500 | |
12048 Here is an example that uses @code{cond}: | |
12049 | |
12050 @smallexample | |
12051 @group | |
12052 (defun keep-three-letter-words (word-list) | |
12053 "Keep three letter words in WORD-LIST." | |
12054 (cond | |
12055 ;; First do-again-test: stop-condition | |
12056 ((not word-list) nil) | |
12057 | |
12058 ;; Second do-again-test: when to act | |
12059 ((eq 3 (length (symbol-name (car word-list)))) | |
12060 ;; combine acted-on element with recursive call on shorter list | |
12061 (cons (car word-list) (keep-three-letter-words (cdr word-list)))) | |
12062 | |
12063 ;; Third do-again-test: when to skip element; | |
12064 ;; recursively call shorter list with next-step expression | |
12065 (t (keep-three-letter-words (cdr word-list))))) | |
12066 @end group | |
12067 | |
12068 @group | |
12069 (keep-three-letter-words '(one two three four five six)) | |
12070 @result{} (one two six) | |
12071 @end group | |
12072 @end smallexample | |
12073 | |
12074 It goes without saying that you need not use @code{nil} as the test for | |
12075 when to stop; and you can, of course, combine these patterns. | |
12076 | |
12077 @node No Deferment, No deferment solution, Recursive Patterns, Recursion | |
12078 @subsection Recursion without Deferments | |
12079 @cindex Deferment in recursion | |
12080 @cindex Recursion without Deferments | |
12081 | |
12082 Let's consider again what happens with the @code{triangle-recursively} | |
12083 function. We will find that the intermediate calculations are | |
12084 deferred until all can be done. | |
12085 | |
12086 @need 800 | |
12087 Here is the function definition: | |
12088 | |
12089 @smallexample | |
12090 @group | |
12091 (defun triangle-recursively (number) | |
12092 "Return the sum of the numbers 1 through NUMBER inclusive. | |
12093 Uses recursion." | |
12094 (if (= number 1) ; @r{do-again-test} | |
12095 1 ; @r{then-part} | |
12096 (+ number ; @r{else-part} | |
12097 (triangle-recursively ; @r{recursive call} | |
12098 (1- number))))) ; @r{next-step-expression} | |
12099 @end group | |
12100 @end smallexample | |
12101 | |
12102 What happens when we call this function with a argument of 7? | |
12103 | |
12104 The first instance of the @code{triangle-recursively} function adds | |
12105 the number 7 to the value returned by a second instance of | |
12106 @code{triangle-recursively}, an instance that has been passed an | |
12107 argument of 6. That is to say, the first calculation is: | |
12108 | |
12109 @smallexample | |
12110 (+ 7 (triangle-recursively 6)) | |
12111 @end smallexample | |
12112 | |
12113 @noindent | |
12114 The first instance of @code{triangle-recursively}---you may want to | |
12115 think of it as a little robot---cannot complete its job. It must hand | |
12116 off the calculation for @code{(triangle-recursively 6)} to a second | |
12117 instance of the program, to a second robot. This second individual is | |
12118 completely different from the first one; it is, in the jargon, a | |
12119 `different instantiation'. Or, put another way, it is a different | |
12120 robot. It is the same model as the first; it calculates triangle | |
12121 numbers recursively; but it has a different serial number. | |
12122 | |
12123 And what does @code{(triangle-recursively 6)} return? It returns the | |
12124 number 6 added to the value returned by evaluating | |
12125 @code{triangle-recursively} with an argument of 5. Using the robot | |
12126 metaphor, it asks yet another robot to help it. | |
12127 | |
12128 @need 800 | |
12129 Now the total is: | |
12130 | |
12131 @smallexample | |
12132 (+ 7 6 (triangle-recursively 5)) | |
12133 @end smallexample | |
12134 | |
12135 @need 800 | |
12136 And what happens next? | |
12137 | |
12138 @smallexample | |
12139 (+ 7 6 5 (triangle-recursively 4)) | |
12140 @end smallexample | |
12141 | |
12142 Each time @code{triangle-recursively} is called, except for the last | |
12143 time, it creates another instance of the program---another robot---and | |
12144 asks it to make a calculation. | |
12145 | |
12146 @need 800 | |
12147 Eventually, the full addition is set up and performed: | |
12148 | |
12149 @smallexample | |
12150 (+ 7 6 5 4 3 2 1) | |
12151 @end smallexample | |
12152 | |
12153 This design for the function defers the calculation of the first step | |
12154 until the second can be done, and defers that until the third can be | |
12155 done, and so on. Each deferment means the computer must remember what | |
12156 is being waited on. This is not a problem when there are only a few | |
12157 steps, as in this example. But it can be a problem when there are | |
12158 more steps. | |
12159 | |
12160 @node No deferment solution, , No Deferment, Recursion | |
12161 @subsection No Deferment Solution | |
12162 @cindex No deferment solution | |
12163 @cindex Defermentless solution | |
12164 @cindex Solution without deferment | |
12165 | |
12166 The solution to the problem of deferred operations is to write in a | |
12167 manner that does not defer operations@footnote{The phrase @dfn{tail | |
12168 recursive} is used to describe such a process, one that uses | |
12169 `constant space'.}. This requires | |
12170 writing to a different pattern, often one that involves writing two | |
12171 function definitions, an `initialization' function and a `helper' | |
12172 function. | |
12173 | |
12174 The `initialization' function sets up the job; the `helper' function | |
12175 does the work. | |
12176 | |
12177 @need 1200 | |
12178 Here are the two function definitions for adding up numbers. They are | |
12179 so simple, I find them hard to understand. | |
12180 | |
12181 @smallexample | |
12182 @group | |
12183 (defun triangle-initialization (number) | |
12184 "Return the sum of the numbers 1 through NUMBER inclusive. | |
12185 This is the `initialization' component of a two function | |
12186 duo that uses recursion." | |
12187 (triangle-recursive-helper 0 0 number)) | |
12188 @end group | |
12189 @end smallexample | |
12190 | |
12191 @smallexample | |
12192 @group | |
12193 (defun triangle-recursive-helper (sum counter number) | |
12194 "Return SUM, using COUNTER, through NUMBER inclusive. | |
12195 This is the `helper' component of a two function duo | |
12196 that uses recursion." | |
12197 (if (> counter number) | |
12198 sum | |
12199 (triangle-recursive-helper (+ sum counter) ; @r{sum} | |
12200 (1+ counter) ; @r{counter} | |
12201 number))) ; @r{number} | |
12202 @end group | |
12203 @end smallexample | |
12204 | |
12205 @need 1250 | |
12206 Install both function definitions by evaluating them, then call | |
12207 @code{triangle-initialization} with 2 rows: | |
12208 | |
12209 @smallexample | |
12210 @group | |
12211 (triangle-initialization 2) | |
12212 @result{} 3 | |
12213 @end group | |
12214 @end smallexample | |
12215 | |
12216 The `initialization' function calls the first instance of the `helper' | |
12217 function with three arguments: zero, zero, and a number which is the | |
12218 number of rows in the triangle. | |
12219 | |
12220 The first two arguments passed to the `helper' function are | |
12221 initialization values. These values are changed when | |
12222 @code{triangle-recursive-helper} invokes new instances.@footnote{The | |
12223 jargon is mildly confusing: @code{triangle-recursive-helper} uses a | |
12224 process that is iterative in a procedure that is recursive. The | |
12225 process is called iterative because the computer need only record the | |
12226 three values, @code{sum}, @code{counter}, and @code{number}; the | |
12227 procedure is recursive because the function `calls itself'. On the | |
12228 other hand, both the process and the procedure used by | |
12229 @code{triangle-recursively} are called recursive. The word | |
12230 `recursive' has different meanings in the two contexts.} | |
12231 | |
12232 Let's see what happens when we have a triangle that has one row. (This | |
12233 triangle will have one pebble in it!) | |
12234 | |
12235 @need 1200 | |
12236 @code{triangle-initialization} will call its helper with | |
12237 the arguments @w{@code{0 0 1}}. That function will run the conditional | |
12238 test whether @code{(> counter number)}: | |
12239 | |
12240 @smallexample | |
12241 (> 0 1) | |
12242 @end smallexample | |
12243 | |
12244 @need 1200 | |
12245 @noindent | |
12246 and find that the result is false, so it will invoke | |
12247 the else-part of the @code{if} clause: | |
12248 | |
12249 @smallexample | |
12250 @group | |
12251 (triangle-recursive-helper | |
12252 (+ sum counter) ; @r{sum plus counter} @result{} @r{sum} | |
12253 (1+ counter) ; @r{increment counter} @result{} @r{counter} | |
12254 number) ; @r{number stays the same} | |
12255 @end group | |
12256 @end smallexample | |
12257 | |
12258 @need 800 | |
12259 @noindent | |
12260 which will first compute: | |
12261 | |
12262 @smallexample | |
12263 @group | |
12264 (triangle-recursive-helper (+ 0 0) ; @r{sum} | |
12265 (1+ 0) ; @r{counter} | |
12266 1) ; @r{number} | |
12267 @exdent which is: | |
12268 | |
12269 (triangle-recursive-helper 0 1 1) | |
12270 @end group | |
12271 @end smallexample | |
12272 | |
12273 Again, @code{(> counter number)} will be false, so again, the Lisp | |
12274 interpreter will evaluate @code{triangle-recursive-helper}, creating a | |
12275 new instance with new arguments. | |
12276 | |
12277 @need 800 | |
12278 This new instance will be; | |
12279 | |
12280 @smallexample | |
12281 @group | |
12282 (triangle-recursive-helper | |
12283 (+ sum counter) ; @r{sum plus counter} @result{} @r{sum} | |
12284 (1+ counter) ; @r{increment counter} @result{} @r{counter} | |
12285 number) ; @r{number stays the same} | |
12286 | |
12287 @exdent which is: | |
12288 | |
12289 (triangle-recursive-helper 1 2 1) | |
12290 @end group | |
12291 @end smallexample | |
12292 | |
12293 In this case, the @code{(> counter number)} test will be true! So the | |
12294 instance will return the value of the sum, which will be 1, as | |
12295 expected. | |
12296 | |
12297 Now, let's pass @code{triangle-initialization} an argument | |
12298 of 2, to find out how many pebbles there are in a triangle with two rows. | |
12299 | |
12300 That function calls @code{(triangle-recursive-helper 0 0 2)}. | |
12301 | |
12302 @need 800 | |
12303 In stages, the instances called will be: | |
12304 | |
12305 @smallexample | |
12306 @group | |
12307 @r{sum counter number} | |
12308 (triangle-recursive-helper 0 1 2) | |
12309 | |
12310 (triangle-recursive-helper 1 2 2) | |
12311 | |
12312 (triangle-recursive-helper 3 3 2) | |
12313 @end group | |
12314 @end smallexample | |
12315 | |
12316 When the last instance is called, the @code{(> counter number)} test | |
12317 will be true, so the instance will return the value of @code{sum}, | |
12318 which will be 3. | |
12319 | |
12320 This kind of pattern helps when you are writing functions that can use | |
12321 many resources in a computer. | |
12322 | |
12323 @need 1500 | |
12324 @node Looping exercise, , Recursion, Loops & Recursion | |
12325 @section Looping Exercise | |
12326 | |
12327 @itemize @bullet | |
12328 @item | |
12329 Write a function similar to @code{triangle} in which each row has a | |
12330 value which is the square of the row number. Use a @code{while} loop. | |
12331 | |
12332 @item | |
12333 Write a function similar to @code{triangle} that multiplies instead of | |
12334 adds the values. | |
12335 | |
12336 @item | |
12337 Rewrite these two functions recursively. Rewrite these functions | |
12338 using @code{cond}. | |
12339 | |
12340 @c comma in printed title causes problem in Info cross reference | |
12341 @item | |
12342 Write a function for Texinfo mode that creates an index entry at the | |
12343 beginning of a paragraph for every @samp{@@dfn} within the paragraph. | |
12344 (In a Texinfo file, @samp{@@dfn} marks a definition. This book is | |
12345 written in Texinfo.) | |
12346 | |
12347 Many of the functions you will need are described in two of the | |
12348 previous chapters, @ref{Cutting & Storing Text, , Cutting and Storing | |
12349 Text}, and @ref{Yanking, , Yanking Text Back}. If you use | |
12350 @code{forward-paragraph} to put the index entry at the beginning of | |
12351 the paragraph, you will have to use @w{@kbd{C-h f}} | |
12352 (@code{describe-function}) to find out how to make the command go | |
12353 backwards. | |
12354 | |
12355 For more information, see | |
12356 @ifinfo | |
12357 @ref{Indicating, , Indicating Definitions, texinfo}. | |
12358 @end ifinfo | |
12359 @ifhtml | |
12360 @ref{Indicating, , Indicating, texinfo, Texinfo Manual}, which goes to | |
12361 a Texinfo manual in the current directory. Or, if you are on the | |
12362 Internet, see | |
12363 @uref{http://www.gnu.org/software/texinfo/manual/texinfo/} | |
12364 @end ifhtml | |
12365 @iftex | |
12366 ``Indicating Definitions, Commands, etc.'' in @cite{Texinfo, The GNU | |
12367 Documentation Format}. | |
12368 @end iftex | |
12369 @end itemize | |
12370 | |
12371 @node Regexp Search, Counting Words, Loops & Recursion, Top | |
12372 @comment node-name, next, previous, up | |
12373 @chapter Regular Expression Searches | |
12374 @cindex Searches, illustrating | |
12375 @cindex Regular expression searches | |
12376 @cindex Patterns, searching for | |
12377 @cindex Motion by sentence and paragraph | |
12378 @cindex Sentences, movement by | |
12379 @cindex Paragraphs, movement by | |
12380 | |
12381 Regular expression searches are used extensively in GNU Emacs. The | |
12382 two functions, @code{forward-sentence} and @code{forward-paragraph}, | |
12383 illustrate these searches well. They use regular expressions to find | |
12384 where to move point. The phrase `regular expression' is often written | |
12385 as `regexp'. | |
12386 | |
12387 Regular expression searches are described in @ref{Regexp Search, , | |
12388 Regular Expression Search, emacs, The GNU Emacs Manual}, as well as in | |
12389 @ref{Regular Expressions, , , elisp, The GNU Emacs Lisp Reference | |
12390 Manual}. In writing this chapter, I am presuming that you have at | |
12391 least a mild acquaintance with them. The major point to remember is | |
12392 that regular expressions permit you to search for patterns as well as | |
12393 for literal strings of characters. For example, the code in | |
12394 @code{forward-sentence} searches for the pattern of possible | |
12395 characters that could mark the end of a sentence, and moves point to | |
12396 that spot. | |
12397 | |
12398 Before looking at the code for the @code{forward-sentence} function, it | |
12399 is worth considering what the pattern that marks the end of a sentence | |
12400 must be. The pattern is discussed in the next section; following that | |
12401 is a description of the regular expression search function, | |
12402 @code{re-search-forward}. The @code{forward-sentence} function | |
12403 is described in the section following. Finally, the | |
12404 @code{forward-paragraph} function is described in the last section of | |
12405 this chapter. @code{forward-paragraph} is a complex function that | |
12406 introduces several new features. | |
12407 | |
12408 @menu | |
12409 * sentence-end:: The regular expression for @code{sentence-end}. | |
12410 * re-search-forward:: Very similar to @code{search-forward}. | |
12411 * forward-sentence:: A straightforward example of regexp search. | |
12412 * forward-paragraph:: A somewhat complex example. | |
12413 * etags:: How to create your own @file{TAGS} table. | |
12414 * Regexp Review:: | |
12415 * re-search Exercises:: | |
12416 @end menu | |
12417 | |
12418 @node sentence-end, re-search-forward, Regexp Search, Regexp Search | |
12419 @comment node-name, next, previous, up | |
12420 @section The Regular Expression for @code{sentence-end} | |
12421 @findex sentence-end | |
12422 | |
12423 The symbol @code{sentence-end} is bound to the pattern that marks the | |
12424 end of a sentence. What should this regular expression be? | |
12425 | |
12426 Clearly, a sentence may be ended by a period, a question mark, or an | |
12427 exclamation mark. Indeed, in English, only clauses that end with one | |
12428 of those three characters should be considered the end of a sentence. | |
12429 This means that the pattern should include the character set: | |
12430 | |
12431 @smallexample | |
12432 [.?!] | |
12433 @end smallexample | |
12434 | |
12435 However, we do not want @code{forward-sentence} merely to jump to a | |
12436 period, a question mark, or an exclamation mark, because such a character | |
12437 might be used in the middle of a sentence. A period, for example, is | |
12438 used after abbreviations. So other information is needed. | |
12439 | |
12440 According to convention, you type two spaces after every sentence, but | |
12441 only one space after a period, a question mark, or an exclamation mark in | |
12442 the body of a sentence. So a period, a question mark, or an exclamation | |
12443 mark followed by two spaces is a good indicator of an end of sentence. | |
12444 However, in a file, the two spaces may instead be a tab or the end of a | |
12445 line. This means that the regular expression should include these three | |
12446 items as alternatives. | |
12447 | |
12448 @need 800 | |
12449 This group of alternatives will look like this: | |
12450 | |
12451 @smallexample | |
12452 @group | |
12453 \\($\\| \\| \\) | |
12454 ^ ^^ | |
12455 TAB SPC | |
12456 @end group | |
12457 @end smallexample | |
12458 | |
12459 @noindent | |
12460 Here, @samp{$} indicates the end of the line, and I have pointed out | |
12461 where the tab and two spaces are inserted in the expression. Both are | |
12462 inserted by putting the actual characters into the expression. | |
12463 | |
12464 Two backslashes, @samp{\\}, are required before the parentheses and | |
12465 vertical bars: the first backslash quotes the following backslash in | |
12466 Emacs; and the second indicates that the following character, the | |
12467 parenthesis or the vertical bar, is special. | |
12468 | |
12469 @need 1000 | |
12470 Also, a sentence may be followed by one or more carriage returns, like | |
12471 this: | |
12472 | |
12473 @smallexample | |
12474 @group | |
12475 [ | |
12476 ]* | |
12477 @end group | |
12478 @end smallexample | |
12479 | |
12480 @noindent | |
12481 Like tabs and spaces, a carriage return is inserted into a regular | |
12482 expression by inserting it literally. The asterisk indicates that the | |
12483 @key{RET} is repeated zero or more times. | |
12484 | |
12485 But a sentence end does not consist only of a period, a question mark or | |
12486 an exclamation mark followed by appropriate space: a closing quotation | |
12487 mark or a closing brace of some kind may precede the space. Indeed more | |
12488 than one such mark or brace may precede the space. These require a | |
12489 expression that looks like this: | |
12490 | |
12491 @smallexample | |
12492 []\"')@}]* | |
12493 @end smallexample | |
12494 | |
12495 In this expression, the first @samp{]} is the first character in the | |
12496 expression; the second character is @samp{"}, which is preceded by a | |
12497 @samp{\} to tell Emacs the @samp{"} is @emph{not} special. The last | |
12498 three characters are @samp{'}, @samp{)}, and @samp{@}}. | |
12499 | |
12500 All this suggests what the regular expression pattern for matching the | |
12501 end of a sentence should be; and, indeed, if we evaluate | |
12502 @code{sentence-end} we find that it returns the following value: | |
12503 | |
12504 @smallexample | |
12505 @group | |
12506 sentence-end | |
12507 @result{} "[.?!][]\"')@}]*\\($\\| \\| \\)[ | |
12508 ]*" | |
12509 @end group | |
12510 @end smallexample | |
12511 | |
12512 @noindent | |
12513 (Well, not in GNU Emacs 22; that is because of an effort to make the | |
12514 process simpler and to handle more glyphs and languages. When the | |
12515 value of @code{sentence-end} is @code{nil}, then use the value defined | |
12516 by the function @code{sentence-end}. (Here is a use of the difference | |
12517 between a value and a function in Emacs Lisp.) The function returns a | |
12518 value constructed from the variables @code{sentence-end-base}, | |
12519 @code{sentence-end-double-space}, @code{sentence-end-without-period}, | |
12520 and @code{sentence-end-without-space}. The critical variable is | |
12521 @code{sentence-end-base}; its global value is similar to the one | |
12522 described above but it also contains two additional quotation marks. | |
12523 These have differing degrees of curliness. The | |
12524 @code{sentence-end-without-period} variable, when true, tells Emacs | |
12525 that a sentence may end without a period, such as text in Thai.) | |
12526 | |
12527 @ignore | |
12528 @noindent | |
12529 (Note that here the @key{TAB}, two spaces, and @key{RET} are shown | |
12530 literally in the pattern.) | |
12531 | |
12532 This regular expression can be deciphered as follows: | |
12533 | |
12534 @table @code | |
12535 @item [.?!] | |
12536 The first part of the pattern is the three characters, a period, a question | |
12537 mark and an exclamation mark, within square brackets. The pattern must | |
12538 begin with one or other of these characters. | |
12539 | |
12540 @item []\"')@}]* | |
12541 The second part of the pattern is the group of closing braces and | |
12542 quotation marks, which can appear zero or more times. These may follow | |
12543 the period, question mark or exclamation mark. In a regular expression, | |
12544 the backslash, @samp{\}, followed by the double quotation mark, | |
12545 @samp{"}, indicates the class of string-quote characters. Usually, the | |
12546 double quotation mark is the only character in this class. The | |
12547 asterisk, @samp{*}, indicates that the items in the previous group (the | |
12548 group surrounded by square brackets, @samp{[]}) may be repeated zero or | |
12549 more times. | |
12550 | |
12551 @item \\($\\| \\| \\) | |
12552 The third part of the pattern is one or other of: either the end of a | |
12553 line, or two blank spaces, or a tab. The double back-slashes are used | |
12554 to prevent Emacs from reading the parentheses and vertical bars as part | |
12555 of the search pattern; the parentheses are used to mark the group and | |
12556 the vertical bars are used to indicated that the patterns to either side | |
12557 of them are alternatives. The dollar sign is used to indicate the end | |
12558 of a line and both the two spaces and the tab are each inserted as is to | |
12559 indicate what they are. | |
12560 | |
12561 @item [@key{RET}]* | |
12562 Finally, the last part of the pattern indicates that the end of the line | |
12563 or the whitespace following the period, question mark or exclamation | |
12564 mark may, but need not, be followed by one or more carriage returns. In | |
12565 the pattern, the carriage return is inserted as an actual carriage | |
12566 return between square brackets but here it is shown as @key{RET}. | |
12567 @end table | |
12568 @end ignore | |
12569 | |
12570 @node re-search-forward, forward-sentence, sentence-end, Regexp Search | |
12571 @comment node-name, next, previous, up | |
12572 @section The @code{re-search-forward} Function | |
12573 @findex re-search-forward | |
12574 | |
12575 The @code{re-search-forward} function is very like the | |
12576 @code{search-forward} function. (@xref{search-forward, , The | |
12577 @code{search-forward} Function}.) | |
12578 | |
12579 @code{re-search-forward} searches for a regular expression. If the | |
12580 search is successful, it leaves point immediately after the last | |
12581 character in the target. If the search is backwards, it leaves point | |
12582 just before the first character in the target. You may tell | |
12583 @code{re-search-forward} to return @code{t} for true. (Moving point | |
12584 is therefore a `side effect'.) | |
12585 | |
12586 Like @code{search-forward}, the @code{re-search-forward} function takes | |
12587 four arguments: | |
12588 | |
12589 @enumerate | |
12590 @item | |
12591 The first argument is the regular expression that the function searches | |
12592 for. The regular expression will be a string between quotations marks. | |
12593 | |
12594 @item | |
12595 The optional second argument limits how far the function will search; it is a | |
12596 bound, which is specified as a position in the buffer. | |
12597 | |
12598 @item | |
12599 The optional third argument specifies how the function responds to | |
12600 failure: @code{nil} as the third argument causes the function to | |
12601 signal an error (and print a message) when the search fails; any other | |
12602 value causes it to return @code{nil} if the search fails and @code{t} | |
12603 if the search succeeds. | |
12604 | |
12605 @item | |
12606 The optional fourth argument is the repeat count. A negative repeat | |
12607 count causes @code{re-search-forward} to search backwards. | |
12608 @end enumerate | |
12609 | |
12610 @need 800 | |
12611 The template for @code{re-search-forward} looks like this: | |
12612 | |
12613 @smallexample | |
12614 @group | |
12615 (re-search-forward "@var{regular-expression}" | |
12616 @var{limit-of-search} | |
12617 @var{what-to-do-if-search-fails} | |
12618 @var{repeat-count}) | |
12619 @end group | |
12620 @end smallexample | |
12621 | |
12622 The second, third, and fourth arguments are optional. However, if you | |
12623 want to pass a value to either or both of the last two arguments, you | |
12624 must also pass a value to all the preceding arguments. Otherwise, the | |
12625 Lisp interpreter will mistake which argument you are passing the value | |
12626 to. | |
12627 | |
12628 @need 1200 | |
12629 In the @code{forward-sentence} function, the regular expression will be | |
12630 the value of the variable @code{sentence-end}. In simple form, that is: | |
12631 | |
12632 @smallexample | |
12633 @group | |
12634 "[.?!][]\"')@}]*\\($\\| \\| \\)[ | |
12635 ]*" | |
12636 @end group | |
12637 @end smallexample | |
12638 | |
12639 @noindent | |
12640 The limit of the search will be the end of the paragraph (since a | |
12641 sentence cannot go beyond a paragraph). If the search fails, the | |
12642 function will return @code{nil}; and the repeat count will be provided | |
12643 by the argument to the @code{forward-sentence} function. | |
12644 | |
12645 @node forward-sentence, forward-paragraph, re-search-forward, Regexp Search | |
12646 @comment node-name, next, previous, up | |
12647 @section @code{forward-sentence} | |
12648 @findex forward-sentence | |
12649 | |
12650 The command to move the cursor forward a sentence is a straightforward | |
12651 illustration of how to use regular expression searches in Emacs Lisp. | |
12652 Indeed, the function looks longer and more complicated than it is; this | |
12653 is because the function is designed to go backwards as well as forwards; | |
12654 and, optionally, over more than one sentence. The function is usually | |
12655 bound to the key command @kbd{M-e}. | |
12656 | |
12657 @menu | |
12658 * Complete forward-sentence:: | |
12659 * fwd-sentence while loops:: Two @code{while} loops. | |
12660 * fwd-sentence re-search:: A regular expression search. | |
12661 @end menu | |
12662 | |
12663 @node Complete forward-sentence, fwd-sentence while loops, forward-sentence, forward-sentence | |
12664 @ifnottex | |
12665 @unnumberedsubsec Complete @code{forward-sentence} function definition | |
12666 @end ifnottex | |
12667 | |
12668 @need 1250 | |
12669 Here is the code for @code{forward-sentence}: | |
12670 | |
12671 @c in GNU Emacs 22 | |
12672 @smallexample | |
12673 @group | |
12674 (defun forward-sentence (&optional arg) | |
12675 "Move forward to next `sentence-end'. With argument, repeat. | |
12676 With negative argument, move backward repeatedly to `sentence-beginning'. | |
12677 | |
12678 The variable `sentence-end' is a regular expression that matches ends of | |
12679 sentences. Also, every paragraph boundary terminates sentences as well." | |
12680 @end group | |
12681 @group | |
12682 (interactive "p") | |
12683 (or arg (setq arg 1)) | |
12684 (let ((opoint (point)) | |
12685 (sentence-end (sentence-end))) | |
12686 (while (< arg 0) | |
12687 (let ((pos (point)) | |
12688 (par-beg (save-excursion (start-of-paragraph-text) (point)))) | |
12689 (if (and (re-search-backward sentence-end par-beg t) | |
12690 (or (< (match-end 0) pos) | |
12691 (re-search-backward sentence-end par-beg t))) | |
12692 (goto-char (match-end 0)) | |
12693 (goto-char par-beg))) | |
12694 (setq arg (1+ arg))) | |
12695 @end group | |
12696 @group | |
12697 (while (> arg 0) | |
12698 (let ((par-end (save-excursion (end-of-paragraph-text) (point)))) | |
12699 (if (re-search-forward sentence-end par-end t) | |
12700 (skip-chars-backward " \t\n") | |
12701 (goto-char par-end))) | |
12702 (setq arg (1- arg))) | |
12703 (constrain-to-field nil opoint t))) | |
12704 @end group | |
12705 @end smallexample | |
12706 | |
12707 @ignore | |
12708 GNU Emacs 21 | |
12709 @smallexample | |
12710 @group | |
12711 (defun forward-sentence (&optional arg) | |
12712 "Move forward to next sentence-end. With argument, repeat. | |
12713 With negative argument, move backward repeatedly to sentence-beginning. | |
12714 Sentence ends are identified by the value of sentence-end | |
12715 treated as a regular expression. Also, every paragraph boundary | |
12716 terminates sentences as well." | |
12717 @end group | |
12718 @group | |
12719 (interactive "p") | |
12720 (or arg (setq arg 1)) | |
12721 (while (< arg 0) | |
12722 (let ((par-beg | |
12723 (save-excursion (start-of-paragraph-text) (point)))) | |
12724 (if (re-search-backward | |
12725 (concat sentence-end "[^ \t\n]") par-beg t) | |
12726 (goto-char (1- (match-end 0))) | |
12727 (goto-char par-beg))) | |
12728 (setq arg (1+ arg))) | |
12729 (while (> arg 0) | |
12730 (let ((par-end | |
12731 (save-excursion (end-of-paragraph-text) (point)))) | |
12732 (if (re-search-forward sentence-end par-end t) | |
12733 (skip-chars-backward " \t\n") | |
12734 (goto-char par-end))) | |
12735 (setq arg (1- arg)))) | |
12736 @end group | |
12737 @end smallexample | |
12738 @end ignore | |
12739 | |
12740 The function looks long at first sight and it is best to look at its | |
12741 skeleton first, and then its muscle. The way to see the skeleton is to | |
12742 look at the expressions that start in the left-most columns: | |
12743 | |
12744 @smallexample | |
12745 @group | |
12746 (defun forward-sentence (&optional arg) | |
12747 "@var{documentation}@dots{}" | |
12748 (interactive "p") | |
12749 (or arg (setq arg 1)) | |
12750 (let ((opoint (point)) (sentence-end (sentence-end))) | |
12751 (while (< arg 0) | |
12752 (let ((pos (point)) | |
12753 (par-beg (save-excursion (start-of-paragraph-text) (point)))) | |
12754 @var{rest-of-body-of-while-loop-when-going-backwards} | |
12755 (while (> arg 0) | |
12756 (let ((par-end (save-excursion (end-of-paragraph-text) (point)))) | |
12757 @var{rest-of-body-of-while-loop-when-going-forwards} | |
12758 @var{handle-forms-and-equivalent} | |
12759 @end group | |
12760 @end smallexample | |
12761 | |
12762 This looks much simpler! The function definition consists of | |
12763 documentation, an @code{interactive} expression, an @code{or} | |
12764 expression, a @code{let} expression, and @code{while} loops. | |
12765 | |
12766 Let's look at each of these parts in turn. | |
12767 | |
12768 We note that the documentation is thorough and understandable. | |
12769 | |
12770 The function has an @code{interactive "p"} declaration. This means | |
12771 that the processed prefix argument, if any, is passed to the | |
12772 function as its argument. (This will be a number.) If the function | |
12773 is not passed an argument (it is optional) then the argument | |
12774 @code{arg} will be bound to 1. | |
12775 | |
12776 When @code{forward-sentence} is called non-interactively without an | |
12777 argument, @code{arg} is bound to @code{nil}. The @code{or} expression | |
12778 handles this. What it does is either leave the value of @code{arg} as | |
12779 it is, but only if @code{arg} is bound to a value; or it sets the | |
12780 value of @code{arg} to 1, in the case when @code{arg} is bound to | |
12781 @code{nil}. | |
12782 | |
12783 Next is a @code{let}. That specifies the values of two local | |
12784 variables, @code{point} and @code{sentence-end}. The local value of | |
12785 point, from before the search, is used in the | |
12786 @code{constrain-to-field} function which handles forms and | |
12787 equivalents. The @code{sentence-end} variable is set by the | |
12788 @code{sentence-end} function. | |
12789 | |
12790 @node fwd-sentence while loops, fwd-sentence re-search, Complete forward-sentence, forward-sentence | |
12791 @unnumberedsubsec The @code{while} loops | |
12792 | |
12793 Two @code{while} loops follow. The first @code{while} has a | |
12794 true-or-false-test that tests true if the prefix argument for | |
12795 @code{forward-sentence} is a negative number. This is for going | |
12796 backwards. The body of this loop is similar to the body of the second | |
12797 @code{while} clause, but it is not exactly the same. We will skip | |
12798 this @code{while} loop and concentrate on the second @code{while} | |
12799 loop. | |
12800 | |
12801 @need 1500 | |
12802 The second @code{while} loop is for moving point forward. Its skeleton | |
12803 looks like this: | |
12804 | |
12805 @smallexample | |
12806 @group | |
12807 (while (> arg 0) ; @r{true-or-false-test} | |
12808 (let @var{varlist} | |
12809 (if (@var{true-or-false-test}) | |
12810 @var{then-part} | |
12811 @var{else-part} | |
12812 (setq arg (1- arg)))) ; @code{while} @r{loop decrementer} | |
12813 @end group | |
12814 @end smallexample | |
12815 | |
12816 The @code{while} loop is of the decrementing kind. | |
12817 (@xref{Decrementing Loop, , A Loop with a Decrementing Counter}.) It | |
12818 has a true-or-false-test that tests true so long as the counter (in | |
12819 this case, the variable @code{arg}) is greater than zero; and it has a | |
12820 decrementer that subtracts 1 from the value of the counter every time | |
12821 the loop repeats. | |
12822 | |
12823 If no prefix argument is given to @code{forward-sentence}, which is | |
12824 the most common way the command is used, this @code{while} loop will | |
12825 run once, since the value of @code{arg} will be 1. | |
12826 | |
12827 The body of the @code{while} loop consists of a @code{let} expression, | |
12828 which creates and binds a local variable, and has, as its body, an | |
12829 @code{if} expression. | |
12830 | |
12831 @need 1250 | |
12832 The body of the @code{while} loop looks like this: | |
12833 | |
12834 @smallexample | |
12835 @group | |
12836 (let ((par-end | |
12837 (save-excursion (end-of-paragraph-text) (point)))) | |
12838 (if (re-search-forward sentence-end par-end t) | |
12839 (skip-chars-backward " \t\n") | |
12840 (goto-char par-end))) | |
12841 @end group | |
12842 @end smallexample | |
12843 | |
12844 The @code{let} expression creates and binds the local variable | |
12845 @code{par-end}. As we shall see, this local variable is designed to | |
12846 provide a bound or limit to the regular expression search. If the | |
12847 search fails to find a proper sentence ending in the paragraph, it will | |
12848 stop on reaching the end of the paragraph. | |
12849 | |
12850 But first, let us examine how @code{par-end} is bound to the value of | |
12851 the end of the paragraph. What happens is that the @code{let} sets the | |
12852 value of @code{par-end} to the value returned when the Lisp interpreter | |
12853 evaluates the expression | |
12854 | |
12855 @smallexample | |
12856 @group | |
12857 (save-excursion (end-of-paragraph-text) (point)) | |
12858 @end group | |
12859 @end smallexample | |
12860 | |
12861 @noindent | |
12862 In this expression, @code{(end-of-paragraph-text)} moves point to the | |
12863 end of the paragraph, @code{(point)} returns the value of point, and then | |
12864 @code{save-excursion} restores point to its original position. Thus, | |
12865 the @code{let} binds @code{par-end} to the value returned by the | |
12866 @code{save-excursion} expression, which is the position of the end of | |
12867 the paragraph. (The @code{end-of-paragraph-text} function uses | |
12868 @code{forward-paragraph}, which we will discuss shortly.) | |
12869 | |
12870 @need 1200 | |
12871 Emacs next evaluates the body of the @code{let}, which is an @code{if} | |
12872 expression that looks like this: | |
12873 | |
12874 @smallexample | |
12875 @group | |
12876 (if (re-search-forward sentence-end par-end t) ; @r{if-part} | |
12877 (skip-chars-backward " \t\n") ; @r{then-part} | |
12878 (goto-char par-end))) ; @r{else-part} | |
12879 @end group | |
12880 @end smallexample | |
12881 | |
12882 The @code{if} tests whether its first argument is true and if so, | |
12883 evaluates its then-part; otherwise, the Emacs Lisp interpreter | |
12884 evaluates the else-part. The true-or-false-test of the @code{if} | |
12885 expression is the regular expression search. | |
12886 | |
12887 It may seem odd to have what looks like the `real work' of | |
12888 the @code{forward-sentence} function buried here, but this is a common | |
12889 way this kind of operation is carried out in Lisp. | |
12890 | |
12891 @node fwd-sentence re-search, , fwd-sentence while loops, forward-sentence | |
12892 @unnumberedsubsec The regular expression search | |
12893 | |
12894 The @code{re-search-forward} function searches for the end of the | |
12895 sentence, that is, for the pattern defined by the @code{sentence-end} | |
12896 regular expression. If the pattern is found---if the end of the sentence is | |
12897 found---then the @code{re-search-forward} function does two things: | |
12898 | |
12899 @enumerate | |
12900 @item | |
12901 The @code{re-search-forward} function carries out a side effect, which | |
12902 is to move point to the end of the occurrence found. | |
12903 | |
12904 @item | |
12905 The @code{re-search-forward} function returns a value of true. This is | |
12906 the value received by the @code{if}, and means that the search was | |
12907 successful. | |
12908 @end enumerate | |
12909 | |
12910 @noindent | |
12911 The side effect, the movement of point, is completed before the | |
12912 @code{if} function is handed the value returned by the successful | |
12913 conclusion of the search. | |
12914 | |
12915 When the @code{if} function receives the value of true from a successful | |
12916 call to @code{re-search-forward}, the @code{if} evaluates the then-part, | |
12917 which is the expression @code{(skip-chars-backward " \t\n")}. This | |
12918 expression moves backwards over any blank spaces, tabs or carriage | |
12919 returns until a printed character is found and then leaves point after | |
12920 the character. Since point has already been moved to the end of the | |
12921 pattern that marks the end of the sentence, this action leaves point | |
12922 right after the closing printed character of the sentence, which is | |
12923 usually a period. | |
12924 | |
12925 On the other hand, if the @code{re-search-forward} function fails to | |
12926 find a pattern marking the end of the sentence, the function returns | |
12927 false. The false then causes the @code{if} to evaluate its third | |
12928 argument, which is @code{(goto-char par-end)}: it moves point to the | |
12929 end of the paragraph. | |
12930 | |
12931 (And if the text is in a form or equivalent, and point may not move | |
12932 fully, then the @code{constrain-to-field} function comes into play.) | |
12933 | |
12934 Regular expression searches are exceptionally useful and the pattern | |
12935 illustrated by @code{re-search-forward}, in which the search is the | |
12936 test of an @code{if} expression, is handy. You will see or write code | |
12937 incorporating this pattern often. | |
12938 | |
12939 @node forward-paragraph, etags, forward-sentence, Regexp Search | |
12940 @comment node-name, next, previous, up | |
12941 @section @code{forward-paragraph}: a Goldmine of Functions | |
12942 @findex forward-paragraph | |
12943 | |
12944 @ignore | |
12945 @c in GNU Emacs 22 | |
12946 (defun forward-paragraph (&optional arg) | |
12947 "Move forward to end of paragraph. | |
12948 With argument ARG, do it ARG times; | |
12949 a negative argument ARG = -N means move backward N paragraphs. | |
12950 | |
12951 A line which `paragraph-start' matches either separates paragraphs | |
12952 \(if `paragraph-separate' matches it also) or is the first line of a paragraph. | |
12953 A paragraph end is the beginning of a line which is not part of the paragraph | |
12954 to which the end of the previous line belongs, or the end of the buffer. | |
12955 Returns the count of paragraphs left to move." | |
12956 (interactive "p") | |
12957 (or arg (setq arg 1)) | |
12958 (let* ((opoint (point)) | |
12959 (fill-prefix-regexp | |
12960 (and fill-prefix (not (equal fill-prefix "")) | |
12961 (not paragraph-ignore-fill-prefix) | |
12962 (regexp-quote fill-prefix))) | |
12963 ;; Remove ^ from paragraph-start and paragraph-sep if they are there. | |
12964 ;; These regexps shouldn't be anchored, because we look for them | |
12965 ;; starting at the left-margin. This allows paragraph commands to | |
12966 ;; work normally with indented text. | |
12967 ;; This hack will not find problem cases like "whatever\\|^something". | |
12968 (parstart (if (and (not (equal "" paragraph-start)) | |
12969 (equal ?^ (aref paragraph-start 0))) | |
12970 (substring paragraph-start 1) | |
12971 paragraph-start)) | |
12972 (parsep (if (and (not (equal "" paragraph-separate)) | |
12973 (equal ?^ (aref paragraph-separate 0))) | |
12974 (substring paragraph-separate 1) | |
12975 paragraph-separate)) | |
12976 (parsep | |
12977 (if fill-prefix-regexp | |
12978 (concat parsep "\\|" | |
12979 fill-prefix-regexp "[ \t]*$") | |
12980 parsep)) | |
12981 ;; This is used for searching. | |
12982 (sp-parstart (concat "^[ \t]*\\(?:" parstart "\\|" parsep "\\)")) | |
12983 start found-start) | |
12984 (while (and (< arg 0) (not (bobp))) | |
12985 (if (and (not (looking-at parsep)) | |
12986 (re-search-backward "^\n" (max (1- (point)) (point-min)) t) | |
12987 (looking-at parsep)) | |
12988 (setq arg (1+ arg)) | |
12989 (setq start (point)) | |
12990 ;; Move back over paragraph-separating lines. | |
12991 (forward-char -1) (beginning-of-line) | |
12992 (while (and (not (bobp)) | |
12993 (progn (move-to-left-margin) | |
12994 (looking-at parsep))) | |
12995 (forward-line -1)) | |
12996 (if (bobp) | |
12997 nil | |
12998 (setq arg (1+ arg)) | |
12999 ;; Go to end of the previous (non-separating) line. | |
13000 (end-of-line) | |
13001 ;; Search back for line that starts or separates paragraphs. | |
13002 (if (if fill-prefix-regexp | |
13003 ;; There is a fill prefix; it overrides parstart. | |
13004 (let (multiple-lines) | |
13005 (while (and (progn (beginning-of-line) (not (bobp))) | |
13006 (progn (move-to-left-margin) | |
13007 (not (looking-at parsep))) | |
13008 (looking-at fill-prefix-regexp)) | |
13009 (unless (= (point) start) | |
13010 (setq multiple-lines t)) | |
13011 (forward-line -1)) | |
13012 (move-to-left-margin) | |
13013 ;; This deleted code caused a long hanging-indent line | |
13014 ;; not to be filled together with the following lines. | |
13015 ;; ;; Don't move back over a line before the paragraph | |
13016 ;; ;; which doesn't start with fill-prefix | |
13017 ;; ;; unless that is the only line we've moved over. | |
13018 ;; (and (not (looking-at fill-prefix-regexp)) | |
13019 ;; multiple-lines | |
13020 ;; (forward-line 1)) | |
13021 (not (bobp))) | |
13022 (while (and (re-search-backward sp-parstart nil 1) | |
13023 (setq found-start t) | |
13024 ;; Found a candidate, but need to check if it is a | |
13025 ;; REAL parstart. | |
13026 (progn (setq start (point)) | |
13027 (move-to-left-margin) | |
13028 (not (looking-at parsep))) | |
13029 (not (and (looking-at parstart) | |
13030 (or (not use-hard-newlines) | |
13031 (bobp) | |
13032 (get-text-property | |
13033 (1- start) 'hard))))) | |
13034 (setq found-start nil) | |
13035 (goto-char start)) | |
13036 found-start) | |
13037 ;; Found one. | |
13038 (progn | |
13039 ;; Move forward over paragraph separators. | |
13040 ;; We know this cannot reach the place we started | |
13041 ;; because we know we moved back over a non-separator. | |
13042 (while (and (not (eobp)) | |
13043 (progn (move-to-left-margin) | |
13044 (looking-at parsep))) | |
13045 (forward-line 1)) | |
13046 ;; If line before paragraph is just margin, back up to there. | |
13047 (end-of-line 0) | |
13048 (if (> (current-column) (current-left-margin)) | |
13049 (forward-char 1) | |
13050 (skip-chars-backward " \t") | |
13051 (if (not (bolp)) | |
13052 (forward-line 1)))) | |
13053 ;; No starter or separator line => use buffer beg. | |
13054 (goto-char (point-min)))))) | |
13055 | |
13056 (while (and (> arg 0) (not (eobp))) | |
13057 ;; Move forward over separator lines... | |
13058 (while (and (not (eobp)) | |
13059 (progn (move-to-left-margin) (not (eobp))) | |
13060 (looking-at parsep)) | |
13061 (forward-line 1)) | |
13062 (unless (eobp) (setq arg (1- arg))) | |
13063 ;; ... and one more line. | |
13064 (forward-line 1) | |
13065 (if fill-prefix-regexp | |
13066 ;; There is a fill prefix; it overrides parstart. | |
13067 (while (and (not (eobp)) | |
13068 (progn (move-to-left-margin) (not (eobp))) | |
13069 (not (looking-at parsep)) | |
13070 (looking-at fill-prefix-regexp)) | |
13071 (forward-line 1)) | |
13072 (while (and (re-search-forward sp-parstart nil 1) | |
13073 (progn (setq start (match-beginning 0)) | |
13074 (goto-char start) | |
13075 (not (eobp))) | |
13076 (progn (move-to-left-margin) | |
13077 (not (looking-at parsep))) | |
13078 (or (not (looking-at parstart)) | |
13079 (and use-hard-newlines | |
13080 (not (get-text-property (1- start) 'hard))))) | |
13081 (forward-char 1)) | |
13082 (if (< (point) (point-max)) | |
13083 (goto-char start)))) | |
13084 (constrain-to-field nil opoint t) | |
13085 ;; Return the number of steps that could not be done. | |
13086 arg)) | |
13087 @end ignore | |
13088 | |
13089 The @code{forward-paragraph} function moves point forward to the end | |
13090 of the paragraph. It is usually bound to @kbd{M-@}} and makes use of a | |
13091 number of functions that are important in themselves, including | |
13092 @code{let*}, @code{match-beginning}, and @code{looking-at}. | |
13093 | |
13094 The function definition for @code{forward-paragraph} is considerably | |
13095 longer than the function definition for @code{forward-sentence} | |
13096 because it works with a paragraph, each line of which may begin with a | |
13097 fill prefix. | |
13098 | |
13099 A fill prefix consists of a string of characters that are repeated at | |
13100 the beginning of each line. For example, in Lisp code, it is a | |
13101 convention to start each line of a paragraph-long comment with | |
13102 @samp{;;; }. In Text mode, four blank spaces make up another common | |
13103 fill prefix, creating an indented paragraph. (@xref{Fill Prefix, , , | |
13104 emacs, The GNU Emacs Manual}, for more information about fill | |
13105 prefixes.) | |
13106 | |
13107 The existence of a fill prefix means that in addition to being able to | |
13108 find the end of a paragraph whose lines begin on the left-most | |
13109 column, the @code{forward-paragraph} function must be able to find the | |
13110 end of a paragraph when all or many of the lines in the buffer begin | |
13111 with the fill prefix. | |
13112 | |
13113 Moreover, it is sometimes practical to ignore a fill prefix that | |
13114 exists, especially when blank lines separate paragraphs. | |
13115 This is an added complication. | |
13116 | |
13117 @menu | |
13118 * forward-paragraph in brief:: Key parts of the function definition. | |
13119 * fwd-para let:: The @code{let*} expression. | |
13120 * fwd-para while:: The forward motion @code{while} loop. | |
13121 @end menu | |
13122 | |
13123 @node forward-paragraph in brief, fwd-para let, forward-paragraph, forward-paragraph | |
13124 @ifnottex | |
13125 @unnumberedsubsec Shortened @code{forward-paragraph} function definition | |
13126 @end ifnottex | |
13127 | |
13128 Rather than print all of the @code{forward-paragraph} function, we | |
13129 will only print parts of it. Read without preparation, the function | |
13130 can be daunting! | |
13131 | |
13132 @need 800 | |
13133 In outline, the function looks like this: | |
13134 | |
13135 @smallexample | |
13136 @group | |
13137 (defun forward-paragraph (&optional arg) | |
13138 "@var{documentation}@dots{}" | |
13139 (interactive "p") | |
13140 (or arg (setq arg 1)) | |
13141 (let* | |
13142 @var{varlist} | |
13143 (while (and (< arg 0) (not (bobp))) ; @r{backward-moving-code} | |
13144 @dots{} | |
13145 (while (and (> arg 0) (not (eobp))) ; @r{forward-moving-code} | |
13146 @dots{} | |
13147 @end group | |
13148 @end smallexample | |
13149 | |
13150 The first parts of the function are routine: the function's argument | |
13151 list consists of one optional argument. Documentation follows. | |
13152 | |
13153 The lower case @samp{p} in the @code{interactive} declaration means | |
13154 that the processed prefix argument, if any, is passed to the function. | |
13155 This will be a number, and is the repeat count of how many paragraphs | |
13156 point will move. The @code{or} expression in the next line handles | |
13157 the common case when no argument is passed to the function, which occurs | |
13158 if the function is called from other code rather than interactively. | |
13159 This case was described earlier. (@xref{forward-sentence, The | |
13160 @code{forward-sentence} function}.) Now we reach the end of the | |
13161 familiar part of this function. | |
13162 | |
13163 @node fwd-para let, fwd-para while, forward-paragraph in brief, forward-paragraph | |
13164 @unnumberedsubsec The @code{let*} expression | |
13165 | |
13166 The next line of the @code{forward-paragraph} function begins a | |
13167 @code{let*} expression. This is a different than @code{let}. The | |
13168 symbol is @code{let*} not @code{let}. | |
13169 | |
13170 The @code{let*} special form is like @code{let} except that Emacs sets | |
13171 each variable in sequence, one after another, and variables in the | |
13172 latter part of the varlist can make use of the values to which Emacs | |
13173 set variables in the earlier part of the varlist. | |
13174 | |
13175 @ignore | |
13176 ( refappend save-excursion, , code save-excursion in code append-to-buffer .) | |
13177 @end ignore | |
13178 | |
13179 (@ref{append save-excursion, , @code{save-excursion} in @code{append-to-buffer}}.) | |
13180 | |
13181 In the @code{let*} expression in this function, Emacs binds a total of | |
13182 seven variables: @code{opoint}, @code{fill-prefix-regexp}, | |
13183 @code{parstart}, @code{parsep}, @code{sp-parstart}, @code{start}, and | |
13184 @code{found-start}. | |
13185 | |
13186 The variable @code{parsep} appears twice, first, to remove instances | |
13187 of @samp{^}, and second, to handle fill prefixes. | |
13188 | |
13189 The variable @code{opoint} is just the value of @code{point}. As you | |
13190 can guess, it is used in a @code{constrain-to-field} expression, just | |
13191 as in @code{forward-sentence}. | |
13192 | |
13193 The variable @code{fill-prefix-regexp} is set to the value returned by | |
13194 evaluating the following list: | |
13195 | |
13196 @smallexample | |
13197 @group | |
13198 (and fill-prefix | |
13199 (not (equal fill-prefix "")) | |
13200 (not paragraph-ignore-fill-prefix) | |
13201 (regexp-quote fill-prefix)) | |
13202 @end group | |
13203 @end smallexample | |
13204 | |
13205 @noindent | |
13206 This is an expression whose first element is the @code{and} special form. | |
13207 | |
13208 As we learned earlier (@pxref{kill-new function, , The @code{kill-new} | |
13209 function}), the @code{and} special form evaluates each of its | |
13210 arguments until one of the arguments returns a value of @code{nil}, in | |
13211 which case the @code{and} expression returns @code{nil}; however, if | |
13212 none of the arguments returns a value of @code{nil}, the value | |
13213 resulting from evaluating the last argument is returned. (Since such | |
13214 a value is not @code{nil}, it is considered true in Lisp.) In other | |
13215 words, an @code{and} expression returns a true value only if all its | |
13216 arguments are true. | |
13217 @findex and | |
13218 | |
13219 In this case, the variable @code{fill-prefix-regexp} is bound to a | |
13220 non-@code{nil} value only if the following four expressions produce a | |
13221 true (i.e., a non-@code{nil}) value when they are evaluated; otherwise, | |
13222 @code{fill-prefix-regexp} is bound to @code{nil}. | |
13223 | |
13224 @table @code | |
13225 @item fill-prefix | |
13226 When this variable is evaluated, the value of the fill prefix, if any, | |
13227 is returned. If there is no fill prefix, this variable returns | |
13228 @code{nil}. | |
13229 | |
13230 @item (not (equal fill-prefix "") | |
13231 This expression checks whether an existing fill prefix is an empty | |
13232 string, that is, a string with no characters in it. An empty string is | |
13233 not a useful fill prefix. | |
13234 | |
13235 @item (not paragraph-ignore-fill-prefix) | |
13236 This expression returns @code{nil} if the variable | |
13237 @code{paragraph-ignore-fill-prefix} has been turned on by being set to a | |
13238 true value such as @code{t}. | |
13239 | |
13240 @item (regexp-quote fill-prefix) | |
13241 This is the last argument to the @code{and} special form. If all the | |
13242 arguments to the @code{and} are true, the value resulting from | |
13243 evaluating this expression will be returned by the @code{and} expression | |
13244 and bound to the variable @code{fill-prefix-regexp}, | |
13245 @end table | |
13246 | |
13247 @findex regexp-quote | |
13248 @noindent | |
13249 The result of evaluating this @code{and} expression successfully is that | |
13250 @code{fill-prefix-regexp} will be bound to the value of | |
13251 @code{fill-prefix} as modified by the @code{regexp-quote} function. | |
13252 What @code{regexp-quote} does is read a string and return a regular | |
13253 expression that will exactly match the string and match nothing else. | |
13254 This means that @code{fill-prefix-regexp} will be set to a value that | |
13255 will exactly match the fill prefix if the fill prefix exists. | |
13256 Otherwise, the variable will be set to @code{nil}. | |
13257 | |
13258 The next two local variables in the @code{let*} expression are | |
13259 designed to remove instances of @samp{^} from @code{parstart} and | |
13260 @code{parsep}, the local variables which indicate the paragraph start | |
13261 and the paragraph separator. The next expression sets @code{parsep} | |
13262 again. That is to handle fill prefixes. | |
13263 | |
13264 This is the setting that requires the definition call @code{let*} | |
13265 rather than @code{let}. The true-or-false-test for the @code{if} | |
13266 depends on whether the variable @code{fill-prefix-regexp} evaluates to | |
13267 @code{nil} or some other value. | |
13268 | |
13269 If @code{fill-prefix-regexp} does not have a value, Emacs evaluates | |
13270 the else-part of the @code{if} expression and binds @code{parsep} to | |
13271 its local value. (@code{parsep} is a regular expression that matches | |
13272 what separates paragraphs.) | |
13273 | |
13274 But if @code{fill-prefix-regexp} does have a value, Emacs evaluates | |
13275 the then-part of the @code{if} expression and binds @code{parsep} to a | |
13276 regular expression that includes the @code{fill-prefix-regexp} as part | |
13277 of the pattern. | |
13278 | |
13279 Specifically, @code{parsep} is set to the original value of the | |
13280 paragraph separate regular expression concatenated with an alternative | |
13281 expression that consists of the @code{fill-prefix-regexp} followed by | |
13282 optional whitespace to the end of the line. The whitespace is defined | |
13283 by @w{@code{"[ \t]*$"}}.) The @samp{\\|} defines this portion of the | |
13284 regexp as an alternative to @code{parsep}. | |
13285 | |
13286 According to a comment in the code, the next local variable, | |
13287 @code{sp-parstart}, is used for searching, and then the final two, | |
13288 @code{start} and @code{found-start}, are set to @code{nil}. | |
13289 | |
13290 Now we get into the body of the @code{let*}. The first part of the body | |
13291 of the @code{let*} deals with the case when the function is given a | |
13292 negative argument and is therefore moving backwards. We will skip this | |
13293 section. | |
13294 | |
13295 @node fwd-para while, , fwd-para let, forward-paragraph | |
13296 @unnumberedsubsec The forward motion @code{while} loop | |
13297 | |
13298 The second part of the body of the @code{let*} deals with forward | |
13299 motion. It is a @code{while} loop that repeats itself so long as the | |
13300 value of @code{arg} is greater than zero. In the most common use of | |
13301 the function, the value of the argument is 1, so the body of the | |
13302 @code{while} loop is evaluated exactly once, and the cursor moves | |
13303 forward one paragraph. | |
13304 | |
13305 @ignore | |
13306 (while (and (> arg 0) (not (eobp))) | |
13307 | |
13308 ;; Move forward over separator lines... | |
13309 (while (and (not (eobp)) | |
13310 (progn (move-to-left-margin) (not (eobp))) | |
13311 (looking-at parsep)) | |
13312 (forward-line 1)) | |
13313 (unless (eobp) (setq arg (1- arg))) | |
13314 ;; ... and one more line. | |
13315 (forward-line 1) | |
13316 | |
13317 (if fill-prefix-regexp | |
13318 ;; There is a fill prefix; it overrides parstart. | |
13319 (while (and (not (eobp)) | |
13320 (progn (move-to-left-margin) (not (eobp))) | |
13321 (not (looking-at parsep)) | |
13322 (looking-at fill-prefix-regexp)) | |
13323 (forward-line 1)) | |
13324 | |
13325 (while (and (re-search-forward sp-parstart nil 1) | |
13326 (progn (setq start (match-beginning 0)) | |
13327 (goto-char start) | |
13328 (not (eobp))) | |
13329 (progn (move-to-left-margin) | |
13330 (not (looking-at parsep))) | |
13331 (or (not (looking-at parstart)) | |
13332 (and use-hard-newlines | |
13333 (not (get-text-property (1- start) 'hard))))) | |
13334 (forward-char 1)) | |
13335 | |
13336 (if (< (point) (point-max)) | |
13337 (goto-char start)))) | |
13338 @end ignore | |
13339 | |
13340 This part handles three situations: when point is between paragraphs, | |
13341 when there is a fill prefix and when there is no fill prefix. | |
13342 | |
13343 @need 800 | |
13344 The @code{while} loop looks like this: | |
13345 | |
13346 @smallexample | |
13347 @group | |
13348 ;; @r{going forwards and not at the end of the buffer} | |
13349 (while (and (> arg 0) (not (eobp))) | |
13350 | |
13351 ;; @r{between paragraphs} | |
13352 ;; Move forward over separator lines... | |
13353 (while (and (not (eobp)) | |
13354 (progn (move-to-left-margin) (not (eobp))) | |
13355 (looking-at parsep)) | |
13356 (forward-line 1)) | |
13357 ;; @r{This decrements the loop} | |
13358 (unless (eobp) (setq arg (1- arg))) | |
13359 ;; ... and one more line. | |
13360 (forward-line 1) | |
13361 @end group | |
13362 | |
13363 @group | |
13364 (if fill-prefix-regexp | |
13365 ;; There is a fill prefix; it overrides parstart; | |
13366 ;; we go forward line by line | |
13367 (while (and (not (eobp)) | |
13368 (progn (move-to-left-margin) (not (eobp))) | |
13369 (not (looking-at parsep)) | |
13370 (looking-at fill-prefix-regexp)) | |
13371 (forward-line 1)) | |
13372 @end group | |
13373 | |
13374 @group | |
13375 ;; There is no fill prefix; | |
13376 ;; we go forward character by character | |
13377 (while (and (re-search-forward sp-parstart nil 1) | |
13378 (progn (setq start (match-beginning 0)) | |
13379 (goto-char start) | |
13380 (not (eobp))) | |
13381 (progn (move-to-left-margin) | |
13382 (not (looking-at parsep))) | |
13383 (or (not (looking-at parstart)) | |
13384 (and use-hard-newlines | |
13385 (not (get-text-property (1- start) 'hard))))) | |
13386 (forward-char 1)) | |
13387 @end group | |
13388 | |
13389 @group | |
13390 ;; and if there is no fill prefix and if we are not at the end, | |
13391 ;; go to whatever was found in the regular expression search | |
13392 ;; for sp-parstart | |
13393 (if (< (point) (point-max)) | |
13394 (goto-char start)))) | |
13395 @end group | |
13396 @end smallexample | |
13397 | |
13398 @findex eobp | |
13399 We can see that this is a decrementing counter @code{while} loop, | |
13400 using the expression @code{(setq arg (1- arg))} as the decrementer. | |
13401 That expression is not far from the @code{while}, but is hidden in | |
13402 another Lisp macro, an @code{unless} macro. Unless we are at the end | |
13403 of the buffer --- that is what the @code{eobp} function determines; it | |
13404 is an abbreviation of @samp{End Of Buffer P} --- we decrease the value | |
13405 of @code{arg} by one. | |
13406 | |
13407 (If we are at the end of the buffer, we cannot go forward any more and | |
13408 the next loop of the @code{while} expression will test false since the | |
13409 test is an @code{and} with @code{(not (eobp))}. The @code{not} | |
13410 function means exactly as you expect; it is another name for | |
13411 @code{null}, a function that returns true when its argument is false.) | |
13412 | |
13413 Interestingly, the loop count is not decremented until we leave the | |
13414 space between paragraphs, unless we come to the end of buffer or stop | |
13415 seeing the local value of the paragraph separator. | |
13416 | |
13417 That second @code{while} also has a @code{(move-to-left-margin)} | |
13418 expression. The function is self-explanatory. It is inside a | |
13419 @code{progn} expression and not the last element of its body, so it is | |
13420 only invoked for its side effect, which is to move point to the left | |
13421 margin of the current line. | |
13422 | |
13423 @findex looking-at | |
13424 The @code{looking-at} function is also self-explanatory; it returns | |
13425 true if the text after point matches the regular expression given as | |
13426 its argument. | |
13427 | |
13428 The rest of the body of the loop looks difficult at first, but makes | |
13429 sense as you come to understand it. | |
13430 | |
13431 @need 800 | |
13432 First consider what happens if there is a fill prefix: | |
13433 | |
13434 @smallexample | |
13435 @group | |
13436 (if fill-prefix-regexp | |
13437 ;; There is a fill prefix; it overrides parstart; | |
13438 ;; we go forward line by line | |
13439 (while (and (not (eobp)) | |
13440 (progn (move-to-left-margin) (not (eobp))) | |
13441 (not (looking-at parsep)) | |
13442 (looking-at fill-prefix-regexp)) | |
13443 (forward-line 1)) | |
13444 @end group | |
13445 @end smallexample | |
13446 | |
13447 @noindent | |
13448 This expression moves point forward line by line so long | |
13449 as four conditions are true: | |
13450 | |
13451 @enumerate | |
13452 @item | |
13453 Point is not at the end of the buffer. | |
13454 | |
13455 @item | |
13456 We can move to the left margin of the text and are | |
13457 not at the end of the buffer. | |
13458 | |
13459 @item | |
13460 The text following point does not separate paragraphs. | |
13461 | |
13462 @item | |
13463 The pattern following point is the fill prefix regular expression. | |
13464 @end enumerate | |
13465 | |
13466 The last condition may be puzzling, until you remember that point was | |
13467 moved to the beginning of the line early in the @code{forward-paragraph} | |
13468 function. This means that if the text has a fill prefix, the | |
13469 @code{looking-at} function will see it. | |
13470 | |
13471 @need 1250 | |
13472 Consider what happens when there is no fill prefix. | |
13473 | |
13474 @smallexample | |
13475 @group | |
13476 (while (and (re-search-forward sp-parstart nil 1) | |
13477 (progn (setq start (match-beginning 0)) | |
13478 (goto-char start) | |
13479 (not (eobp))) | |
13480 (progn (move-to-left-margin) | |
13481 (not (looking-at parsep))) | |
13482 (or (not (looking-at parstart)) | |
13483 (and use-hard-newlines | |
13484 (not (get-text-property (1- start) 'hard))))) | |
13485 (forward-char 1)) | |
13486 @end group | |
13487 @end smallexample | |
13488 | |
13489 @noindent | |
13490 This @code{while} loop has us searching forward for | |
13491 @code{sp-parstart}, which is the combination of possible whitespace | |
13492 with a the local value of the start of a paragraph or of a paragraph | |
13493 separator. (The latter two are within an expression starting | |
13494 @code{\(?:} so that they are not referenced by the | |
13495 @code{match-beginning} function.) | |
13496 | |
13497 @need 800 | |
13498 The two expressions, | |
13499 | |
13500 @smallexample | |
13501 @group | |
13502 (setq start (match-beginning 0)) | |
13503 (goto-char start) | |
13504 @end group | |
13505 @end smallexample | |
13506 | |
13507 @noindent | |
13508 mean go to the start of the text matched by the regular expression | |
13509 search. | |
13510 | |
13511 The @code{(match-beginning 0)} expression is new. It returns a number | |
13512 specifying the location of the start of the text that was matched by | |
13513 the last search. | |
13514 | |
13515 The @code{match-beginning} function is used here because of a | |
13516 characteristic of a forward search: a successful forward search, | |
13517 regardless of whether it is a plain search or a regular expression | |
13518 search, moves point to the end of the text that is found. In this | |
13519 case, a successful search moves point to the end of the pattern for | |
13520 @code{sp-parstart}. | |
13521 | |
13522 However, we want to put point at the end of the current paragraph, not | |
13523 somewhere else. Indeed, since the search possibly includes the | |
13524 paragraph separator, point may end up at the beginning of the next one | |
13525 unless we use an expression that includes @code{match-beginning}. | |
13526 | |
13527 @findex match-beginning | |
13528 When given an argument of 0, @code{match-beginning} returns the | |
13529 position that is the start of the text matched by the most recent | |
13530 search. In this case, the most recent search looks for | |
13531 @code{sp-parstart}. The @code{(match-beginning 0)} expression returns | |
13532 the beginning position of that pattern, rather than the end position | |
13533 of that pattern. | |
13534 | |
13535 (Incidentally, when passed a positive number as an argument, the | |
13536 @code{match-beginning} function returns the location of point at that | |
13537 parenthesized expression in the last search unless that parenthesized | |
13538 expression begins with @code{\(?:}. I don't know why @code{\(?:} | |
13539 appears here since the argument is 0.) | |
13540 | |
13541 @need 1250 | |
13542 The last expression when there is no fill prefix is | |
13543 | |
13544 @smallexample | |
13545 @group | |
13546 (if (< (point) (point-max)) | |
13547 (goto-char start)))) | |
13548 @end group | |
13549 @end smallexample | |
13550 | |
13551 @noindent | |
13552 This says that if there is no fill prefix and if we are not at the | |
13553 end, point should move to the beginning of whatever was found by the | |
13554 regular expression search for @code{sp-parstart}. | |
13555 | |
13556 The full definition for the @code{forward-paragraph} function not only | |
13557 includes code for going forwards, but also code for going backwards. | |
13558 | |
13559 If you are reading this inside of GNU Emacs and you want to see the | |
13560 whole function, you can type @kbd{C-h f} (@code{describe-function}) | |
13561 and the name of the function. This gives you the function | |
13562 documentation and the name of the library containing the function's | |
13563 source. Place point over the name of the library and press the RET | |
13564 key; you will be taken directly to the source. (Be sure to install | |
13565 your sources! Without them, you are like a person who tries to drive | |
13566 a car with his eyes shut!) | |
13567 | |
13568 @node etags, Regexp Review, forward-paragraph, Regexp Search | |
13569 @section Create Your Own @file{TAGS} File | |
13570 @findex etags | |
13571 @cindex @file{TAGS} file, create own | |
13572 | |
13573 Besides @kbd{C-h f} (@code{describe-function}), another way to see the | |
13574 source of a function is to type @kbd{M-.} (@code{find-tag}) and the | |
13575 name of the function when prompted for it. This is a good habit to | |
13576 get into. The @kbd{M-.} (@code{find-tag}) command takes you directly | |
13577 to the source for a function, variable, or node. The function depends | |
13578 on tags tables to tell it where to go. | |
13579 | |
13580 If the @code{find-tag} function first asks you for the name of a | |
13581 @file{TAGS} table, give it the name of a @file{TAGS} file such as | |
13582 @file{/usr/local/src/emacs/src/TAGS}. (The exact path to your | |
13583 @file{TAGS} file depends on how your copy of Emacs was installed. I | |
13584 just told you the location that provides both my C and my Emacs Lisp | |
13585 sources.) | |
13586 | |
13587 You can also create your own @file{TAGS} file for directories that | |
13588 lack one. | |
13589 | |
13590 You often need to build and install tags tables yourself. They are | |
13591 not built automatically. A tags table is called a @file{TAGS} file; | |
13592 the name is in upper case letters. | |
13593 | |
13594 You can create a @file{TAGS} file by calling the @code{etags} program | |
13595 that comes as a part of the Emacs distribution. Usually, @code{etags} | |
13596 is compiled and installed when Emacs is built. (@code{etags} is not | |
13597 an Emacs Lisp function or a part of Emacs; it is a C program.) | |
13598 | |
13599 @need 1250 | |
13600 To create a @file{TAGS} file, first switch to the directory in which | |
13601 you want to create the file. In Emacs you can do this with the | |
13602 @kbd{M-x cd} command, or by visiting a file in the directory, or by | |
13603 listing the directory with @kbd{C-x d} (@code{dired}). Then run the | |
13604 compile command, with @w{@code{etags *.el}} as the command to execute | |
13605 | |
13606 @smallexample | |
13607 M-x compile RET etags *.el RET | |
13608 @end smallexample | |
13609 | |
13610 @noindent | |
13611 to create a @file{TAGS} file for Emacs Lisp. | |
13612 | |
13613 For example, if you have a large number of files in your | |
13614 @file{~/emacs} directory, as I do---I have 137 @file{.el} files in it, | |
13615 of which I load 12---you can create a @file{TAGS} file for the Emacs | |
13616 Lisp files in that directory. | |
13617 | |
13618 @need 1250 | |
13619 The @code{etags} program takes all the usual shell `wildcards'. For | |
13620 example, if you have two directories for which you want a single | |
13621 @file{TAGS} file, type @w{@code{etags *.el ../elisp/*.el}}, where | |
13622 @file{../elisp/} is the second directory: | |
13623 | |
13624 @smallexample | |
13625 M-x compile RET etags *.el ../elisp/*.el RET | |
13626 @end smallexample | |
13627 | |
13628 @need 1250 | |
13629 Type | |
13630 | |
13631 @smallexample | |
13632 M-x compile RET etags --help RET | |
13633 @end smallexample | |
13634 | |
13635 @noindent | |
13636 to see a list of the options accepted by @code{etags} as well as a | |
13637 list of supported languages. | |
13638 | |
13639 The @code{etags} program handles more than 20 languages, including | |
13640 Emacs Lisp, Common Lisp, Scheme, C, C++, Ada, Fortran, HTML, Java, | |
13641 LaTeX, Pascal, Perl, Postscript, Python, TeX, Texinfo, makefiles, and | |
13642 most assemblers. The program has no switches for specifying the | |
13643 language; it recognizes the language in an input file according to its | |
13644 file name and contents. | |
13645 | |
13646 @file{etags} is very helpful when you are writing code yourself and | |
13647 want to refer back to functions you have already written. Just run | |
13648 @code{etags} again at intervals as you write new functions, so they | |
13649 become part of the @file{TAGS} file. | |
13650 | |
13651 If you think an appropriate @file{TAGS} file already exists for what | |
13652 you want, but do not know where it is, you can use the @code{locate} | |
13653 program to attempt to find it. | |
13654 | |
13655 Type @w{@kbd{M-x locate @key{RET} TAGS @key{RET}}} and Emacs will list | |
13656 for you the full path names of all your @file{TAGS} files. On my | |
13657 system, this command lists 34 @file{TAGS} files. On the other hand, a | |
13658 `plain vanilla' system I recently installed did not contain any | |
13659 @file{TAGS} files. | |
13660 | |
13661 If the tags table you want has been created, you can use the @code{M-x | |
13662 visit-tags-table} command to specify it. Otherwise, you will need to | |
13663 create the tag table yourself and then use @code{M-x | |
13664 visit-tags-table}. | |
13665 | |
13666 @subsubheading Building Tags in the Emacs sources | |
13667 @cindex Building Tags in the Emacs sources | |
13668 @cindex Tags in the Emacs sources | |
13669 @findex make tags | |
13670 | |
13671 The GNU Emacs sources come with a @file{Makefile} that contains a | |
13672 sophisticated @code{etags} command that creates, collects, and merges | |
13673 tags tables from all over the Emacs sources and puts the information | |
13674 into one @file{TAGS} file in the @file{src/} directory. (The | |
13675 @file{src/} directory is below the top level of your Emacs directory.) | |
13676 | |
13677 @need 1250 | |
13678 To build this @file{TAGS} file, go to the top level of your Emacs | |
13679 source directory and run the compile command @code{make tags}: | |
13680 | |
13681 @smallexample | |
13682 M-x compile RET make tags RET | |
13683 @end smallexample | |
13684 | |
13685 @noindent | |
13686 (The @code{make tags} command works well with the GNU Emacs sources, | |
13687 as well as with some other source packages.) | |
13688 | |
13689 For more information, see @ref{Tags, , Tag Tables, emacs, The GNU Emacs | |
13690 Manual}. | |
13691 | |
13692 @node Regexp Review, re-search Exercises, etags, Regexp Search | |
13693 @comment node-name, next, previous, up | |
13694 @section Review | |
13695 | |
13696 Here is a brief summary of some recently introduced functions. | |
13697 | |
13698 @table @code | |
13699 @item while | |
13700 Repeatedly evaluate the body of the expression so long as the first | |
13701 element of the body tests true. Then return @code{nil}. (The | |
13702 expression is evaluated only for its side effects.) | |
13703 | |
13704 @need 1250 | |
13705 For example: | |
13706 | |
13707 @smallexample | |
13708 @group | |
13709 (let ((foo 2)) | |
13710 (while (> foo 0) | |
13711 (insert (format "foo is %d.\n" foo)) | |
13712 (setq foo (1- foo)))) | |
13713 | |
13714 @result{} foo is 2. | |
13715 foo is 1. | |
13716 nil | |
13717 @end group | |
13718 @end smallexample | |
13719 | |
13720 @noindent | |
13721 (The @code{insert} function inserts its arguments at point; the | |
13722 @code{format} function returns a string formatted from its arguments | |
13723 the way @code{message} formats its arguments; @code{\n} produces a new | |
13724 line.) | |
13725 | |
13726 @item re-search-forward | |
13727 Search for a pattern, and if the pattern is found, move point to rest | |
13728 just after it. | |
13729 | |
13730 @noindent | |
13731 Takes four arguments, like @code{search-forward}: | |
13732 | |
13733 @enumerate | |
13734 @item | |
13735 A regular expression that specifies the pattern to search for. | |
13736 (Remember to put quotation marks around this argument!) | |
13737 | |
13738 @item | |
13739 Optionally, the limit of the search. | |
13740 | |
13741 @item | |
13742 Optionally, what to do if the search fails, return @code{nil} or an | |
13743 error message. | |
13744 | |
13745 @item | |
13746 Optionally, how many times to repeat the search; if negative, the | |
13747 search goes backwards. | |
13748 @end enumerate | |
13749 | |
13750 @item let* | |
13751 Bind some variables locally to particular values, | |
13752 and then evaluate the remaining arguments, returning the value of the | |
13753 last one. While binding the local variables, use the local values of | |
13754 variables bound earlier, if any. | |
13755 | |
13756 @need 1250 | |
13757 For example: | |
13758 | |
13759 @smallexample | |
13760 @group | |
13761 (let* ((foo 7) | |
13762 (bar (* 3 foo))) | |
13763 (message "`bar' is %d." bar)) | |
13764 @result{} `bar' is 21. | |
13765 @end group | |
13766 @end smallexample | |
13767 | |
13768 @item match-beginning | |
13769 Return the position of the start of the text found by the last regular | |
13770 expression search. | |
13771 | |
13772 @item looking-at | |
13773 Return @code{t} for true if the text after point matches the argument, | |
13774 which should be a regular expression. | |
13775 | |
13776 @item eobp | |
13777 Return @code{t} for true if point is at the end of the accessible part | |
13778 of a buffer. The end of the accessible part is the end of the buffer | |
13779 if the buffer is not narrowed; it is the end of the narrowed part if | |
13780 the buffer is narrowed. | |
13781 @end table | |
13782 | |
13783 @need 1500 | |
13784 @node re-search Exercises, , Regexp Review, Regexp Search | |
13785 @section Exercises with @code{re-search-forward} | |
13786 | |
13787 @itemize @bullet | |
13788 @item | |
13789 Write a function to search for a regular expression that matches two | |
13790 or more blank lines in sequence. | |
13791 | |
13792 @item | |
13793 Write a function to search for duplicated words, such as `the the'. | |
13794 @xref{Regexps, , Syntax of Regular Expressions, emacs, The GNU Emacs | |
13795 Manual}, for information on how to write a regexp (a regular | |
13796 expression) to match a string that is composed of two identical | |
13797 halves. You can devise several regexps; some are better than others. | |
13798 The function I use is described in an appendix, along with several | |
13799 regexps. @xref{the-the, , @code{the-the} Duplicated Words Function}. | |
13800 @end itemize | |
13801 | |
13802 @node Counting Words, Words in a defun, Regexp Search, Top | |
13803 @chapter Counting: Repetition and Regexps | |
13804 @cindex Repetition for word counting | |
13805 @cindex Regular expressions for word counting | |
13806 | |
13807 Repetition and regular expression searches are powerful tools that you | |
13808 often use when you write code in Emacs Lisp. This chapter illustrates | |
13809 the use of regular expression searches through the construction of | |
13810 word count commands using @code{while} loops and recursion. | |
13811 | |
13812 @menu | |
13813 * Why Count Words:: | |
13814 * count-words-region:: Use a regexp, but find a problem. | |
13815 * recursive-count-words:: Start with case of no words in region. | |
13816 * Counting Exercise:: | |
13817 @end menu | |
13818 | |
13819 @node Why Count Words, count-words-region, Counting Words, Counting Words | |
13820 @ifnottex | |
13821 @unnumberedsec Counting words | |
13822 @end ifnottex | |
13823 | |
13824 The standard Emacs distribution contains a function for counting the | |
13825 number of lines within a region. However, there is no corresponding | |
13826 function for counting words. | |
13827 | |
13828 Certain types of writing ask you to count words. Thus, if you write | |
13829 an essay, you may be limited to 800 words; if you write a novel, you | |
13830 may discipline yourself to write 1000 words a day. It seems odd to me | |
13831 that Emacs lacks a word count command. Perhaps people use Emacs | |
13832 mostly for code or types of documentation that do not require word | |
13833 counts; or perhaps they restrict themselves to the operating system | |
13834 word count command, @code{wc}. Alternatively, people may follow | |
13835 the publishers' convention and compute a word count by dividing the | |
13836 number of characters in a document by five. In any event, here are | |
13837 commands to count words. | |
13838 | |
13839 @node count-words-region, recursive-count-words, Why Count Words, Counting Words | |
13840 @comment node-name, next, previous, up | |
13841 @section The @code{count-words-region} Function | |
13842 @findex count-words-region | |
13843 | |
13844 A word count command could count words in a line, paragraph, region, | |
13845 or buffer. What should the command cover? You could design the | |
13846 command to count the number of words in a complete buffer. However, | |
13847 the Emacs tradition encourages flexibility---you may want to count | |
13848 words in just a section, rather than all of a buffer. So it makes | |
13849 more sense to design the command to count the number of words in a | |
13850 region. Once you have a @code{count-words-region} command, you can, | |
13851 if you wish, count words in a whole buffer by marking it with | |
13852 @w{@kbd{C-x h}} (@code{mark-whole-buffer}). | |
13853 | |
13854 Clearly, counting words is a repetitive act: starting from the | |
13855 beginning of the region, you count the first word, then the second | |
13856 word, then the third word, and so on, until you reach the end of the | |
13857 region. This means that word counting is ideally suited to recursion | |
13858 or to a @code{while} loop. | |
13859 | |
13860 @menu | |
13861 * Design count-words-region:: The definition using a @code{while} loop. | |
13862 * Whitespace Bug:: The Whitespace Bug in @code{count-words-region}. | |
13863 @end menu | |
13864 | |
13865 @node Design count-words-region, Whitespace Bug, count-words-region, count-words-region | |
13866 @ifnottex | |
13867 @unnumberedsubsec Designing @code{count-words-region} | |
13868 @end ifnottex | |
13869 | |
13870 First, we will implement the word count command with a @code{while} | |
13871 loop, then with recursion. The command will, of course, be | |
13872 interactive. | |
13873 | |
13874 @need 800 | |
13875 The template for an interactive function definition is, as always: | |
13876 | |
13877 @smallexample | |
13878 @group | |
13879 (defun @var{name-of-function} (@var{argument-list}) | |
13880 "@var{documentation}@dots{}" | |
13881 (@var{interactive-expression}@dots{}) | |
13882 @var{body}@dots{}) | |
13883 @end group | |
13884 @end smallexample | |
13885 | |
13886 What we need to do is fill in the slots. | |
13887 | |
13888 The name of the function should be self-explanatory and similar to the | |
13889 existing @code{count-lines-region} name. This makes the name easier | |
13890 to remember. @code{count-words-region} is a good choice. | |
13891 | |
13892 The function counts words within a region. This means that the | |
13893 argument list must contain symbols that are bound to the two | |
13894 positions, the beginning and end of the region. These two positions | |
13895 can be called @samp{beginning} and @samp{end} respectively. The first | |
13896 line of the documentation should be a single sentence, since that is | |
13897 all that is printed as documentation by a command such as | |
13898 @code{apropos}. The interactive expression will be of the form | |
13899 @samp{(interactive "r")}, since that will cause Emacs to pass the | |
13900 beginning and end of the region to the function's argument list. All | |
13901 this is routine. | |
13902 | |
13903 The body of the function needs to be written to do three tasks: | |
13904 first, to set up conditions under which the @code{while} loop can | |
13905 count words, second, to run the @code{while} loop, and third, to send | |
13906 a message to the user. | |
13907 | |
13908 When a user calls @code{count-words-region}, point may be at the | |
13909 beginning or the end of the region. However, the counting process | |
13910 must start at the beginning of the region. This means we will want | |
13911 to put point there if it is not already there. Executing | |
13912 @code{(goto-char beginning)} ensures this. Of course, we will want to | |
13913 return point to its expected position when the function finishes its | |
13914 work. For this reason, the body must be enclosed in a | |
13915 @code{save-excursion} expression. | |
13916 | |
13917 The central part of the body of the function consists of a | |
13918 @code{while} loop in which one expression jumps point forward word by | |
13919 word, and another expression counts those jumps. The true-or-false-test | |
13920 of the @code{while} loop should test true so long as point should jump | |
13921 forward, and false when point is at the end of the region. | |
13922 | |
13923 We could use @code{(forward-word 1)} as the expression for moving point | |
13924 forward word by word, but it is easier to see what Emacs identifies as a | |
13925 `word' if we use a regular expression search. | |
13926 | |
13927 A regular expression search that finds the pattern for which it is | |
13928 searching leaves point after the last character matched. This means | |
13929 that a succession of successful word searches will move point forward | |
13930 word by word. | |
13931 | |
13932 As a practical matter, we want the regular expression search to jump | |
13933 over whitespace and punctuation between words as well as over the | |
13934 words themselves. A regexp that refuses to jump over interword | |
13935 whitespace would never jump more than one word! This means that | |
13936 the regexp should include the whitespace and punctuation that follows | |
13937 a word, if any, as well as the word itself. (A word may end a buffer | |
13938 and not have any following whitespace or punctuation, so that part of | |
13939 the regexp must be optional.) | |
13940 | |
13941 Thus, what we want for the regexp is a pattern defining one or more | |
13942 word constituent characters followed, optionally, by one or more | |
13943 characters that are not word constituents. The regular expression for | |
13944 this is: | |
13945 | |
13946 @smallexample | |
13947 \w+\W* | |
13948 @end smallexample | |
13949 | |
13950 @noindent | |
13951 The buffer's syntax table determines which characters are and are not | |
13952 word constituents. (@xref{Syntax, , What Constitutes a Word or | |
13953 Symbol?}, for more about syntax. Also, see @ref{Syntax, Syntax, The | |
13954 Syntax Table, emacs, The GNU Emacs Manual}, and @ref{Syntax Tables, , | |
13955 Syntax Tables, elisp, The GNU Emacs Lisp Reference Manual}.) | |
13956 | |
13957 @need 800 | |
13958 The search expression looks like this: | |
13959 | |
13960 @smallexample | |
13961 (re-search-forward "\\w+\\W*") | |
13962 @end smallexample | |
13963 | |
13964 @noindent | |
13965 (Note that paired backslashes precede the @samp{w} and @samp{W}. A | |
13966 single backslash has special meaning to the Emacs Lisp interpreter. | |
13967 It indicates that the following character is interpreted differently | |
13968 than usual. For example, the two characters, @samp{\n}, stand for | |
13969 @samp{newline}, rather than for a backslash followed by @samp{n}. Two | |
13970 backslashes in a row stand for an ordinary, `unspecial' backslash, so | |
13971 Emacs Lisp interpreter ends of seeing a single backslash followed by a | |
13972 letter. So it discovers the letter is special.) | |
13973 | |
13974 We need a counter to count how many words there are; this variable | |
13975 must first be set to 0 and then incremented each time Emacs goes | |
13976 around the @code{while} loop. The incrementing expression is simply: | |
13977 | |
13978 @smallexample | |
13979 (setq count (1+ count)) | |
13980 @end smallexample | |
13981 | |
13982 Finally, we want to tell the user how many words there are in the | |
13983 region. The @code{message} function is intended for presenting this | |
13984 kind of information to the user. The message has to be phrased so | |
13985 that it reads properly regardless of how many words there are in the | |
13986 region: we don't want to say that ``there are 1 words in the region''. | |
13987 The conflict between singular and plural is ungrammatical. We can | |
13988 solve this problem by using a conditional expression that evaluates | |
13989 different messages depending on the number of words in the region. | |
13990 There are three possibilities: no words in the region, one word in the | |
13991 region, and more than one word. This means that the @code{cond} | |
13992 special form is appropriate. | |
13993 | |
13994 @need 1500 | |
13995 All this leads to the following function definition: | |
13996 | |
13997 @smallexample | |
13998 @group | |
13999 ;;; @r{First version; has bugs!} | |
14000 (defun count-words-region (beginning end) | |
14001 "Print number of words in the region. | |
14002 Words are defined as at least one word-constituent | |
14003 character followed by at least one character that | |
14004 is not a word-constituent. The buffer's syntax | |
14005 table determines which characters these are." | |
14006 (interactive "r") | |
14007 (message "Counting words in region ... ") | |
14008 @end group | |
14009 | |
14010 @group | |
14011 ;;; @r{1. Set up appropriate conditions.} | |
14012 (save-excursion | |
14013 (goto-char beginning) | |
14014 (let ((count 0)) | |
14015 @end group | |
14016 | |
14017 @group | |
14018 ;;; @r{2. Run the} while @r{loop.} | |
14019 (while (< (point) end) | |
14020 (re-search-forward "\\w+\\W*") | |
14021 (setq count (1+ count))) | |
14022 @end group | |
14023 | |
14024 @group | |
14025 ;;; @r{3. Send a message to the user.} | |
14026 (cond ((zerop count) | |
14027 (message | |
14028 "The region does NOT have any words.")) | |
14029 ((= 1 count) | |
14030 (message | |
14031 "The region has 1 word.")) | |
14032 (t | |
14033 (message | |
14034 "The region has %d words." count)))))) | |
14035 @end group | |
14036 @end smallexample | |
14037 | |
14038 @noindent | |
14039 As written, the function works, but not in all circumstances. | |
14040 | |
14041 @node Whitespace Bug, , Design count-words-region, count-words-region | |
14042 @comment node-name, next, previous, up | |
14043 @subsection The Whitespace Bug in @code{count-words-region} | |
14044 | |
14045 The @code{count-words-region} command described in the preceding | |
14046 section has two bugs, or rather, one bug with two manifestations. | |
14047 First, if you mark a region containing only whitespace in the middle | |
14048 of some text, the @code{count-words-region} command tells you that the | |
14049 region contains one word! Second, if you mark a region containing | |
14050 only whitespace at the end of the buffer or the accessible portion of | |
14051 a narrowed buffer, the command displays an error message that looks | |
14052 like this: | |
14053 | |
14054 @smallexample | |
14055 Search failed: "\\w+\\W*" | |
14056 @end smallexample | |
14057 | |
14058 If you are reading this in Info in GNU Emacs, you can test for these | |
14059 bugs yourself. | |
14060 | |
14061 First, evaluate the function in the usual manner to install it. | |
14062 @ifinfo | |
14063 Here is a copy of the definition. Place your cursor after the closing | |
14064 parenthesis and type @kbd{C-x C-e} to install it. | |
14065 | |
14066 @smallexample | |
14067 @group | |
14068 ;; @r{First version; has bugs!} | |
14069 (defun count-words-region (beginning end) | |
14070 "Print number of words in the region. | |
14071 Words are defined as at least one word-constituent character followed | |
14072 by at least one character that is not a word-constituent. The buffer's | |
14073 syntax table determines which characters these are." | |
14074 @end group | |
14075 @group | |
14076 (interactive "r") | |
14077 (message "Counting words in region ... ") | |
14078 @end group | |
14079 | |
14080 @group | |
14081 ;;; @r{1. Set up appropriate conditions.} | |
14082 (save-excursion | |
14083 (goto-char beginning) | |
14084 (let ((count 0)) | |
14085 @end group | |
14086 | |
14087 @group | |
14088 ;;; @r{2. Run the} while @r{loop.} | |
14089 (while (< (point) end) | |
14090 (re-search-forward "\\w+\\W*") | |
14091 (setq count (1+ count))) | |
14092 @end group | |
14093 | |
14094 @group | |
14095 ;;; @r{3. Send a message to the user.} | |
14096 (cond ((zerop count) | |
14097 (message "The region does NOT have any words.")) | |
14098 ((= 1 count) (message "The region has 1 word.")) | |
14099 (t (message "The region has %d words." count)))))) | |
14100 @end group | |
14101 @end smallexample | |
14102 @end ifinfo | |
14103 | |
14104 @need 1000 | |
14105 If you wish, you can also install this keybinding by evaluating it: | |
14106 | |
14107 @smallexample | |
14108 (global-set-key "\C-c=" 'count-words-region) | |
14109 @end smallexample | |
14110 | |
14111 To conduct the first test, set mark and point to the beginning and end | |
14112 of the following line and then type @kbd{C-c =} (or @kbd{M-x | |
14113 count-words-region} if you have not bound @kbd{C-c =}): | |
14114 | |
14115 @smallexample | |
14116 one two three | |
14117 @end smallexample | |
14118 | |
14119 @noindent | |
14120 Emacs will tell you, correctly, that the region has three words. | |
14121 | |
14122 Repeat the test, but place mark at the beginning of the line and place | |
14123 point just @emph{before} the word @samp{one}. Again type the command | |
14124 @kbd{C-c =} (or @kbd{M-x count-words-region}). Emacs should tell you | |
14125 that the region has no words, since it is composed only of the | |
14126 whitespace at the beginning of the line. But instead Emacs tells you | |
14127 that the region has one word! | |
14128 | |
14129 For the third test, copy the sample line to the end of the | |
14130 @file{*scratch*} buffer and then type several spaces at the end of the | |
14131 line. Place mark right after the word @samp{three} and point at the | |
14132 end of line. (The end of the line will be the end of the buffer.) | |
14133 Type @kbd{C-c =} (or @kbd{M-x count-words-region}) as you did before. | |
14134 Again, Emacs should tell you that the region has no words, since it is | |
14135 composed only of the whitespace at the end of the line. Instead, | |
14136 Emacs displays an error message saying @samp{Search failed}. | |
14137 | |
14138 The two bugs stem from the same problem. | |
14139 | |
14140 Consider the first manifestation of the bug, in which the command | |
14141 tells you that the whitespace at the beginning of the line contains | |
14142 one word. What happens is this: The @code{M-x count-words-region} | |
14143 command moves point to the beginning of the region. The @code{while} | |
14144 tests whether the value of point is smaller than the value of | |
14145 @code{end}, which it is. Consequently, the regular expression search | |
14146 looks for and finds the first word. It leaves point after the word. | |
14147 @code{count} is set to one. The @code{while} loop repeats; but this | |
14148 time the value of point is larger than the value of @code{end}, the | |
14149 loop is exited; and the function displays a message saying the number | |
14150 of words in the region is one. In brief, the regular expression | |
14151 search looks for and finds the word even though it is outside | |
14152 the marked region. | |
14153 | |
14154 In the second manifestation of the bug, the region is whitespace at | |
14155 the end of the buffer. Emacs says @samp{Search failed}. What happens | |
14156 is that the true-or-false-test in the @code{while} loop tests true, so | |
14157 the search expression is executed. But since there are no more words | |
14158 in the buffer, the search fails. | |
14159 | |
14160 In both manifestations of the bug, the search extends or attempts to | |
14161 extend outside of the region. | |
14162 | |
14163 The solution is to limit the search to the region---this is a fairly | |
14164 simple action, but as you may have come to expect, it is not quite as | |
14165 simple as you might think. | |
14166 | |
14167 As we have seen, the @code{re-search-forward} function takes a search | |
14168 pattern as its first argument. But in addition to this first, | |
14169 mandatory argument, it accepts three optional arguments. The optional | |
14170 second argument bounds the search. The optional third argument, if | |
14171 @code{t}, causes the function to return @code{nil} rather than signal | |
14172 an error if the search fails. The optional fourth argument is a | |
14173 repeat count. (In Emacs, you can see a function's documentation by | |
14174 typing @kbd{C-h f}, the name of the function, and then @key{RET}.) | |
14175 | |
14176 In the @code{count-words-region} definition, the value of the end of | |
14177 the region is held by the variable @code{end} which is passed as an | |
14178 argument to the function. Thus, we can add @code{end} as an argument | |
14179 to the regular expression search expression: | |
14180 | |
14181 @smallexample | |
14182 (re-search-forward "\\w+\\W*" end) | |
14183 @end smallexample | |
14184 | |
14185 However, if you make only this change to the @code{count-words-region} | |
14186 definition and then test the new version of the definition on a | |
14187 stretch of whitespace, you will receive an error message saying | |
14188 @samp{Search failed}. | |
14189 | |
14190 What happens is this: the search is limited to the region, and fails | |
14191 as you expect because there are no word-constituent characters in the | |
14192 region. Since it fails, we receive an error message. But we do not | |
14193 want to receive an error message in this case; we want to receive the | |
14194 message that "The region does NOT have any words." | |
14195 | |
14196 The solution to this problem is to provide @code{re-search-forward} | |
14197 with a third argument of @code{t}, which causes the function to return | |
14198 @code{nil} rather than signal an error if the search fails. | |
14199 | |
14200 However, if you make this change and try it, you will see the message | |
14201 ``Counting words in region ... '' and @dots{} you will keep on seeing | |
14202 that message @dots{}, until you type @kbd{C-g} (@code{keyboard-quit}). | |
14203 | |
14204 Here is what happens: the search is limited to the region, as before, | |
14205 and it fails because there are no word-constituent characters in the | |
14206 region, as expected. Consequently, the @code{re-search-forward} | |
14207 expression returns @code{nil}. It does nothing else. In particular, | |
14208 it does not move point, which it does as a side effect if it finds the | |
14209 search target. After the @code{re-search-forward} expression returns | |
14210 @code{nil}, the next expression in the @code{while} loop is evaluated. | |
14211 This expression increments the count. Then the loop repeats. The | |
14212 true-or-false-test tests true because the value of point is still less | |
14213 than the value of end, since the @code{re-search-forward} expression | |
14214 did not move point. @dots{} and the cycle repeats @dots{} | |
14215 | |
14216 The @code{count-words-region} definition requires yet another | |
14217 modification, to cause the true-or-false-test of the @code{while} loop | |
14218 to test false if the search fails. Put another way, there are two | |
14219 conditions that must be satisfied in the true-or-false-test before the | |
14220 word count variable is incremented: point must still be within the | |
14221 region and the search expression must have found a word to count. | |
14222 | |
14223 Since both the first condition and the second condition must be true | |
14224 together, the two expressions, the region test and the search | |
14225 expression, can be joined with an @code{and} special form and embedded in | |
14226 the @code{while} loop as the true-or-false-test, like this: | |
14227 | |
14228 @smallexample | |
14229 (and (< (point) end) (re-search-forward "\\w+\\W*" end t)) | |
14230 @end smallexample | |
14231 | |
14232 @c colon in printed section title causes problem in Info cross reference | |
14233 @c also trouble with an overfull hbox | |
14234 @iftex | |
14235 @noindent | |
14236 (For information about @code{and}, see | |
14237 @ref{kill-new function, , The @code{kill-new} function}.) | |
14238 @end iftex | |
14239 @ifinfo | |
14240 @noindent | |
14241 (@xref{kill-new function, , The @code{kill-new} function}, for | |
14242 information about @code{and}.) | |
14243 @end ifinfo | |
14244 | |
14245 The @code{re-search-forward} expression returns @code{t} if the search | |
14246 succeeds and as a side effect moves point. Consequently, as words are | |
14247 found, point is moved through the region. When the search expression | |
14248 fails to find another word, or when point reaches the end of the | |
14249 region, the true-or-false-test tests false, the @code{while} loop | |
14250 exits, and the @code{count-words-region} function displays one or | |
14251 other of its messages. | |
14252 | |
14253 After incorporating these final changes, the @code{count-words-region} | |
14254 works without bugs (or at least, without bugs that I have found!). | |
14255 Here is what it looks like: | |
14256 | |
14257 @smallexample | |
14258 @group | |
14259 ;;; @r{Final version:} @code{while} | |
14260 (defun count-words-region (beginning end) | |
14261 "Print number of words in the region." | |
14262 (interactive "r") | |
14263 (message "Counting words in region ... ") | |
14264 @end group | |
14265 | |
14266 @group | |
14267 ;;; @r{1. Set up appropriate conditions.} | |
14268 (save-excursion | |
14269 (let ((count 0)) | |
14270 (goto-char beginning) | |
14271 @end group | |
14272 | |
14273 @group | |
14274 ;;; @r{2. Run the} while @r{loop.} | |
14275 (while (and (< (point) end) | |
14276 (re-search-forward "\\w+\\W*" end t)) | |
14277 (setq count (1+ count))) | |
14278 @end group | |
14279 | |
14280 @group | |
14281 ;;; @r{3. Send a message to the user.} | |
14282 (cond ((zerop count) | |
14283 (message | |
14284 "The region does NOT have any words.")) | |
14285 ((= 1 count) | |
14286 (message | |
14287 "The region has 1 word.")) | |
14288 (t | |
14289 (message | |
14290 "The region has %d words." count)))))) | |
14291 @end group | |
14292 @end smallexample | |
14293 | |
14294 @node recursive-count-words, Counting Exercise, count-words-region, Counting Words | |
14295 @comment node-name, next, previous, up | |
14296 @section Count Words Recursively | |
14297 @cindex Count words recursively | |
14298 @cindex Recursively counting words | |
14299 @cindex Words, counted recursively | |
14300 | |
14301 You can write the function for counting words recursively as well as | |
14302 with a @code{while} loop. Let's see how this is done. | |
14303 | |
14304 First, we need to recognize that the @code{count-words-region} | |
14305 function has three jobs: it sets up the appropriate conditions for | |
14306 counting to occur; it counts the words in the region; and it sends a | |
14307 message to the user telling how many words there are. | |
14308 | |
14309 If we write a single recursive function to do everything, we will | |
14310 receive a message for every recursive call. If the region contains 13 | |
14311 words, we will receive thirteen messages, one right after the other. | |
14312 We don't want this! Instead, we must write two functions to do the | |
14313 job, one of which (the recursive function) will be used inside of the | |
14314 other. One function will set up the conditions and display the | |
14315 message; the other will return the word count. | |
14316 | |
14317 Let us start with the function that causes the message to be displayed. | |
14318 We can continue to call this @code{count-words-region}. | |
14319 | |
14320 This is the function that the user will call. It will be interactive. | |
14321 Indeed, it will be similar to our previous versions of this | |
14322 function, except that it will call @code{recursive-count-words} to | |
14323 determine how many words are in the region. | |
14324 | |
14325 @need 1250 | |
14326 We can readily construct a template for this function, based on our | |
14327 previous versions: | |
14328 | |
14329 @smallexample | |
14330 @group | |
14331 ;; @r{Recursive version; uses regular expression search} | |
14332 (defun count-words-region (beginning end) | |
14333 "@var{documentation}@dots{}" | |
14334 (@var{interactive-expression}@dots{}) | |
14335 @end group | |
14336 @group | |
14337 | |
14338 ;;; @r{1. Set up appropriate conditions.} | |
14339 (@var{explanatory message}) | |
14340 (@var{set-up functions}@dots{} | |
14341 @end group | |
14342 @group | |
14343 | |
14344 ;;; @r{2. Count the words.} | |
14345 @var{recursive call} | |
14346 @end group | |
14347 @group | |
14348 | |
14349 ;;; @r{3. Send a message to the user.} | |
14350 @var{message providing word count})) | |
14351 @end group | |
14352 @end smallexample | |
14353 | |
14354 The definition looks straightforward, except that somehow the count | |
14355 returned by the recursive call must be passed to the message | |
14356 displaying the word count. A little thought suggests that this can be | |
14357 done by making use of a @code{let} expression: we can bind a variable | |
14358 in the varlist of a @code{let} expression to the number of words in | |
14359 the region, as returned by the recursive call; and then the | |
14360 @code{cond} expression, using binding, can display the value to the | |
14361 user. | |
14362 | |
14363 Often, one thinks of the binding within a @code{let} expression as | |
14364 somehow secondary to the `primary' work of a function. But in this | |
14365 case, what you might consider the `primary' job of the function, | |
14366 counting words, is done within the @code{let} expression. | |
14367 | |
14368 @need 1250 | |
14369 Using @code{let}, the function definition looks like this: | |
14370 | |
14371 @smallexample | |
14372 @group | |
14373 (defun count-words-region (beginning end) | |
14374 "Print number of words in the region." | |
14375 (interactive "r") | |
14376 @end group | |
14377 | |
14378 @group | |
14379 ;;; @r{1. Set up appropriate conditions.} | |
14380 (message "Counting words in region ... ") | |
14381 (save-excursion | |
14382 (goto-char beginning) | |
14383 @end group | |
14384 | |
14385 @group | |
14386 ;;; @r{2. Count the words.} | |
14387 (let ((count (recursive-count-words end))) | |
14388 @end group | |
14389 | |
14390 @group | |
14391 ;;; @r{3. Send a message to the user.} | |
14392 (cond ((zerop count) | |
14393 (message | |
14394 "The region does NOT have any words.")) | |
14395 ((= 1 count) | |
14396 (message | |
14397 "The region has 1 word.")) | |
14398 (t | |
14399 (message | |
14400 "The region has %d words." count)))))) | |
14401 @end group | |
14402 @end smallexample | |
14403 | |
14404 Next, we need to write the recursive counting function. | |
14405 | |
14406 A recursive function has at least three parts: the `do-again-test', the | |
14407 `next-step-expression', and the recursive call. | |
14408 | |
14409 The do-again-test determines whether the function will or will not be | |
14410 called again. Since we are counting words in a region and can use a | |
14411 function that moves point forward for every word, the do-again-test | |
14412 can check whether point is still within the region. The do-again-test | |
14413 should find the value of point and determine whether point is before, | |
14414 at, or after the value of the end of the region. We can use the | |
14415 @code{point} function to locate point. Clearly, we must pass the | |
14416 value of the end of the region to the recursive counting function as an | |
14417 argument. | |
14418 | |
14419 In addition, the do-again-test should also test whether the search finds a | |
14420 word. If it does not, the function should not call itself again. | |
14421 | |
14422 The next-step-expression changes a value so that when the recursive | |
14423 function is supposed to stop calling itself, it stops. More | |
14424 precisely, the next-step-expression changes a value so that at the | |
14425 right time, the do-again-test stops the recursive function from | |
14426 calling itself again. In this case, the next-step-expression can be | |
14427 the expression that moves point forward, word by word. | |
14428 | |
14429 The third part of a recursive function is the recursive call. | |
14430 | |
14431 Somewhere, also, we also need a part that does the `work' of the | |
14432 function, a part that does the counting. A vital part! | |
14433 | |
14434 @need 1250 | |
14435 But already, we have an outline of the recursive counting function: | |
14436 | |
14437 @smallexample | |
14438 @group | |
14439 (defun recursive-count-words (region-end) | |
14440 "@var{documentation}@dots{}" | |
14441 @var{do-again-test} | |
14442 @var{next-step-expression} | |
14443 @var{recursive call}) | |
14444 @end group | |
14445 @end smallexample | |
14446 | |
14447 Now we need to fill in the slots. Let's start with the simplest cases | |
14448 first: if point is at or beyond the end of the region, there cannot | |
14449 be any words in the region, so the function should return zero. | |
14450 Likewise, if the search fails, there are no words to count, so the | |
14451 function should return zero. | |
14452 | |
14453 On the other hand, if point is within the region and the search | |
14454 succeeds, the function should call itself again. | |
14455 | |
14456 @need 800 | |
14457 Thus, the do-again-test should look like this: | |
14458 | |
14459 @smallexample | |
14460 @group | |
14461 (and (< (point) region-end) | |
14462 (re-search-forward "\\w+\\W*" region-end t)) | |
14463 @end group | |
14464 @end smallexample | |
14465 | |
14466 Note that the search expression is part of the do-again-test---the | |
14467 function returns @code{t} if its search succeeds and @code{nil} if it | |
14468 fails. (@xref{Whitespace Bug, , The Whitespace Bug in | |
14469 @code{count-words-region}}, for an explanation of how | |
14470 @code{re-search-forward} works.) | |
14471 | |
14472 The do-again-test is the true-or-false test of an @code{if} clause. | |
14473 Clearly, if the do-again-test succeeds, the then-part of the @code{if} | |
14474 clause should call the function again; but if it fails, the else-part | |
14475 should return zero since either point is outside the region or the | |
14476 search failed because there were no words to find. | |
14477 | |
14478 But before considering the recursive call, we need to consider the | |
14479 next-step-expression. What is it? Interestingly, it is the search | |
14480 part of the do-again-test. | |
14481 | |
14482 In addition to returning @code{t} or @code{nil} for the | |
14483 do-again-test, @code{re-search-forward} moves point forward as a side | |
14484 effect of a successful search. This is the action that changes the | |
14485 value of point so that the recursive function stops calling itself | |
14486 when point completes its movement through the region. Consequently, | |
14487 the @code{re-search-forward} expression is the next-step-expression. | |
14488 | |
14489 @need 1200 | |
14490 In outline, then, the body of the @code{recursive-count-words} | |
14491 function looks like this: | |
14492 | |
14493 @smallexample | |
14494 @group | |
14495 (if @var{do-again-test-and-next-step-combined} | |
14496 ;; @r{then} | |
14497 @var{recursive-call-returning-count} | |
14498 ;; @r{else} | |
14499 @var{return-zero}) | |
14500 @end group | |
14501 @end smallexample | |
14502 | |
14503 How to incorporate the mechanism that counts? | |
14504 | |
14505 If you are not used to writing recursive functions, a question like | |
14506 this can be troublesome. But it can and should be approached | |
14507 systematically. | |
14508 | |
14509 We know that the counting mechanism should be associated in some way | |
14510 with the recursive call. Indeed, since the next-step-expression moves | |
14511 point forward by one word, and since a recursive call is made for | |
14512 each word, the counting mechanism must be an expression that adds one | |
14513 to the value returned by a call to @code{recursive-count-words}. | |
14514 | |
14515 @need 800 | |
14516 Consider several cases: | |
14517 | |
14518 @itemize @bullet | |
14519 @item | |
14520 If there are two words in the region, the function should return | |
14521 a value resulting from adding one to the value returned when it counts | |
14522 the first word, plus the number returned when it counts the remaining | |
14523 words in the region, which in this case is one. | |
14524 | |
14525 @item | |
14526 If there is one word in the region, the function should return | |
14527 a value resulting from adding one to the value returned when it counts | |
14528 that word, plus the number returned when it counts the remaining | |
14529 words in the region, which in this case is zero. | |
14530 | |
14531 @item | |
14532 If there are no words in the region, the function should return zero. | |
14533 @end itemize | |
14534 | |
14535 From the sketch we can see that the else-part of the @code{if} returns | |
14536 zero for the case of no words. This means that the then-part of the | |
14537 @code{if} must return a value resulting from adding one to the value | |
14538 returned from a count of the remaining words. | |
14539 | |
14540 @need 1200 | |
14541 The expression will look like this, where @code{1+} is a function that | |
14542 adds one to its argument. | |
14543 | |
14544 @smallexample | |
14545 (1+ (recursive-count-words region-end)) | |
14546 @end smallexample | |
14547 | |
14548 @need 1200 | |
14549 The whole @code{recursive-count-words} function will then look like | |
14550 this: | |
14551 | |
14552 @smallexample | |
14553 @group | |
14554 (defun recursive-count-words (region-end) | |
14555 "@var{documentation}@dots{}" | |
14556 | |
14557 ;;; @r{1. do-again-test} | |
14558 (if (and (< (point) region-end) | |
14559 (re-search-forward "\\w+\\W*" region-end t)) | |
14560 @end group | |
14561 | |
14562 @group | |
14563 ;;; @r{2. then-part: the recursive call} | |
14564 (1+ (recursive-count-words region-end)) | |
14565 | |
14566 ;;; @r{3. else-part} | |
14567 0)) | |
14568 @end group | |
14569 @end smallexample | |
14570 | |
14571 @need 1250 | |
14572 Let's examine how this works: | |
14573 | |
14574 If there are no words in the region, the else part of the @code{if} | |
14575 expression is evaluated and consequently the function returns zero. | |
14576 | |
14577 If there is one word in the region, the value of point is less than | |
14578 the value of @code{region-end} and the search succeeds. In this case, | |
14579 the true-or-false-test of the @code{if} expression tests true, and the | |
14580 then-part of the @code{if} expression is evaluated. The counting | |
14581 expression is evaluated. This expression returns a value (which will | |
14582 be the value returned by the whole function) that is the sum of one | |
14583 added to the value returned by a recursive call. | |
14584 | |
14585 Meanwhile, the next-step-expression has caused point to jump over the | |
14586 first (and in this case only) word in the region. This means that | |
14587 when @code{(recursive-count-words region-end)} is evaluated a second | |
14588 time, as a result of the recursive call, the value of point will be | |
14589 equal to or greater than the value of region end. So this time, | |
14590 @code{recursive-count-words} will return zero. The zero will be added | |
14591 to one, and the original evaluation of @code{recursive-count-words} | |
14592 will return one plus zero, which is one, which is the correct amount. | |
14593 | |
14594 Clearly, if there are two words in the region, the first call to | |
14595 @code{recursive-count-words} returns one added to the value returned | |
14596 by calling @code{recursive-count-words} on a region containing the | |
14597 remaining word---that is, it adds one to one, producing two, which is | |
14598 the correct amount. | |
14599 | |
14600 Similarly, if there are three words in the region, the first call to | |
14601 @code{recursive-count-words} returns one added to the value returned | |
14602 by calling @code{recursive-count-words} on a region containing the | |
14603 remaining two words---and so on and so on. | |
14604 | |
14605 @need 1250 | |
14606 @noindent | |
14607 With full documentation the two functions look like this: | |
14608 | |
14609 @need 1250 | |
14610 @noindent | |
14611 The recursive function: | |
14612 | |
14613 @findex recursive-count-words | |
14614 @smallexample | |
14615 @group | |
14616 (defun recursive-count-words (region-end) | |
14617 "Number of words between point and REGION-END." | |
14618 @end group | |
14619 | |
14620 @group | |
14621 ;;; @r{1. do-again-test} | |
14622 (if (and (< (point) region-end) | |
14623 (re-search-forward "\\w+\\W*" region-end t)) | |
14624 @end group | |
14625 | |
14626 @group | |
14627 ;;; @r{2. then-part: the recursive call} | |
14628 (1+ (recursive-count-words region-end)) | |
14629 | |
14630 ;;; @r{3. else-part} | |
14631 0)) | |
14632 @end group | |
14633 @end smallexample | |
14634 | |
14635 @need 800 | |
14636 @noindent | |
14637 The wrapper: | |
14638 | |
14639 @smallexample | |
14640 @group | |
14641 ;;; @r{Recursive version} | |
14642 (defun count-words-region (beginning end) | |
14643 "Print number of words in the region. | |
14644 @end group | |
14645 | |
14646 @group | |
14647 Words are defined as at least one word-constituent | |
14648 character followed by at least one character that is | |
14649 not a word-constituent. The buffer's syntax table | |
14650 determines which characters these are." | |
14651 @end group | |
14652 @group | |
14653 (interactive "r") | |
14654 (message "Counting words in region ... ") | |
14655 (save-excursion | |
14656 (goto-char beginning) | |
14657 (let ((count (recursive-count-words end))) | |
14658 @end group | |
14659 @group | |
14660 (cond ((zerop count) | |
14661 (message | |
14662 "The region does NOT have any words.")) | |
14663 @end group | |
14664 @group | |
14665 ((= 1 count) | |
14666 (message "The region has 1 word.")) | |
14667 (t | |
14668 (message | |
14669 "The region has %d words." count)))))) | |
14670 @end group | |
14671 @end smallexample | |
14672 | |
14673 @node Counting Exercise, , recursive-count-words, Counting Words | |
14674 @section Exercise: Counting Punctuation | |
14675 | |
14676 Using a @code{while} loop, write a function to count the number of | |
14677 punctuation marks in a region---period, comma, semicolon, colon, | |
14678 exclamation mark, and question mark. Do the same using recursion. | |
14679 | |
14680 @node Words in a defun, Readying a Graph, Counting Words, Top | |
14681 @chapter Counting Words in a @code{defun} | |
14682 @cindex Counting words in a @code{defun} | |
14683 @cindex Word counting in a @code{defun} | |
14684 | |
14685 Our next project is to count the number of words in a function | |
14686 definition. Clearly, this can be done using some variant of | |
14687 @code{count-word-region}. @xref{Counting Words, , Counting Words: | |
14688 Repetition and Regexps}. If we are just going to count the words in | |
14689 one definition, it is easy enough to mark the definition with the | |
14690 @kbd{C-M-h} (@code{mark-defun}) command, and then call | |
14691 @code{count-word-region}. | |
14692 | |
14693 However, I am more ambitious: I want to count the words and symbols in | |
14694 every definition in the Emacs sources and then print a graph that | |
14695 shows how many functions there are of each length: how many contain 40 | |
14696 to 49 words or symbols, how many contain 50 to 59 words or symbols, | |
14697 and so on. I have often been curious how long a typical function is, | |
14698 and this will tell. | |
14699 | |
14700 @menu | |
14701 * Divide and Conquer:: | |
14702 * Words and Symbols:: What to count? | |
14703 * Syntax:: What constitutes a word or symbol? | |
14704 * count-words-in-defun:: Very like @code{count-words}. | |
14705 * Several defuns:: Counting several defuns in a file. | |
14706 * Find a File:: Do you want to look at a file? | |
14707 * lengths-list-file:: A list of the lengths of many definitions. | |
14708 * Several files:: Counting in definitions in different files. | |
14709 * Several files recursively:: Recursively counting in different files. | |
14710 * Prepare the data:: Prepare the data for display in a graph. | |
14711 @end menu | |
14712 | |
14713 @node Divide and Conquer, Words and Symbols, Words in a defun, Words in a defun | |
14714 @ifnottex | |
14715 @unnumberedsec Divide and Conquer | |
14716 @end ifnottex | |
14717 | |
14718 Described in one phrase, the histogram project is daunting; but | |
14719 divided into numerous small steps, each of which we can take one at a | |
14720 time, the project becomes less fearsome. Let us consider what the | |
14721 steps must be: | |
14722 | |
14723 @itemize @bullet | |
14724 @item | |
14725 First, write a function to count the words in one definition. This | |
14726 includes the problem of handling symbols as well as words. | |
14727 | |
14728 @item | |
14729 Second, write a function to list the numbers of words in each function | |
14730 in a file. This function can use the @code{count-words-in-defun} | |
14731 function. | |
14732 | |
14733 @item | |
14734 Third, write a function to list the numbers of words in each function | |
14735 in each of several files. This entails automatically finding the | |
14736 various files, switching to them, and counting the words in the | |
14737 definitions within them. | |
14738 | |
14739 @item | |
14740 Fourth, write a function to convert the list of numbers that we | |
14741 created in step three to a form that will be suitable for printing as | |
14742 a graph. | |
14743 | |
14744 @item | |
14745 Fifth, write a function to print the results as a graph. | |
14746 @end itemize | |
14747 | |
14748 This is quite a project! But if we take each step slowly, it will not | |
14749 be difficult. | |
14750 | |
14751 @node Words and Symbols, Syntax, Divide and Conquer, Words in a defun | |
14752 @section What to Count? | |
14753 @cindex Words and symbols in defun | |
14754 | |
14755 When we first start thinking about how to count the words in a | |
14756 function definition, the first question is (or ought to be) what are | |
14757 we going to count? When we speak of `words' with respect to a Lisp | |
14758 function definition, we are actually speaking, in large part, of | |
14759 `symbols'. For example, the following @code{multiply-by-seven} | |
14760 function contains the five symbols @code{defun}, | |
14761 @code{multiply-by-seven}, @code{number}, @code{*}, and @code{7}. In | |
14762 addition, in the documentation string, it contains the four words | |
14763 @samp{Multiply}, @samp{NUMBER}, @samp{by}, and @samp{seven}. The | |
14764 symbol @samp{number} is repeated, so the definition contains a total | |
14765 of ten words and symbols. | |
14766 | |
14767 @smallexample | |
14768 @group | |
14769 (defun multiply-by-seven (number) | |
14770 "Multiply NUMBER by seven." | |
14771 (* 7 number)) | |
14772 @end group | |
14773 @end smallexample | |
14774 | |
14775 @noindent | |
14776 However, if we mark the @code{multiply-by-seven} definition with | |
14777 @kbd{C-M-h} (@code{mark-defun}), and then call | |
14778 @code{count-words-region} on it, we will find that | |
14779 @code{count-words-region} claims the definition has eleven words, not | |
14780 ten! Something is wrong! | |
14781 | |
14782 The problem is twofold: @code{count-words-region} does not count the | |
14783 @samp{*} as a word, and it counts the single symbol, | |
14784 @code{multiply-by-seven}, as containing three words. The hyphens are | |
14785 treated as if they were interword spaces rather than intraword | |
14786 connectors: @samp{multiply-by-seven} is counted as if it were written | |
14787 @samp{multiply by seven}. | |
14788 | |
14789 The cause of this confusion is the regular expression search within | |
14790 the @code{count-words-region} definition that moves point forward word | |
14791 by word. In the canonical version of @code{count-words-region}, the | |
14792 regexp is: | |
14793 | |
14794 @smallexample | |
14795 "\\w+\\W*" | |
14796 @end smallexample | |
14797 | |
14798 @noindent | |
14799 This regular expression is a pattern defining one or more word | |
14800 constituent characters possibly followed by one or more characters | |
14801 that are not word constituents. What is meant by `word constituent | |
14802 characters' brings us to the issue of syntax, which is worth a section | |
14803 of its own. | |
14804 | |
14805 @node Syntax, count-words-in-defun, Words and Symbols, Words in a defun | |
14806 @section What Constitutes a Word or Symbol? | |
14807 @cindex Syntax categories and tables | |
14808 | |
14809 Emacs treats different characters as belonging to different | |
14810 @dfn{syntax categories}. For example, the regular expression, | |
14811 @samp{\\w+}, is a pattern specifying one or more @emph{word | |
14812 constituent} characters. Word constituent characters are members of | |
14813 one syntax category. Other syntax categories include the class of | |
14814 punctuation characters, such as the period and the comma, and the | |
14815 class of whitespace characters, such as the blank space and the tab | |
14816 character. (For more information, see @ref{Syntax, Syntax, The Syntax | |
14817 Table, emacs, The GNU Emacs Manual}, and @ref{Syntax Tables, , Syntax | |
14818 Tables, elisp, The GNU Emacs Lisp Reference Manual}.) | |
14819 | |
14820 Syntax tables specify which characters belong to which categories. | |
14821 Usually, a hyphen is not specified as a `word constituent character'. | |
14822 Instead, it is specified as being in the `class of characters that are | |
14823 part of symbol names but not words.' This means that the | |
14824 @code{count-words-region} function treats it in the same way it treats | |
14825 an interword white space, which is why @code{count-words-region} | |
14826 counts @samp{multiply-by-seven} as three words. | |
14827 | |
14828 There are two ways to cause Emacs to count @samp{multiply-by-seven} as | |
14829 one symbol: modify the syntax table or modify the regular expression. | |
14830 | |
14831 We could redefine a hyphen as a word constituent character by | |
14832 modifying the syntax table that Emacs keeps for each mode. This | |
14833 action would serve our purpose, except that a hyphen is merely the | |
14834 most common character within symbols that is not typically a word | |
14835 constituent character; there are others, too. | |
14836 | |
14837 Alternatively, we can redefine the regular expression used in the | |
14838 @code{count-words} definition so as to include symbols. This | |
14839 procedure has the merit of clarity, but the task is a little tricky. | |
14840 | |
14841 @need 1200 | |
14842 The first part is simple enough: the pattern must match ``at least one | |
14843 character that is a word or symbol constituent''. Thus: | |
14844 | |
14845 @smallexample | |
14846 "\\(\\w\\|\\s_\\)+" | |
14847 @end smallexample | |
14848 | |
14849 @noindent | |
14850 The @samp{\\(} is the first part of the grouping construct that | |
14851 includes the @samp{\\w} and the @samp{\\s_} as alternatives, separated | |
14852 by the @samp{\\|}. The @samp{\\w} matches any word-constituent | |
14853 character and the @samp{\\s_} matches any character that is part of a | |
14854 symbol name but not a word-constituent character. The @samp{+} | |
14855 following the group indicates that the word or symbol constituent | |
14856 characters must be matched at least once. | |
14857 | |
14858 However, the second part of the regexp is more difficult to design. | |
14859 What we want is to follow the first part with ``optionally one or more | |
14860 characters that are not constituents of a word or symbol''. At first, | |
14861 I thought I could define this with the following: | |
14862 | |
14863 @smallexample | |
14864 "\\(\\W\\|\\S_\\)*" | |
14865 @end smallexample | |
14866 | |
14867 @noindent | |
14868 The upper case @samp{W} and @samp{S} match characters that are | |
14869 @emph{not} word or symbol constituents. Unfortunately, this | |
14870 expression matches any character that is either not a word constituent | |
14871 or not a symbol constituent. This matches any character! | |
14872 | |
14873 I then noticed that every word or symbol in my test region was | |
14874 followed by white space (blank space, tab, or newline). So I tried | |
14875 placing a pattern to match one or more blank spaces after the pattern | |
14876 for one or more word or symbol constituents. This failed, too. Words | |
14877 and symbols are often separated by whitespace, but in actual code | |
14878 parentheses may follow symbols and punctuation may follow words. So | |
14879 finally, I designed a pattern in which the word or symbol constituents | |
14880 are followed optionally by characters that are not white space and | |
14881 then followed optionally by white space. | |
14882 | |
14883 @need 800 | |
14884 Here is the full regular expression: | |
14885 | |
14886 @smallexample | |
14887 "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" | |
14888 @end smallexample | |
14889 | |
14890 @node count-words-in-defun, Several defuns, Syntax, Words in a defun | |
14891 @section The @code{count-words-in-defun} Function | |
14892 @cindex Counting words in a @code{defun} | |
14893 | |
14894 We have seen that there are several ways to write a | |
14895 @code{count-word-region} function. To write a | |
14896 @code{count-words-in-defun}, we need merely adapt one of these | |
14897 versions. | |
14898 | |
14899 The version that uses a @code{while} loop is easy to understand, so I | |
14900 am going to adapt that. Because @code{count-words-in-defun} will be | |
14901 part of a more complex program, it need not be interactive and it need | |
14902 not display a message but just return the count. These considerations | |
14903 simplify the definition a little. | |
14904 | |
14905 On the other hand, @code{count-words-in-defun} will be used within a | |
14906 buffer that contains function definitions. Consequently, it is | |
14907 reasonable to ask that the function determine whether it is called | |
14908 when point is within a function definition, and if it is, to return | |
14909 the count for that definition. This adds complexity to the | |
14910 definition, but saves us from needing to pass arguments to the | |
14911 function. | |
14912 | |
14913 @need 1250 | |
14914 These considerations lead us to prepare the following template: | |
14915 | |
14916 @smallexample | |
14917 @group | |
14918 (defun count-words-in-defun () | |
14919 "@var{documentation}@dots{}" | |
14920 (@var{set up}@dots{} | |
14921 (@var{while loop}@dots{}) | |
14922 @var{return count}) | |
14923 @end group | |
14924 @end smallexample | |
14925 | |
14926 @noindent | |
14927 As usual, our job is to fill in the slots. | |
14928 | |
14929 First, the set up. | |
14930 | |
14931 We are presuming that this function will be called within a buffer | |
14932 containing function definitions. Point will either be within a | |
14933 function definition or not. For @code{count-words-in-defun} to work, | |
14934 point must move to the beginning of the definition, a counter must | |
14935 start at zero, and the counting loop must stop when point reaches the | |
14936 end of the definition. | |
14937 | |
14938 The @code{beginning-of-defun} function searches backwards for an | |
14939 opening delimiter such as a @samp{(} at the beginning of a line, and | |
14940 moves point to that position, or else to the limit of the search. In | |
14941 practice, this means that @code{beginning-of-defun} moves point to the | |
14942 beginning of an enclosing or preceding function definition, or else to | |
14943 the beginning of the buffer. We can use @code{beginning-of-defun} to | |
14944 place point where we wish to start. | |
14945 | |
14946 The @code{while} loop requires a counter to keep track of the words or | |
14947 symbols being counted. A @code{let} expression can be used to create | |
14948 a local variable for this purpose, and bind it to an initial value of zero. | |
14949 | |
14950 The @code{end-of-defun} function works like @code{beginning-of-defun} | |
14951 except that it moves point to the end of the definition. | |
14952 @code{end-of-defun} can be used as part of an expression that | |
14953 determines the position of the end of the definition. | |
14954 | |
14955 The set up for @code{count-words-in-defun} takes shape rapidly: first | |
14956 we move point to the beginning of the definition, then we create a | |
14957 local variable to hold the count, and finally, we record the position | |
14958 of the end of the definition so the @code{while} loop will know when to stop | |
14959 looping. | |
14960 | |
14961 @need 1250 | |
14962 The code looks like this: | |
14963 | |
14964 @smallexample | |
14965 @group | |
14966 (beginning-of-defun) | |
14967 (let ((count 0) | |
14968 (end (save-excursion (end-of-defun) (point)))) | |
14969 @end group | |
14970 @end smallexample | |
14971 | |
14972 @noindent | |
14973 The code is simple. The only slight complication is likely to concern | |
14974 @code{end}: it is bound to the position of the end of the definition | |
14975 by a @code{save-excursion} expression that returns the value of point | |
14976 after @code{end-of-defun} temporarily moves it to the end of the | |
14977 definition. | |
14978 | |
14979 The second part of the @code{count-words-in-defun}, after the set up, | |
14980 is the @code{while} loop. | |
14981 | |
14982 The loop must contain an expression that jumps point forward word by | |
14983 word and symbol by symbol, and another expression that counts the | |
14984 jumps. The true-or-false-test for the @code{while} loop should test | |
14985 true so long as point should jump forward, and false when point is at | |
14986 the end of the definition. We have already redefined the regular | |
14987 expression for this (@pxref{Syntax}), so the loop is straightforward: | |
14988 | |
14989 @smallexample | |
14990 @group | |
14991 (while (and (< (point) end) | |
14992 (re-search-forward | |
14993 "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" end t) | |
14994 (setq count (1+ count))) | |
14995 @end group | |
14996 @end smallexample | |
14997 | |
14998 The third part of the function definition returns the count of words | |
14999 and symbols. This part is the last expression within the body of the | |
15000 @code{let} expression, and can be, very simply, the local variable | |
15001 @code{count}, which when evaluated returns the count. | |
15002 | |
15003 @need 1250 | |
15004 Put together, the @code{count-words-in-defun} definition looks like this: | |
15005 | |
15006 @findex count-words-in-defun | |
15007 @smallexample | |
15008 @group | |
15009 (defun count-words-in-defun () | |
15010 "Return the number of words and symbols in a defun." | |
15011 (beginning-of-defun) | |
15012 (let ((count 0) | |
15013 (end (save-excursion (end-of-defun) (point)))) | |
15014 @end group | |
15015 @group | |
15016 (while | |
15017 (and (< (point) end) | |
15018 (re-search-forward | |
15019 "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" | |
15020 end t)) | |
15021 (setq count (1+ count))) | |
15022 count)) | |
15023 @end group | |
15024 @end smallexample | |
15025 | |
15026 How to test this? The function is not interactive, but it is easy to | |
15027 put a wrapper around the function to make it interactive; we can use | |
15028 almost the same code as for the recursive version of | |
15029 @code{count-words-region}: | |
15030 | |
15031 @smallexample | |
15032 @group | |
15033 ;;; @r{Interactive version.} | |
15034 (defun count-words-defun () | |
15035 "Number of words and symbols in a function definition." | |
15036 (interactive) | |
15037 (message | |
15038 "Counting words and symbols in function definition ... ") | |
15039 @end group | |
15040 @group | |
15041 (let ((count (count-words-in-defun))) | |
15042 (cond | |
15043 ((zerop count) | |
15044 (message | |
15045 "The definition does NOT have any words or symbols.")) | |
15046 @end group | |
15047 @group | |
15048 ((= 1 count) | |
15049 (message | |
15050 "The definition has 1 word or symbol.")) | |
15051 (t | |
15052 (message | |
15053 "The definition has %d words or symbols." count))))) | |
15054 @end group | |
15055 @end smallexample | |
15056 | |
15057 @need 800 | |
15058 @noindent | |
15059 Let's re-use @kbd{C-c =} as a convenient keybinding: | |
15060 | |
15061 @smallexample | |
15062 (global-set-key "\C-c=" 'count-words-defun) | |
15063 @end smallexample | |
15064 | |
15065 Now we can try out @code{count-words-defun}: install both | |
15066 @code{count-words-in-defun} and @code{count-words-defun}, and set the | |
15067 keybinding, and then place the cursor within the following definition: | |
15068 | |
15069 @smallexample | |
15070 @group | |
15071 (defun multiply-by-seven (number) | |
15072 "Multiply NUMBER by seven." | |
15073 (* 7 number)) | |
15074 @result{} 10 | |
15075 @end group | |
15076 @end smallexample | |
15077 | |
15078 @noindent | |
15079 Success! The definition has 10 words and symbols. | |
15080 | |
15081 The next problem is to count the numbers of words and symbols in | |
15082 several definitions within a single file. | |
15083 | |
15084 @node Several defuns, Find a File, count-words-in-defun, Words in a defun | |
15085 @section Count Several @code{defuns} Within a File | |
15086 | |
15087 A file such as @file{simple.el} may have a hundred or more function | |
15088 definitions within it. Our long term goal is to collect statistics on | |
15089 many files, but as a first step, our immediate goal is to collect | |
15090 statistics on one file. | |
15091 | |
15092 The information will be a series of numbers, each number being the | |
15093 length of a function definition. We can store the numbers in a list. | |
15094 | |
15095 We know that we will want to incorporate the information regarding one | |
15096 file with information about many other files; this means that the | |
15097 function for counting definition lengths within one file need only | |
15098 return the list of lengths. It need not and should not display any | |
15099 messages. | |
15100 | |
15101 The word count commands contain one expression to jump point forward | |
15102 word by word and another expression to count the jumps. The function | |
15103 to return the lengths of definitions can be designed to work the same | |
15104 way, with one expression to jump point forward definition by | |
15105 definition and another expression to construct the lengths' list. | |
15106 | |
15107 This statement of the problem makes it elementary to write the | |
15108 function definition. Clearly, we will start the count at the | |
15109 beginning of the file, so the first command will be @code{(goto-char | |
15110 (point-min))}. Next, we start the @code{while} loop; and the | |
15111 true-or-false test of the loop can be a regular expression search for | |
15112 the next function definition---so long as the search succeeds, point | |
15113 is moved forward and then the body of the loop is evaluated. The body | |
15114 needs an expression that constructs the lengths' list. @code{cons}, | |
15115 the list construction command, can be used to create the list. That | |
15116 is almost all there is to it. | |
15117 | |
15118 @need 800 | |
15119 Here is what this fragment of code looks like: | |
15120 | |
15121 @smallexample | |
15122 @group | |
15123 (goto-char (point-min)) | |
15124 (while (re-search-forward "^(defun" nil t) | |
15125 (setq lengths-list | |
15126 (cons (count-words-in-defun) lengths-list))) | |
15127 @end group | |
15128 @end smallexample | |
15129 | |
15130 What we have left out is the mechanism for finding the file that | |
15131 contains the function definitions. | |
15132 | |
15133 In previous examples, we either used this, the Info file, or we | |
15134 switched back and forth to some other buffer, such as the | |
15135 @file{*scratch*} buffer. | |
15136 | |
15137 Finding a file is a new process that we have not yet discussed. | |
15138 | |
15139 @node Find a File, lengths-list-file, Several defuns, Words in a defun | |
15140 @comment node-name, next, previous, up | |
15141 @section Find a File | |
15142 @cindex Find a File | |
15143 | |
15144 To find a file in Emacs, you use the @kbd{C-x C-f} (@code{find-file}) | |
15145 command. This command is almost, but not quite right for the lengths | |
15146 problem. | |
15147 | |
15148 @need 1200 | |
15149 Let's look at the source for @code{find-file}: | |
15150 | |
15151 @smallexample | |
15152 @group | |
15153 (defun find-file (filename) | |
15154 "Edit file FILENAME. | |
15155 Switch to a buffer visiting file FILENAME, | |
15156 creating one if none already exists." | |
15157 (interactive "FFind file: ") | |
15158 (switch-to-buffer (find-file-noselect filename))) | |
15159 @end group | |
15160 @end smallexample | |
15161 | |
15162 @noindent | |
15163 (The most recent version of the @code{find-file} function definition | |
15164 permits you to specify optional wildcards to visit multiple files; that | |
15165 makes the definition more complex and we will not discuss it here, | |
15166 since it is not relevant. You can see its source using either | |
15167 @kbd{M-.} (@code{find-tag}) or @kbd{C-h f} (@code{describe-function}).) | |
15168 | |
15169 @ignore | |
15170 In Emacs 22 | |
15171 (defun find-file (filename &optional wildcards) | |
15172 "Edit file FILENAME. | |
15173 Switch to a buffer visiting file FILENAME, | |
15174 creating one if none already exists. | |
15175 Interactively, the default if you just type RET is the current directory, | |
15176 but the visited file name is available through the minibuffer history: | |
15177 type M-n to pull it into the minibuffer. | |
15178 | |
15179 Interactively, or if WILDCARDS is non-nil in a call from Lisp, | |
15180 expand wildcards (if any) and visit multiple files. You can | |
15181 suppress wildcard expansion by setting `find-file-wildcards' to nil. | |
15182 | |
15183 To visit a file without any kind of conversion and without | |
15184 automatically choosing a major mode, use \\[find-file-literally]." | |
15185 (interactive (find-file-read-args "Find file: " nil)) | |
15186 (let ((value (find-file-noselect filename nil nil wildcards))) | |
15187 (if (listp value) | |
15188 (mapcar 'switch-to-buffer (nreverse value)) | |
15189 (switch-to-buffer value)))) | |
15190 @end ignore | |
15191 | |
15192 The definition I am showing possesses short but complete documentation | |
15193 and an interactive specification that prompts you for a file name when | |
15194 you use the command interactively. The body of the definition | |
15195 contains two functions, @code{find-file-noselect} and | |
15196 @code{switch-to-buffer}. | |
15197 | |
15198 According to its documentation as shown by @kbd{C-h f} (the | |
15199 @code{describe-function} command), the @code{find-file-noselect} | |
15200 function reads the named file into a buffer and returns the buffer. | |
15201 (Its most recent version includes an optional wildcards argument, | |
15202 too, as well as another to read a file literally and an other you | |
15203 suppress warning messages. These optional arguments are irrelevant.) | |
15204 | |
15205 However, the @code{find-file-noselect} function does not select the | |
15206 buffer in which it puts the file. Emacs does not switch its attention | |
15207 (or yours if you are using @code{find-file-noselect}) to the selected | |
15208 buffer. That is what @code{switch-to-buffer} does: it switches the | |
15209 buffer to which Emacs attention is directed; and it switches the | |
15210 buffer displayed in the window to the new buffer. We have discussed | |
15211 buffer switching elsewhere. (@xref{Switching Buffers}.) | |
15212 | |
15213 In this histogram project, we do not need to display each file on the | |
15214 screen as the program determines the length of each definition within | |
15215 it. Instead of employing @code{switch-to-buffer}, we can work with | |
15216 @code{set-buffer}, which redirects the attention of the computer | |
15217 program to a different buffer but does not redisplay it on the screen. | |
15218 So instead of calling on @code{find-file} to do the job, we must write | |
15219 our own expression. | |
15220 | |
15221 The task is easy: use @code{find-file-noselect} and @code{set-buffer}. | |
15222 | |
15223 @node lengths-list-file, Several files, Find a File, Words in a defun | |
15224 @section @code{lengths-list-file} in Detail | |
15225 | |
15226 The core of the @code{lengths-list-file} function is a @code{while} | |
15227 loop containing a function to move point forward `defun by defun' and | |
15228 a function to count the number of words and symbols in each defun. | |
15229 This core must be surrounded by functions that do various other tasks, | |
15230 including finding the file, and ensuring that point starts out at the | |
15231 beginning of the file. The function definition looks like this: | |
15232 @findex lengths-list-file | |
15233 | |
15234 @smallexample | |
15235 @group | |
15236 (defun lengths-list-file (filename) | |
15237 "Return list of definitions' lengths within FILE. | |
15238 The returned list is a list of numbers. | |
15239 Each number is the number of words or | |
15240 symbols in one function definition." | |
15241 @end group | |
15242 @group | |
15243 (message "Working on `%s' ... " filename) | |
15244 (save-excursion | |
15245 (let ((buffer (find-file-noselect filename)) | |
15246 (lengths-list)) | |
15247 (set-buffer buffer) | |
15248 (setq buffer-read-only t) | |
15249 (widen) | |
15250 (goto-char (point-min)) | |
15251 (while (re-search-forward "^(defun" nil t) | |
15252 (setq lengths-list | |
15253 (cons (count-words-in-defun) lengths-list))) | |
15254 (kill-buffer buffer) | |
15255 lengths-list))) | |
15256 @end group | |
15257 @end smallexample | |
15258 | |
15259 @noindent | |
15260 The function is passed one argument, the name of the file on which it | |
15261 will work. It has four lines of documentation, but no interactive | |
15262 specification. Since people worry that a computer is broken if they | |
15263 don't see anything going on, the first line of the body is a | |
15264 message. | |
15265 | |
15266 The next line contains a @code{save-excursion} that returns Emacs' | |
15267 attention to the current buffer when the function completes. This is | |
15268 useful in case you embed this function in another function that | |
15269 presumes point is restored to the original buffer. | |
15270 | |
15271 In the varlist of the @code{let} expression, Emacs finds the file and | |
15272 binds the local variable @code{buffer} to the buffer containing the | |
15273 file. At the same time, Emacs creates @code{lengths-list} as a local | |
15274 variable. | |
15275 | |
15276 Next, Emacs switches its attention to the buffer. | |
15277 | |
15278 In the following line, Emacs makes the buffer read-only. Ideally, | |
15279 this line is not necessary. None of the functions for counting words | |
15280 and symbols in a function definition should change the buffer. | |
15281 Besides, the buffer is not going to be saved, even if it were changed. | |
15282 This line is entirely the consequence of great, perhaps excessive, | |
15283 caution. The reason for the caution is that this function and those | |
15284 it calls work on the sources for Emacs and it is inconvenient if they | |
15285 are inadvertently modified. It goes without saying that I did not | |
15286 realize a need for this line until an experiment went awry and started | |
15287 to modify my Emacs source files @dots{} | |
15288 | |
15289 Next comes a call to widen the buffer if it is narrowed. This | |
15290 function is usually not needed---Emacs creates a fresh buffer if none | |
15291 already exists; but if a buffer visiting the file already exists Emacs | |
15292 returns that one. In this case, the buffer may be narrowed and must | |
15293 be widened. If we wanted to be fully `user-friendly', we would | |
15294 arrange to save the restriction and the location of point, but we | |
15295 won't. | |
15296 | |
15297 The @code{(goto-char (point-min))} expression moves point to the | |
15298 beginning of the buffer. | |
15299 | |
15300 Then comes a @code{while} loop in which the `work' of the function is | |
15301 carried out. In the loop, Emacs determines the length of each | |
15302 definition and constructs a lengths' list containing the information. | |
15303 | |
15304 Emacs kills the buffer after working through it. This is to save | |
15305 space inside of Emacs. My version of GNU Emacs 19 contained over 300 | |
15306 source files of interest; GNU Emacs 22 contains over a thousand source | |
15307 files. Another function will apply @code{lengths-list-file} to each | |
15308 of the files. | |
15309 | |
15310 Finally, the last expression within the @code{let} expression is the | |
15311 @code{lengths-list} variable; its value is returned as the value of | |
15312 the whole function. | |
15313 | |
15314 You can try this function by installing it in the usual fashion. Then | |
15315 place your cursor after the following expression and type @kbd{C-x | |
15316 C-e} (@code{eval-last-sexp}). | |
15317 | |
15318 @c !!! 22.1.1 lisp sources location here | |
15319 @smallexample | |
15320 (lengths-list-file | |
15321 "/usr/local/share/emacs/22.1.1/lisp/emacs-lisp/debug.el") | |
15322 @end smallexample | |
15323 | |
15324 @noindent | |
15325 (You may need to change the pathname of the file; the one here is for | |
15326 GNU Emacs version 22.1.1. To change the expression, copy it to | |
15327 the @file{*scratch*} buffer and edit it. | |
15328 | |
15329 @need 1200 | |
15330 @noindent | |
15331 (Also, to see the full length of the list, rather than a truncated | |
15332 version, you may have to evaluate the following: | |
15333 | |
15334 @smallexample | |
15335 (custom-set-variables '(eval-expression-print-length nil)) | |
15336 @end smallexample | |
15337 | |
15338 @noindent | |
15339 (@xref{defcustom, , Specifying Variables using @code{defcustom}}. | |
15340 Then evaluate the @code{lengths-list-file} expression.) | |
15341 | |
15342 @need 1200 | |
15343 The lengths' list for @file{debug.el} takes less than a second to | |
15344 produce and looks like this in GNU Emacs 22: | |
15345 | |
15346 @smallexample | |
15347 (83 113 105 144 289 22 30 97 48 89 25 52 52 88 28 29 77 49 43 290 232 587) | |
15348 @end smallexample | |
15349 | |
15350 @need 1500 | |
15351 (Using my old machine, the version 19 lengths' list for @file{debug.el} | |
15352 took seven seconds to produce and looked like this: | |
15353 | |
15354 @smallexample | |
15355 (75 41 80 62 20 45 44 68 45 12 34 235) | |
15356 @end smallexample | |
15357 | |
15358 (The newer version of @file{debug.el} contains more defuns than the | |
15359 earlier one; and my new machine is much faster than the old one.) | |
15360 | |
15361 Note that the length of the last definition in the file is first in | |
15362 the list. | |
15363 | |
15364 @node Several files, Several files recursively, lengths-list-file, Words in a defun | |
15365 @section Count Words in @code{defuns} in Different Files | |
15366 | |
15367 In the previous section, we created a function that returns a list of | |
15368 the lengths of each definition in a file. Now, we want to define a | |
15369 function to return a master list of the lengths of the definitions in | |
15370 a list of files. | |
15371 | |
15372 Working on each of a list of files is a repetitious act, so we can use | |
15373 either a @code{while} loop or recursion. | |
15374 | |
15375 @menu | |
15376 * lengths-list-many-files:: Return a list of the lengths of defuns. | |
15377 * append:: Attach one list to another. | |
15378 @end menu | |
15379 | |
15380 @node lengths-list-many-files, append, Several files, Several files | |
15381 @ifnottex | |
15382 @unnumberedsubsec Determine the lengths of @code{defuns} | |
15383 @end ifnottex | |
15384 | |
15385 The design using a @code{while} loop is routine. The argument passed | |
15386 the function is a list of files. As we saw earlier (@pxref{Loop | |
15387 Example}), you can write a @code{while} loop so that the body of the | |
15388 loop is evaluated if such a list contains elements, but to exit the | |
15389 loop if the list is empty. For this design to work, the body of the | |
15390 loop must contain an expression that shortens the list each time the | |
15391 body is evaluated, so that eventually the list is empty. The usual | |
15392 technique is to set the value of the list to the value of the @sc{cdr} | |
15393 of the list each time the body is evaluated. | |
15394 | |
15395 @need 800 | |
15396 The template looks like this: | |
15397 | |
15398 @smallexample | |
15399 @group | |
15400 (while @var{test-whether-list-is-empty} | |
15401 @var{body}@dots{} | |
15402 @var{set-list-to-cdr-of-list}) | |
15403 @end group | |
15404 @end smallexample | |
15405 | |
15406 Also, we remember that a @code{while} loop returns @code{nil} (the | |
15407 result of evaluating the true-or-false-test), not the result of any | |
15408 evaluation within its body. (The evaluations within the body of the | |
15409 loop are done for their side effects.) However, the expression that | |
15410 sets the lengths' list is part of the body---and that is the value | |
15411 that we want returned by the function as a whole. To do this, we | |
15412 enclose the @code{while} loop within a @code{let} expression, and | |
15413 arrange that the last element of the @code{let} expression contains | |
15414 the value of the lengths' list. (@xref{Incrementing Example, , Loop | |
15415 Example with an Incrementing Counter}.) | |
15416 | |
15417 @findex lengths-list-many-files | |
15418 @need 1250 | |
15419 These considerations lead us directly to the function itself: | |
15420 | |
15421 @smallexample | |
15422 @group | |
15423 ;;; @r{Use @code{while} loop.} | |
15424 (defun lengths-list-many-files (list-of-files) | |
15425 "Return list of lengths of defuns in LIST-OF-FILES." | |
15426 @end group | |
15427 @group | |
15428 (let (lengths-list) | |
15429 | |
15430 ;;; @r{true-or-false-test} | |
15431 (while list-of-files | |
15432 (setq lengths-list | |
15433 (append | |
15434 lengths-list | |
15435 | |
15436 ;;; @r{Generate a lengths' list.} | |
15437 (lengths-list-file | |
15438 (expand-file-name (car list-of-files))))) | |
15439 @end group | |
15440 | |
15441 @group | |
15442 ;;; @r{Make files' list shorter.} | |
15443 (setq list-of-files (cdr list-of-files))) | |
15444 | |
15445 ;;; @r{Return final value of lengths' list.} | |
15446 lengths-list)) | |
15447 @end group | |
15448 @end smallexample | |
15449 | |
15450 @code{expand-file-name} is a built-in function that converts a file | |
15451 name to the absolute, long, path name form. The function employs the | |
15452 name of the directory in which the function is called. | |
15453 | |
15454 @c !!! 22.1.1 lisp sources location here | |
15455 @need 1500 | |
15456 Thus, if @code{expand-file-name} is called on @code{debug.el} when | |
15457 Emacs is visiting the | |
15458 @file{/usr/local/share/emacs/22.1.1/lisp/emacs-lisp/} directory, | |
15459 | |
15460 @smallexample | |
15461 debug.el | |
15462 @end smallexample | |
15463 | |
15464 @need 800 | |
15465 @noindent | |
15466 becomes | |
15467 | |
15468 @c !!! 22.1.1 lisp sources location here | |
15469 @smallexample | |
15470 /usr/local/share/emacs/22.1.1/lisp/emacs-lisp/debug.el | |
15471 @end smallexample | |
15472 | |
15473 The only other new element of this function definition is the as yet | |
15474 unstudied function @code{append}, which merits a short section for | |
15475 itself. | |
15476 | |
15477 @node append, , lengths-list-many-files, Several files | |
15478 @subsection The @code{append} Function | |
15479 | |
15480 @need 800 | |
15481 The @code{append} function attaches one list to another. Thus, | |
15482 | |
15483 @smallexample | |
15484 (append '(1 2 3 4) '(5 6 7 8)) | |
15485 @end smallexample | |
15486 | |
15487 @need 800 | |
15488 @noindent | |
15489 produces the list | |
15490 | |
15491 @smallexample | |
15492 (1 2 3 4 5 6 7 8) | |
15493 @end smallexample | |
15494 | |
15495 This is exactly how we want to attach two lengths' lists produced by | |
15496 @code{lengths-list-file} to each other. The results contrast with | |
15497 @code{cons}, | |
15498 | |
15499 @smallexample | |
15500 (cons '(1 2 3 4) '(5 6 7 8)) | |
15501 @end smallexample | |
15502 | |
15503 @need 1250 | |
15504 @noindent | |
15505 which constructs a new list in which the first argument to @code{cons} | |
15506 becomes the first element of the new list: | |
15507 | |
15508 @smallexample | |
15509 ((1 2 3 4) 5 6 7 8) | |
15510 @end smallexample | |
15511 | |
15512 @node Several files recursively, Prepare the data, Several files, Words in a defun | |
15513 @section Recursively Count Words in Different Files | |
15514 | |
15515 Besides a @code{while} loop, you can work on each of a list of files | |
15516 with recursion. A recursive version of @code{lengths-list-many-files} | |
15517 is short and simple. | |
15518 | |
15519 The recursive function has the usual parts: the `do-again-test', the | |
15520 `next-step-expression', and the recursive call. The `do-again-test' | |
15521 determines whether the function should call itself again, which it | |
15522 will do if the @code{list-of-files} contains any remaining elements; | |
15523 the `next-step-expression' resets the @code{list-of-files} to the | |
15524 @sc{cdr} of itself, so eventually the list will be empty; and the | |
15525 recursive call calls itself on the shorter list. The complete | |
15526 function is shorter than this description! | |
15527 @findex recursive-lengths-list-many-files | |
15528 | |
15529 @smallexample | |
15530 @group | |
15531 (defun recursive-lengths-list-many-files (list-of-files) | |
15532 "Return list of lengths of each defun in LIST-OF-FILES." | |
15533 (if list-of-files ; @r{do-again-test} | |
15534 (append | |
15535 (lengths-list-file | |
15536 (expand-file-name (car list-of-files))) | |
15537 (recursive-lengths-list-many-files | |
15538 (cdr list-of-files))))) | |
15539 @end group | |
15540 @end smallexample | |
15541 | |
15542 @noindent | |
15543 In a sentence, the function returns the lengths' list for the first of | |
15544 the @code{list-of-files} appended to the result of calling itself on | |
15545 the rest of the @code{list-of-files}. | |
15546 | |
15547 Here is a test of @code{recursive-lengths-list-many-files}, along with | |
15548 the results of running @code{lengths-list-file} on each of the files | |
15549 individually. | |
15550 | |
15551 Install @code{recursive-lengths-list-many-files} and | |
15552 @code{lengths-list-file}, if necessary, and then evaluate the | |
15553 following expressions. You may need to change the files' pathnames; | |
15554 those here work when this Info file and the Emacs sources are located | |
15555 in their customary places. To change the expressions, copy them to | |
15556 the @file{*scratch*} buffer, edit them, and then evaluate them. | |
15557 | |
15558 The results are shown after the @samp{@result{}}. (These results are | |
15559 for files from Emacs version 22.1.1; files from other versions of | |
15560 Emacs may produce different results.) | |
15561 | |
15562 @c !!! 22.1.1 lisp sources location here | |
15563 @smallexample | |
15564 @group | |
15565 (cd "/usr/local/share/emacs/22.1.1/") | |
15566 | |
15567 (lengths-list-file "./lisp/macros.el") | |
15568 @result{} (283 263 480 90) | |
15569 @end group | |
15570 | |
15571 @group | |
15572 (lengths-list-file "./lisp/mail/mailalias.el") | |
15573 @result{} (38 32 29 95 178 180 321 218 324) | |
15574 @end group | |
15575 | |
15576 @group | |
15577 (lengths-list-file "./lisp/makesum.el") | |
15578 @result{} (85 181) | |
15579 @end group | |
15580 | |
15581 @group | |
15582 (recursive-lengths-list-many-files | |
15583 '("./lisp/macros.el" | |
15584 "./lisp/mail/mailalias.el" | |
15585 "./lisp/makesum.el")) | |
15586 @result{} (283 263 480 90 38 32 29 95 178 180 321 218 324 85 181) | |
15587 @end group | |
15588 @end smallexample | |
15589 | |
15590 The @code{recursive-lengths-list-many-files} function produces the | |
15591 output we want. | |
15592 | |
15593 The next step is to prepare the data in the list for display in a graph. | |
15594 | |
15595 @node Prepare the data, , Several files recursively, Words in a defun | |
15596 @section Prepare the Data for Display in a Graph | |
15597 | |
15598 The @code{recursive-lengths-list-many-files} function returns a list | |
15599 of numbers. Each number records the length of a function definition. | |
15600 What we need to do now is transform this data into a list of numbers | |
15601 suitable for generating a graph. The new list will tell how many | |
15602 functions definitions contain less than 10 words and | |
15603 symbols, how many contain between 10 and 19 words and symbols, how | |
15604 many contain between 20 and 29 words and symbols, and so on. | |
15605 | |
15606 In brief, we need to go through the lengths' list produced by the | |
15607 @code{recursive-lengths-list-many-files} function and count the number | |
15608 of defuns within each range of lengths, and produce a list of those | |
15609 numbers. | |
15610 | |
15611 @menu | |
15612 * Data for Display in Detail:: | |
15613 * Sorting:: Sorting lists. | |
15614 * Files List:: Making a list of files. | |
15615 * Counting function definitions:: | |
15616 @end menu | |
15617 | |
15618 @node Data for Display in Detail, Sorting, Prepare the data, Prepare the data | |
15619 @ifnottex | |
15620 @unnumberedsubsec The Data for Display in Detail | |
15621 @end ifnottex | |
15622 | |
15623 Based on what we have done before, we can readily foresee that it | |
15624 should not be too hard to write a function that `@sc{cdr}s' down the | |
15625 lengths' list, looks at each element, determines which length range it | |
15626 is in, and increments a counter for that range. | |
15627 | |
15628 However, before beginning to write such a function, we should consider | |
15629 the advantages of sorting the lengths' list first, so the numbers are | |
15630 ordered from smallest to largest. First, sorting will make it easier | |
15631 to count the numbers in each range, since two adjacent numbers will | |
15632 either be in the same length range or in adjacent ranges. Second, by | |
15633 inspecting a sorted list, we can discover the highest and lowest | |
15634 number, and thereby determine the largest and smallest length range | |
15635 that we will need. | |
15636 | |
15637 @node Sorting, Files List, Data for Display in Detail, Prepare the data | |
15638 @subsection Sorting Lists | |
15639 @findex sort | |
15640 | |
15641 Emacs contains a function to sort lists, called (as you might guess) | |
15642 @code{sort}. The @code{sort} function takes two arguments, the list | |
15643 to be sorted, and a predicate that determines whether the first of | |
15644 two list elements is ``less'' than the second. | |
15645 | |
15646 As we saw earlier (@pxref{Wrong Type of Argument, , Using the Wrong | |
15647 Type Object as an Argument}), a predicate is a function that | |
15648 determines whether some property is true or false. The @code{sort} | |
15649 function will reorder a list according to whatever property the | |
15650 predicate uses; this means that @code{sort} can be used to sort | |
15651 non-numeric lists by non-numeric criteria---it can, for example, | |
15652 alphabetize a list. | |
15653 | |
15654 @need 1250 | |
15655 The @code{<} function is used when sorting a numeric list. For example, | |
15656 | |
15657 @smallexample | |
15658 (sort '(4 8 21 17 33 7 21 7) '<) | |
15659 @end smallexample | |
15660 | |
15661 @need 800 | |
15662 @noindent | |
15663 produces this: | |
15664 | |
15665 @smallexample | |
15666 (4 7 7 8 17 21 21 33) | |
15667 @end smallexample | |
15668 | |
15669 @noindent | |
15670 (Note that in this example, both the arguments are quoted so that the | |
15671 symbols are not evaluated before being passed to @code{sort} as | |
15672 arguments.) | |
15673 | |
15674 Sorting the list returned by the | |
15675 @code{recursive-lengths-list-many-files} function is straightforward; | |
15676 it uses the @code{<} function: | |
15677 | |
15678 @ignore | |
15679 2006 Oct 29 | |
15680 In GNU Emacs 22, eval | |
15681 (progn | |
15682 (cd "/usr/local/share/emacs/22.0.50/") | |
15683 (sort | |
15684 (recursive-lengths-list-many-files | |
15685 '("./lisp/macros.el" | |
15686 "./lisp/mail/mailalias.el" | |
15687 "./lisp/makesum.el")) | |
15688 '<)) | |
15689 | |
15690 @end ignore | |
15691 | |
15692 @smallexample | |
15693 @group | |
15694 (sort | |
15695 (recursive-lengths-list-many-files | |
15696 '("./lisp/macros.el" | |
15697 "./lisp/mailalias.el" | |
15698 "./lisp/makesum.el")) | |
15699 '<) | |
15700 @end group | |
15701 @end smallexample | |
15702 | |
15703 @need 800 | |
15704 @noindent | |
15705 which produces: | |
15706 | |
15707 @smallexample | |
15708 (29 32 38 85 90 95 178 180 181 218 263 283 321 324 480) | |
15709 @end smallexample | |
15710 | |
15711 @noindent | |
15712 (Note that in this example, the first argument to @code{sort} is not | |
15713 quoted, since the expression must be evaluated so as to produce the | |
15714 list that is passed to @code{sort}.) | |
15715 | |
15716 @node Files List, Counting function definitions, Sorting, Prepare the data | |
15717 @subsection Making a List of Files | |
15718 | |
15719 The @code{recursive-lengths-list-many-files} function requires a list | |
15720 of files as its argument. For our test examples, we constructed such | |
15721 a list by hand; but the Emacs Lisp source directory is too large for | |
15722 us to do for that. Instead, we will write a function to do the job | |
15723 for us. In this function, we will use both a @code{while} loop and a | |
15724 recursive call. | |
15725 | |
15726 @findex directory-files | |
15727 We did not have to write a function like this for older versions of | |
15728 GNU Emacs, since they placed all the @samp{.el} files in one | |
15729 directory. Instead, we were able to use the @code{directory-files} | |
15730 function, which lists the names of files that match a specified | |
15731 pattern within a single directory. | |
15732 | |
15733 However, recent versions of Emacs place Emacs Lisp files in | |
15734 sub-directories of the top level @file{lisp} directory. This | |
15735 re-arrangement eases navigation. For example, all the mail related | |
15736 files are in a @file{lisp} sub-directory called @file{mail}. But at | |
15737 the same time, this arrangement forces us to create a file listing | |
15738 function that descends into the sub-directories. | |
15739 | |
15740 @findex files-in-below-directory | |
15741 We can create this function, called @code{files-in-below-directory}, | |
15742 using familiar functions such as @code{car}, @code{nthcdr}, and | |
15743 @code{substring} in conjunction with an existing function called | |
15744 @code{directory-files-and-attributes}. This latter function not only | |
15745 lists all the filenames in a directory, including the names | |
15746 of sub-directories, but also their attributes. | |
15747 | |
15748 To restate our goal: to create a function that will enable us | |
15749 to feed filenames to @code{recursive-lengths-list-many-files} | |
15750 as a list that looks like this (but with more elements): | |
15751 | |
15752 @smallexample | |
15753 @group | |
15754 ("./lisp/macros.el" | |
15755 "./lisp/mail/rmail.el" | |
15756 "./lisp/makesum.el") | |
15757 @end group | |
15758 @end smallexample | |
15759 | |
15760 The @code{directory-files-and-attributes} function returns a list of | |
15761 lists. Each of the lists within the main list consists of 13 | |
15762 elements. The first element is a string that contains the name of the | |
15763 file -- which, in GNU/Linux, may be a `directory file', that is to | |
15764 say, a file with the special attributes of a directory. The second | |
15765 element of the list is @code{t} for a directory, a string | |
15766 for symbolic link (the string is the name linked to), or @code{nil}. | |
15767 | |
15768 For example, the first @samp{.el} file in the @file{lisp/} directory | |
15769 is @file{abbrev.el}. Its name is | |
15770 @file{/usr/local/share/emacs/22.1.1/lisp/abbrev.el} and it is not a | |
15771 directory or a symbolic link. | |
15772 | |
15773 @need 1000 | |
15774 This is how @code{directory-files-and-attributes} lists that file and | |
15775 its attributes: | |
15776 | |
15777 @smallexample | |
15778 @group | |
15779 ("abbrev.el" | |
15780 nil | |
15781 1 | |
15782 1000 | |
15783 100 | |
15784 @end group | |
15785 @group | |
15786 (17733 259) | |
15787 (17491 28834) | |
15788 (17596 62124) | |
15789 13157 | |
15790 "-rw-rw-r--" | |
15791 @end group | |
15792 @group | |
15793 nil | |
15794 2971624 | |
15795 773) | |
15796 @end group | |
15797 @end smallexample | |
15798 | |
15799 @need 1200 | |
15800 On the other hand, @file{mail/} is a directory within the @file{lisp/} | |
15801 directory. The beginning of its listing looks like this: | |
15802 | |
15803 @smallexample | |
15804 @group | |
15805 ("mail" | |
15806 t | |
15807 @dots{} | |
15808 ) | |
15809 @end group | |
15810 @end smallexample | |
15811 | |
15812 (To learn about the different attributes, look at the documentation of | |
15813 @code{file-attributes}. Bear in mind that the @code{file-attributes} | |
15814 function does not list the filename, so its first element is | |
15815 @code{directory-files-and-attributes}'s second element.) | |
15816 | |
15817 We will want our new function, @code{files-in-below-directory}, to | |
15818 list the @samp{.el} files in the directory it is told to check, and in | |
15819 any directories below that directory. | |
15820 | |
15821 This gives us a hint on how to construct | |
15822 @code{files-in-below-directory}: within a directory, the function | |
15823 should add @samp{.el} filenames to a list; and if, within a directory, | |
15824 the function comes upon a sub-directory, it should go into that | |
15825 sub-directory and repeat its actions. | |
15826 | |
15827 However, we should note that every directory contains a name that | |
15828 refers to itself, called @file{.}, (``dot'') and a name that refers to | |
15829 its parent directory, called @file{..} (``double dot''). (In | |
15830 @file{/}, the root directory, @file{..} refers to itself, since | |
15831 @file{/} has no parent.) Clearly, we do not want our | |
15832 @code{files-in-below-directory} function to enter those directories, | |
15833 since they always lead us, directly or indirectly, to the current | |
15834 directory. | |
15835 | |
15836 Consequently, our @code{files-in-below-directory} function must do | |
15837 several tasks: | |
15838 | |
15839 @itemize @bullet | |
15840 @item | |
15841 Check to see whether it is looking at a filename that ends in | |
15842 @samp{.el}; and if so, add its name to a list. | |
15843 | |
15844 @item | |
15845 Check to see whether it is looking at a filename that is the name of a | |
15846 directory; and if so, | |
15847 | |
15848 @itemize @minus | |
15849 @item | |
15850 Check to see whether it is looking at @file{.} or @file{..}; and if | |
15851 so skip it. | |
15852 | |
15853 @item | |
15854 Or else, go into that directory and repeat the process. | |
15855 @end itemize | |
15856 @end itemize | |
15857 | |
15858 Let's write a function definition to do these tasks. We will use a | |
15859 @code{while} loop to move from one filename to another within a | |
15860 directory, checking what needs to be done; and we will use a recursive | |
15861 call to repeat the actions on each sub-directory. The recursive | |
15862 pattern is `accumulate' | |
15863 (@pxref{Accumulate, , Recursive Pattern: @emph{accumulate}}), | |
15864 using @code{append} as the combiner. | |
15865 | |
15866 @ignore | |
15867 (directory-files "/usr/local/src/emacs/lisp/" t "\\.el$") | |
15868 (shell-command "find /usr/local/src/emacs/lisp/ -name '*.el'") | |
15869 | |
15870 (directory-files "/usr/local/share/emacs/22.1.1/lisp/" t "\\.el$") | |
15871 (shell-command "find /usr/local/share/emacs/22.1.1/lisp/ -name '*.el'") | |
15872 @end ignore | |
15873 | |
15874 @c /usr/local/share/emacs/22.1.1/lisp/ | |
15875 | |
15876 @need 800 | |
15877 Here is the function: | |
15878 | |
15879 @smallexample | |
15880 @group | |
15881 (defun files-in-below-directory (directory) | |
15882 "List the .el files in DIRECTORY and in its sub-directories." | |
15883 ;; Although the function will be used non-interactively, | |
15884 ;; it will be easier to test if we make it interactive. | |
15885 ;; The directory will have a name such as | |
15886 ;; "/usr/local/share/emacs/22.1.1/lisp/" | |
15887 (interactive "DDirectory name: ") | |
15888 @end group | |
15889 @group | |
15890 (let (el-files-list | |
15891 (current-directory-list | |
15892 (directory-files-and-attributes directory t))) | |
15893 ;; while we are in the current directory | |
15894 (while current-directory-list | |
15895 @end group | |
15896 @group | |
15897 (cond | |
15898 ;; check to see whether filename ends in `.el' | |
15899 ;; and if so, append its name to a list. | |
15900 ((equal ".el" (substring (car (car current-directory-list)) -3)) | |
15901 (setq el-files-list | |
15902 (cons (car (car current-directory-list)) el-files-list))) | |
15903 @end group | |
15904 @group | |
15905 ;; check whether filename is that of a directory | |
15906 ((eq t (car (cdr (car current-directory-list)))) | |
15907 ;; decide whether to skip or recurse | |
15908 (if | |
15909 (equal "." | |
15910 (substring (car (car current-directory-list)) -1)) | |
15911 ;; then do nothing since filename is that of | |
15912 ;; current directory or parent, "." or ".." | |
15913 () | |
15914 @end group | |
15915 @group | |
15916 ;; else descend into the directory and repeat the process | |
15917 (setq el-files-list | |
15918 (append | |
15919 (files-in-below-directory | |
15920 (car (car current-directory-list))) | |
15921 el-files-list))))) | |
15922 ;; move to the next filename in the list; this also | |
15923 ;; shortens the list so the while loop eventually comes to an end | |
15924 (setq current-directory-list (cdr current-directory-list))) | |
15925 ;; return the filenames | |
15926 el-files-list)) | |
15927 @end group | |
15928 @end smallexample | |
15929 | |
15930 @c (files-in-below-directory "/usr/local/src/emacs/lisp/") | |
15931 @c (files-in-below-directory "/usr/local/share/emacs/22.1.1/lisp/") | |
15932 | |
15933 The @code{files-in-below-directory} @code{directory-files} function | |
15934 takes one argument, the name of a directory. | |
15935 | |
15936 @need 1250 | |
15937 Thus, on my system, | |
15938 | |
15939 @c (length (files-in-below-directory "/usr/local/src/emacs/lisp/")) | |
15940 | |
15941 @c !!! 22.1.1 lisp sources location here | |
15942 @smallexample | |
15943 @group | |
15944 (length | |
15945 (files-in-below-directory "/usr/local/share/emacs/22.1.1/lisp/")) | |
15946 @end group | |
15947 @end smallexample | |
15948 | |
15949 @noindent | |
15950 tells me that in and below my Lisp sources directory are 1031 | |
15951 @samp{.el} files. | |
15952 | |
15953 @code{files-in-below-directory} returns a list in reverse alphabetical | |
15954 order. An expression to sort the list in alphabetical order looks | |
15955 like this: | |
15956 | |
15957 @smallexample | |
15958 @group | |
15959 (sort | |
15960 (files-in-below-directory "/usr/local/share/emacs/22.1.1/lisp/") | |
15961 'string-lessp) | |
15962 @end group | |
15963 @end smallexample | |
15964 | |
15965 @ignore | |
15966 (defun test () | |
15967 "Test how long it takes to find lengths of all sorted elisp defuns." | |
15968 (insert "\n" (current-time-string) "\n") | |
15969 (sit-for 0) | |
15970 (sort | |
15971 (recursive-lengths-list-many-files | |
15972 (files-in-below-directory "/usr/local/src/emacs/lisp/")) | |
15973 '<) | |
15974 (insert (format "%s" (current-time-string)))) | |
15975 @end ignore | |
15976 | |
15977 @node Counting function definitions, , Files List, Prepare the data | |
15978 @subsection Counting function definitions | |
15979 | |
15980 Our immediate goal is to generate a list that tells us how many | |
15981 function definitions contain fewer than 10 words and symbols, how many | |
15982 contain between 10 and 19 words and symbols, how many contain between | |
15983 20 and 29 words and symbols, and so on. | |
15984 | |
15985 With a sorted list of numbers, this is easy: count how many elements | |
15986 of the list are smaller than 10, then, after moving past the numbers | |
15987 just counted, count how many are smaller than 20, then, after moving | |
15988 past the numbers just counted, count how many are smaller than 30, and | |
15989 so on. Each of the numbers, 10, 20, 30, 40, and the like, is one | |
15990 larger than the top of that range. We can call the list of such | |
15991 numbers the @code{top-of-ranges} list. | |
15992 | |
15993 @need 1200 | |
15994 If we wished, we could generate this list automatically, but it is | |
15995 simpler to write a list manually. Here it is: | |
15996 @vindex top-of-ranges | |
15997 | |
15998 @smallexample | |
15999 @group | |
16000 (defvar top-of-ranges | |
16001 '(10 20 30 40 50 | |
16002 60 70 80 90 100 | |
16003 110 120 130 140 150 | |
16004 160 170 180 190 200 | |
16005 210 220 230 240 250 | |
16006 260 270 280 290 300) | |
16007 "List specifying ranges for `defuns-per-range'.") | |
16008 @end group | |
16009 @end smallexample | |
16010 | |
16011 To change the ranges, we edit this list. | |
16012 | |
16013 Next, we need to write the function that creates the list of the | |
16014 number of definitions within each range. Clearly, this function must | |
16015 take the @code{sorted-lengths} and the @code{top-of-ranges} lists | |
16016 as arguments. | |
16017 | |
16018 The @code{defuns-per-range} function must do two things again and | |
16019 again: it must count the number of definitions within a range | |
16020 specified by the current top-of-range value; and it must shift to the | |
16021 next higher value in the @code{top-of-ranges} list after counting the | |
16022 number of definitions in the current range. Since each of these | |
16023 actions is repetitive, we can use @code{while} loops for the job. | |
16024 One loop counts the number of definitions in the range defined by the | |
16025 current top-of-range value, and the other loop selects each of the | |
16026 top-of-range values in turn. | |
16027 | |
16028 Several entries of the @code{sorted-lengths} list are counted for each | |
16029 range; this means that the loop for the @code{sorted-lengths} list | |
16030 will be inside the loop for the @code{top-of-ranges} list, like a | |
16031 small gear inside a big gear. | |
16032 | |
16033 The inner loop counts the number of definitions within the range. It | |
16034 is a simple counting loop of the type we have seen before. | |
16035 (@xref{Incrementing Loop, , A loop with an incrementing counter}.) | |
16036 The true-or-false test of the loop tests whether the value from the | |
16037 @code{sorted-lengths} list is smaller than the current value of the | |
16038 top of the range. If it is, the function increments the counter and | |
16039 tests the next value from the @code{sorted-lengths} list. | |
16040 | |
16041 @need 1250 | |
16042 The inner loop looks like this: | |
16043 | |
16044 @smallexample | |
16045 @group | |
16046 (while @var{length-element-smaller-than-top-of-range} | |
16047 (setq number-within-range (1+ number-within-range)) | |
16048 (setq sorted-lengths (cdr sorted-lengths))) | |
16049 @end group | |
16050 @end smallexample | |
16051 | |
16052 The outer loop must start with the lowest value of the | |
16053 @code{top-of-ranges} list, and then be set to each of the succeeding | |
16054 higher values in turn. This can be done with a loop like this: | |
16055 | |
16056 @smallexample | |
16057 @group | |
16058 (while top-of-ranges | |
16059 @var{body-of-loop}@dots{} | |
16060 (setq top-of-ranges (cdr top-of-ranges))) | |
16061 @end group | |
16062 @end smallexample | |
16063 | |
16064 @need 1200 | |
16065 Put together, the two loops look like this: | |
16066 | |
16067 @smallexample | |
16068 @group | |
16069 (while top-of-ranges | |
16070 | |
16071 ;; @r{Count the number of elements within the current range.} | |
16072 (while @var{length-element-smaller-than-top-of-range} | |
16073 (setq number-within-range (1+ number-within-range)) | |
16074 (setq sorted-lengths (cdr sorted-lengths))) | |
16075 | |
16076 ;; @r{Move to next range.} | |
16077 (setq top-of-ranges (cdr top-of-ranges))) | |
16078 @end group | |
16079 @end smallexample | |
16080 | |
16081 In addition, in each circuit of the outer loop, Emacs should record | |
16082 the number of definitions within that range (the value of | |
16083 @code{number-within-range}) in a list. We can use @code{cons} for | |
16084 this purpose. (@xref{cons, , @code{cons}}.) | |
16085 | |
16086 The @code{cons} function works fine, except that the list it | |
16087 constructs will contain the number of definitions for the highest | |
16088 range at its beginning and the number of definitions for the lowest | |
16089 range at its end. This is because @code{cons} attaches new elements | |
16090 of the list to the beginning of the list, and since the two loops are | |
16091 working their way through the lengths' list from the lower end first, | |
16092 the @code{defuns-per-range-list} will end up largest number first. | |
16093 But we will want to print our graph with smallest values first and the | |
16094 larger later. The solution is to reverse the order of the | |
16095 @code{defuns-per-range-list}. We can do this using the | |
16096 @code{nreverse} function, which reverses the order of a list. | |
16097 @findex nreverse | |
16098 | |
16099 @need 800 | |
16100 For example, | |
16101 | |
16102 @smallexample | |
16103 (nreverse '(1 2 3 4)) | |
16104 @end smallexample | |
16105 | |
16106 @need 800 | |
16107 @noindent | |
16108 produces: | |
16109 | |
16110 @smallexample | |
16111 (4 3 2 1) | |
16112 @end smallexample | |
16113 | |
16114 Note that the @code{nreverse} function is ``destructive''---that is, | |
16115 it changes the list to which it is applied; this contrasts with the | |
16116 @code{car} and @code{cdr} functions, which are non-destructive. In | |
16117 this case, we do not want the original @code{defuns-per-range-list}, | |
16118 so it does not matter that it is destroyed. (The @code{reverse} | |
16119 function provides a reversed copy of a list, leaving the original list | |
16120 as is.) | |
16121 @findex reverse | |
16122 | |
16123 @need 1250 | |
16124 Put all together, the @code{defuns-per-range} looks like this: | |
16125 | |
16126 @smallexample | |
16127 @group | |
16128 (defun defuns-per-range (sorted-lengths top-of-ranges) | |
16129 "SORTED-LENGTHS defuns in each TOP-OF-RANGES range." | |
16130 (let ((top-of-range (car top-of-ranges)) | |
16131 (number-within-range 0) | |
16132 defuns-per-range-list) | |
16133 @end group | |
16134 | |
16135 @group | |
16136 ;; @r{Outer loop.} | |
16137 (while top-of-ranges | |
16138 @end group | |
16139 | |
16140 @group | |
16141 ;; @r{Inner loop.} | |
16142 (while (and | |
16143 ;; @r{Need number for numeric test.} | |
16144 (car sorted-lengths) | |
16145 (< (car sorted-lengths) top-of-range)) | |
16146 @end group | |
16147 | |
16148 @group | |
16149 ;; @r{Count number of definitions within current range.} | |
16150 (setq number-within-range (1+ number-within-range)) | |
16151 (setq sorted-lengths (cdr sorted-lengths))) | |
16152 | |
16153 ;; @r{Exit inner loop but remain within outer loop.} | |
16154 @end group | |
16155 | |
16156 @group | |
16157 (setq defuns-per-range-list | |
16158 (cons number-within-range defuns-per-range-list)) | |
16159 (setq number-within-range 0) ; @r{Reset count to zero.} | |
16160 @end group | |
16161 | |
16162 @group | |
16163 ;; @r{Move to next range.} | |
16164 (setq top-of-ranges (cdr top-of-ranges)) | |
16165 ;; @r{Specify next top of range value.} | |
16166 (setq top-of-range (car top-of-ranges))) | |
16167 @end group | |
16168 | |
16169 @group | |
16170 ;; @r{Exit outer loop and count the number of defuns larger than} | |
16171 ;; @r{ the largest top-of-range value.} | |
16172 (setq defuns-per-range-list | |
16173 (cons | |
16174 (length sorted-lengths) | |
16175 defuns-per-range-list)) | |
16176 @end group | |
16177 | |
16178 @group | |
16179 ;; @r{Return a list of the number of definitions within each range,} | |
16180 ;; @r{ smallest to largest.} | |
16181 (nreverse defuns-per-range-list))) | |
16182 @end group | |
16183 @end smallexample | |
16184 | |
16185 @need 1200 | |
16186 @noindent | |
16187 The function is straightforward except for one subtle feature. The | |
16188 true-or-false test of the inner loop looks like this: | |
16189 | |
16190 @smallexample | |
16191 @group | |
16192 (and (car sorted-lengths) | |
16193 (< (car sorted-lengths) top-of-range)) | |
16194 @end group | |
16195 @end smallexample | |
16196 | |
16197 @need 800 | |
16198 @noindent | |
16199 instead of like this: | |
16200 | |
16201 @smallexample | |
16202 (< (car sorted-lengths) top-of-range) | |
16203 @end smallexample | |
16204 | |
16205 The purpose of the test is to determine whether the first item in the | |
16206 @code{sorted-lengths} list is less than the value of the top of the | |
16207 range. | |
16208 | |
16209 The simple version of the test works fine unless the | |
16210 @code{sorted-lengths} list has a @code{nil} value. In that case, the | |
16211 @code{(car sorted-lengths)} expression function returns | |
16212 @code{nil}. The @code{<} function cannot compare a number to | |
16213 @code{nil}, which is an empty list, so Emacs signals an error and | |
16214 stops the function from attempting to continue to execute. | |
16215 | |
16216 The @code{sorted-lengths} list always becomes @code{nil} when the | |
16217 counter reaches the end of the list. This means that any attempt to | |
16218 use the @code{defuns-per-range} function with the simple version of | |
16219 the test will fail. | |
16220 | |
16221 We solve the problem by using the @code{(car sorted-lengths)} | |
16222 expression in conjunction with the @code{and} expression. The | |
16223 @code{(car sorted-lengths)} expression returns a non-@code{nil} | |
16224 value so long as the list has at least one number within it, but | |
16225 returns @code{nil} if the list is empty. The @code{and} expression | |
16226 first evaluates the @code{(car sorted-lengths)} expression, and | |
16227 if it is @code{nil}, returns false @emph{without} evaluating the | |
16228 @code{<} expression. But if the @code{(car sorted-lengths)} | |
16229 expression returns a non-@code{nil} value, the @code{and} expression | |
16230 evaluates the @code{<} expression, and returns that value as the value | |
16231 of the @code{and} expression. | |
16232 | |
16233 @c colon in printed section title causes problem in Info cross reference | |
16234 This way, we avoid an error. | |
16235 @iftex | |
16236 @noindent | |
16237 (For information about @code{and}, see | |
16238 @ref{kill-new function, , The @code{kill-new} function}.) | |
16239 @end iftex | |
16240 @ifinfo | |
16241 @noindent | |
16242 (@xref{kill-new function, , The @code{kill-new} function}, for | |
16243 information about @code{and}.) | |
16244 @end ifinfo | |
16245 | |
16246 Here is a short test of the @code{defuns-per-range} function. First, | |
16247 evaluate the expression that binds (a shortened) | |
16248 @code{top-of-ranges} list to the list of values, then evaluate the | |
16249 expression for binding the @code{sorted-lengths} list, and then | |
16250 evaluate the @code{defuns-per-range} function. | |
16251 | |
16252 @smallexample | |
16253 @group | |
16254 ;; @r{(Shorter list than we will use later.)} | |
16255 (setq top-of-ranges | |
16256 '(110 120 130 140 150 | |
16257 160 170 180 190 200)) | |
16258 | |
16259 (setq sorted-lengths | |
16260 '(85 86 110 116 122 129 154 176 179 200 265 300 300)) | |
16261 | |
16262 (defuns-per-range sorted-lengths top-of-ranges) | |
16263 @end group | |
16264 @end smallexample | |
16265 | |
16266 @need 800 | |
16267 @noindent | |
16268 The list returned looks like this: | |
16269 | |
16270 @smallexample | |
16271 (2 2 2 0 0 1 0 2 0 0 4) | |
16272 @end smallexample | |
16273 | |
16274 @noindent | |
16275 Indeed, there are two elements of the @code{sorted-lengths} list | |
16276 smaller than 110, two elements between 110 and 119, two elements | |
16277 between 120 and 129, and so on. There are four elements with a value | |
16278 of 200 or larger. | |
16279 | |
16280 @c The next step is to turn this numbers' list into a graph. | |
16281 @node Readying a Graph, Emacs Initialization, Words in a defun, Top | |
16282 @chapter Readying a Graph | |
16283 @cindex Readying a graph | |
16284 @cindex Graph prototype | |
16285 @cindex Prototype graph | |
16286 @cindex Body of graph | |
16287 | |
16288 Our goal is to construct a graph showing the numbers of function | |
16289 definitions of various lengths in the Emacs lisp sources. | |
16290 | |
16291 As a practical matter, if you were creating a graph, you would | |
16292 probably use a program such as @code{gnuplot} to do the job. | |
16293 (@code{gnuplot} is nicely integrated into GNU Emacs.) In this case, | |
16294 however, we create one from scratch, and in the process we will | |
16295 re-acquaint ourselves with some of what we learned before and learn | |
16296 more. | |
16297 | |
16298 In this chapter, we will first write a simple graph printing function. | |
16299 This first definition will be a @dfn{prototype}, a rapidly written | |
16300 function that enables us to reconnoiter this unknown graph-making | |
16301 territory. We will discover dragons, or find that they are myth. | |
16302 After scouting the terrain, we will feel more confident and enhance | |
16303 the function to label the axes automatically. | |
16304 | |
16305 @menu | |
16306 * Columns of a graph:: | |
16307 * graph-body-print:: How to print the body of a graph. | |
16308 * recursive-graph-body-print:: | |
16309 * Printed Axes:: | |
16310 * Line Graph Exercise:: | |
16311 @end menu | |
16312 | |
16313 @node Columns of a graph, graph-body-print, Readying a Graph, Readying a Graph | |
16314 @ifnottex | |
16315 @unnumberedsec Printing the Columns of a Graph | |
16316 @end ifnottex | |
16317 | |
16318 Since Emacs is designed to be flexible and work with all kinds of | |
16319 terminals, including character-only terminals, the graph will need to | |
16320 be made from one of the `typewriter' symbols. An asterisk will do; as | |
16321 we enhance the graph-printing function, we can make the choice of | |
16322 symbol a user option. | |
16323 | |
16324 We can call this function @code{graph-body-print}; it will take a | |
16325 @code{numbers-list} as its only argument. At this stage, we will not | |
16326 label the graph, but only print its body. | |
16327 | |
16328 The @code{graph-body-print} function inserts a vertical column of | |
16329 asterisks for each element in the @code{numbers-list}. The height of | |
16330 each line is determined by the value of that element of the | |
16331 @code{numbers-list}. | |
16332 | |
16333 Inserting columns is a repetitive act; that means that this function can | |
16334 be written either with a @code{while} loop or recursively. | |
16335 | |
16336 Our first challenge is to discover how to print a column of asterisks. | |
16337 Usually, in Emacs, we print characters onto a screen horizontally, | |
16338 line by line, by typing. We have two routes we can follow: write our | |
16339 own column-insertion function or discover whether one exists in Emacs. | |
16340 | |
16341 To see whether there is one in Emacs, we can use the @kbd{M-x apropos} | |
16342 command. This command is like the @kbd{C-h a} (@code{command-apropos}) | |
16343 command, except that the latter finds only those functions that are | |
16344 commands. The @kbd{M-x apropos} command lists all symbols that match | |
16345 a regular expression, including functions that are not interactive. | |
16346 @findex apropos | |
16347 | |
16348 What we want to look for is some command that prints or inserts | |
16349 columns. Very likely, the name of the function will contain either | |
16350 the word `print' or the word `insert' or the word `column'. | |
16351 Therefore, we can simply type @kbd{M-x apropos RET | |
16352 print\|insert\|column RET} and look at the result. On my system, this | |
16353 command once too takes quite some time, and then produced a list of 79 | |
16354 functions and variables. Now it does not take much time at all and | |
16355 produces a list of 211 functions and variables. Scanning down the | |
16356 list, the only function that looks as if it might do the job is | |
16357 @code{insert-rectangle}. | |
16358 | |
16359 @need 1200 | |
16360 Indeed, this is the function we want; its documentation says: | |
16361 | |
16362 @smallexample | |
16363 @group | |
16364 insert-rectangle: | |
16365 Insert text of RECTANGLE with upper left corner at point. | |
16366 RECTANGLE's first line is inserted at point, | |
16367 its second line is inserted at a point vertically under point, etc. | |
16368 RECTANGLE should be a list of strings. | |
16369 After this command, the mark is at the upper left corner | |
16370 and point is at the lower right corner. | |
16371 @end group | |
16372 @end smallexample | |
16373 | |
16374 We can run a quick test, to make sure it does what we expect of it. | |
16375 | |
16376 Here is the result of placing the cursor after the | |
16377 @code{insert-rectangle} expression and typing @kbd{C-u C-x C-e} | |
16378 (@code{eval-last-sexp}). The function inserts the strings | |
16379 @samp{"first"}, @samp{"second"}, and @samp{"third"} at and below | |
16380 point. Also the function returns @code{nil}. | |
16381 | |
16382 @smallexample | |
16383 @group | |
16384 (insert-rectangle '("first" "second" "third"))first | |
16385 second | |
16386 thirdnil | |
16387 @end group | |
16388 @end smallexample | |
16389 | |
16390 @noindent | |
16391 Of course, we won't be inserting the text of the | |
16392 @code{insert-rectangle} expression itself into the buffer in which we | |
16393 are making the graph, but will call the function from our program. We | |
16394 shall, however, have to make sure that point is in the buffer at the | |
16395 place where the @code{insert-rectangle} function will insert its | |
16396 column of strings. | |
16397 | |
16398 If you are reading this in Info, you can see how this works by | |
16399 switching to another buffer, such as the @file{*scratch*} buffer, | |
16400 placing point somewhere in the buffer, typing @kbd{M-:}, typing the | |
16401 @code{insert-rectangle} expression into the minibuffer at the prompt, | |
16402 and then typing @key{RET}. This causes Emacs to evaluate the | |
16403 expression in the minibuffer, but to use as the value of point the | |
16404 position of point in the @file{*scratch*} buffer. (@kbd{M-:} is the | |
16405 keybinding for @code{eval-expression}. Also, @code{nil} does not | |
16406 appear in the @file{*scratch*} buffer since the expression is | |
16407 evaluated in the minibuffer.) | |
16408 | |
16409 We find when we do this that point ends up at the end of the last | |
16410 inserted line---that is to say, this function moves point as a | |
16411 side-effect. If we were to repeat the command, with point at this | |
16412 position, the next insertion would be below and to the right of the | |
16413 previous insertion. We don't want this! If we are going to make a | |
16414 bar graph, the columns need to be beside each other. | |
16415 | |
16416 So we discover that each cycle of the column-inserting @code{while} | |
16417 loop must reposition point to the place we want it, and that place | |
16418 will be at the top, not the bottom, of the column. Moreover, we | |
16419 remember that when we print a graph, we do not expect all the columns | |
16420 to be the same height. This means that the top of each column may be | |
16421 at a different height from the previous one. We cannot simply | |
16422 reposition point to the same line each time, but moved over to the | |
16423 right---or perhaps we can@dots{} | |
16424 | |
16425 We are planning to make the columns of the bar graph out of asterisks. | |
16426 The number of asterisks in the column is the number specified by the | |
16427 current element of the @code{numbers-list}. We need to construct a | |
16428 list of asterisks of the right length for each call to | |
16429 @code{insert-rectangle}. If this list consists solely of the requisite | |
16430 number of asterisks, then we will have position point the right number | |
16431 of lines above the base for the graph to print correctly. This could | |
16432 be difficult. | |
16433 | |
16434 Alternatively, if we can figure out some way to pass | |
16435 @code{insert-rectangle} a list of the same length each time, then we | |
16436 can place point on the same line each time, but move it over one | |
16437 column to the right for each new column. If we do this, however, some | |
16438 of the entries in the list passed to @code{insert-rectangle} must be | |
16439 blanks rather than asterisks. For example, if the maximum height of | |
16440 the graph is 5, but the height of the column is 3, then | |
16441 @code{insert-rectangle} requires an argument that looks like this: | |
16442 | |
16443 @smallexample | |
16444 (" " " " "*" "*" "*") | |
16445 @end smallexample | |
16446 | |
16447 This last proposal is not so difficult, so long as we can determine | |
16448 the column height. There are two ways for us to specify the column | |
16449 height: we can arbitrarily state what it will be, which would work | |
16450 fine for graphs of that height; or we can search through the list of | |
16451 numbers and use the maximum height of the list as the maximum height | |
16452 of the graph. If the latter operation were difficult, then the former | |
16453 procedure would be easiest, but there is a function built into Emacs | |
16454 that determines the maximum of its arguments. We can use that | |
16455 function. The function is called @code{max} and it returns the | |
16456 largest of all its arguments, which must be numbers. Thus, for | |
16457 example, | |
16458 | |
16459 @smallexample | |
16460 (max 3 4 6 5 7 3) | |
16461 @end smallexample | |
16462 | |
16463 @noindent | |
16464 returns 7. (A corresponding function called @code{min} returns the | |
16465 smallest of all its arguments.) | |
16466 @findex max | |
16467 @findex min | |
16468 | |
16469 However, we cannot simply call @code{max} on the @code{numbers-list}; | |
16470 the @code{max} function expects numbers as its argument, not a list of | |
16471 numbers. Thus, the following expression, | |
16472 | |
16473 @smallexample | |
16474 (max '(3 4 6 5 7 3)) | |
16475 @end smallexample | |
16476 | |
16477 @need 800 | |
16478 @noindent | |
16479 produces the following error message; | |
16480 | |
16481 @smallexample | |
16482 Wrong type of argument: number-or-marker-p, (3 4 6 5 7 3) | |
16483 @end smallexample | |
16484 | |
16485 @findex apply | |
16486 We need a function that passes a list of arguments to a function. | |
16487 This function is @code{apply}. This function `applies' its first | |
16488 argument (a function) to its remaining arguments, the last of which | |
16489 may be a list. | |
16490 | |
16491 @need 1250 | |
16492 For example, | |
16493 | |
16494 @smallexample | |
16495 (apply 'max 3 4 7 3 '(4 8 5)) | |
16496 @end smallexample | |
16497 | |
16498 @noindent | |
16499 returns 8. | |
16500 | |
16501 (Incidentally, I don't know how you would learn of this function | |
16502 without a book such as this. It is possible to discover other | |
16503 functions, like @code{search-forward} or @code{insert-rectangle}, by | |
16504 guessing at a part of their names and then using @code{apropos}. Even | |
16505 though its base in metaphor is clear---`apply' its first argument to | |
16506 the rest---I doubt a novice would come up with that particular word | |
16507 when using @code{apropos} or other aid. Of course, I could be wrong; | |
16508 after all, the function was first named by someone who had to invent | |
16509 it.) | |
16510 | |
16511 The second and subsequent arguments to @code{apply} are optional, so | |
16512 we can use @code{apply} to call a function and pass the elements of a | |
16513 list to it, like this, which also returns 8: | |
16514 | |
16515 @smallexample | |
16516 (apply 'max '(4 8 5)) | |
16517 @end smallexample | |
16518 | |
16519 This latter way is how we will use @code{apply}. The | |
16520 @code{recursive-lengths-list-many-files} function returns a numbers' | |
16521 list to which we can apply @code{max} (we could also apply @code{max} to | |
16522 the sorted numbers' list; it does not matter whether the list is | |
16523 sorted or not.) | |
16524 | |
16525 @need 800 | |
16526 Hence, the operation for finding the maximum height of the graph is this: | |
16527 | |
16528 @smallexample | |
16529 (setq max-graph-height (apply 'max numbers-list)) | |
16530 @end smallexample | |
16531 | |
16532 Now we can return to the question of how to create a list of strings | |
16533 for a column of the graph. Told the maximum height of the graph | |
16534 and the number of asterisks that should appear in the column, the | |
16535 function should return a list of strings for the | |
16536 @code{insert-rectangle} command to insert. | |
16537 | |
16538 Each column is made up of asterisks or blanks. Since the function is | |
16539 passed the value of the height of the column and the number of | |
16540 asterisks in the column, the number of blanks can be found by | |
16541 subtracting the number of asterisks from the height of the column. | |
16542 Given the number of blanks and the number of asterisks, two | |
16543 @code{while} loops can be used to construct the list: | |
16544 | |
16545 @smallexample | |
16546 @group | |
16547 ;;; @r{First version.} | |
16548 (defun column-of-graph (max-graph-height actual-height) | |
16549 "Return list of strings that is one column of a graph." | |
16550 (let ((insert-list nil) | |
16551 (number-of-top-blanks | |
16552 (- max-graph-height actual-height))) | |
16553 @end group | |
16554 | |
16555 @group | |
16556 ;; @r{Fill in asterisks.} | |
16557 (while (> actual-height 0) | |
16558 (setq insert-list (cons "*" insert-list)) | |
16559 (setq actual-height (1- actual-height))) | |
16560 @end group | |
16561 | |
16562 @group | |
16563 ;; @r{Fill in blanks.} | |
16564 (while (> number-of-top-blanks 0) | |
16565 (setq insert-list (cons " " insert-list)) | |
16566 (setq number-of-top-blanks | |
16567 (1- number-of-top-blanks))) | |
16568 @end group | |
16569 | |
16570 @group | |
16571 ;; @r{Return whole list.} | |
16572 insert-list)) | |
16573 @end group | |
16574 @end smallexample | |
16575 | |
16576 If you install this function and then evaluate the following | |
16577 expression you will see that it returns the list as desired: | |
16578 | |
16579 @smallexample | |
16580 (column-of-graph 5 3) | |
16581 @end smallexample | |
16582 | |
16583 @need 800 | |
16584 @noindent | |
16585 returns | |
16586 | |
16587 @smallexample | |
16588 (" " " " "*" "*" "*") | |
16589 @end smallexample | |
16590 | |
16591 As written, @code{column-of-graph} contains a major flaw: the symbols | |
16592 used for the blank and for the marked entries in the column are | |
16593 `hard-coded' as a space and asterisk. This is fine for a prototype, | |
16594 but you, or another user, may wish to use other symbols. For example, | |
16595 in testing the graph function, you many want to use a period in place | |
16596 of the space, to make sure the point is being repositioned properly | |
16597 each time the @code{insert-rectangle} function is called; or you might | |
16598 want to substitute a @samp{+} sign or other symbol for the asterisk. | |
16599 You might even want to make a graph-column that is more than one | |
16600 display column wide. The program should be more flexible. The way to | |
16601 do that is to replace the blank and the asterisk with two variables | |
16602 that we can call @code{graph-blank} and @code{graph-symbol} and define | |
16603 those variables separately. | |
16604 | |
16605 Also, the documentation is not well written. These considerations | |
16606 lead us to the second version of the function: | |
16607 | |
16608 @smallexample | |
16609 @group | |
16610 (defvar graph-symbol "*" | |
16611 "String used as symbol in graph, usually an asterisk.") | |
16612 @end group | |
16613 | |
16614 @group | |
16615 (defvar graph-blank " " | |
16616 "String used as blank in graph, usually a blank space. | |
16617 graph-blank must be the same number of columns wide | |
16618 as graph-symbol.") | |
16619 @end group | |
16620 @end smallexample | |
16621 | |
16622 @noindent | |
16623 (For an explanation of @code{defvar}, see | |
16624 @ref{defvar, , Initializing a Variable with @code{defvar}}.) | |
16625 | |
16626 @smallexample | |
16627 @group | |
16628 ;;; @r{Second version.} | |
16629 (defun column-of-graph (max-graph-height actual-height) | |
16630 "Return MAX-GRAPH-HEIGHT strings; ACTUAL-HEIGHT are graph-symbols. | |
16631 | |
16632 @end group | |
16633 @group | |
16634 The graph-symbols are contiguous entries at the end | |
16635 of the list. | |
16636 The list will be inserted as one column of a graph. | |
16637 The strings are either graph-blank or graph-symbol." | |
16638 @end group | |
16639 | |
16640 @group | |
16641 (let ((insert-list nil) | |
16642 (number-of-top-blanks | |
16643 (- max-graph-height actual-height))) | |
16644 @end group | |
16645 | |
16646 @group | |
16647 ;; @r{Fill in @code{graph-symbols}.} | |
16648 (while (> actual-height 0) | |
16649 (setq insert-list (cons graph-symbol insert-list)) | |
16650 (setq actual-height (1- actual-height))) | |
16651 @end group | |
16652 | |
16653 @group | |
16654 ;; @r{Fill in @code{graph-blanks}.} | |
16655 (while (> number-of-top-blanks 0) | |
16656 (setq insert-list (cons graph-blank insert-list)) | |
16657 (setq number-of-top-blanks | |
16658 (1- number-of-top-blanks))) | |
16659 | |
16660 ;; @r{Return whole list.} | |
16661 insert-list)) | |
16662 @end group | |
16663 @end smallexample | |
16664 | |
16665 If we wished, we could rewrite @code{column-of-graph} a third time to | |
16666 provide optionally for a line graph as well as for a bar graph. This | |
16667 would not be hard to do. One way to think of a line graph is that it | |
16668 is no more than a bar graph in which the part of each bar that is | |
16669 below the top is blank. To construct a column for a line graph, the | |
16670 function first constructs a list of blanks that is one shorter than | |
16671 the value, then it uses @code{cons} to attach a graph symbol to the | |
16672 list; then it uses @code{cons} again to attach the `top blanks' to | |
16673 the list. | |
16674 | |
16675 It is easy to see how to write such a function, but since we don't | |
16676 need it, we will not do it. But the job could be done, and if it were | |
16677 done, it would be done with @code{column-of-graph}. Even more | |
16678 important, it is worth noting that few changes would have to be made | |
16679 anywhere else. The enhancement, if we ever wish to make it, is | |
16680 simple. | |
16681 | |
16682 Now, finally, we come to our first actual graph printing function. | |
16683 This prints the body of a graph, not the labels for the vertical and | |
16684 horizontal axes, so we can call this @code{graph-body-print}. | |
16685 | |
16686 @node graph-body-print, recursive-graph-body-print, Columns of a graph, Readying a Graph | |
16687 @section The @code{graph-body-print} Function | |
16688 @findex graph-body-print | |
16689 | |
16690 After our preparation in the preceding section, the | |
16691 @code{graph-body-print} function is straightforward. The function | |
16692 will print column after column of asterisks and blanks, using the | |
16693 elements of a numbers' list to specify the number of asterisks in each | |
16694 column. This is a repetitive act, which means we can use a | |
16695 decrementing @code{while} loop or recursive function for the job. In | |
16696 this section, we will write the definition using a @code{while} loop. | |
16697 | |
16698 The @code{column-of-graph} function requires the height of the graph | |
16699 as an argument, so we should determine and record that as a local variable. | |
16700 | |
16701 This leads us to the following template for the @code{while} loop | |
16702 version of this function: | |
16703 | |
16704 @smallexample | |
16705 @group | |
16706 (defun graph-body-print (numbers-list) | |
16707 "@var{documentation}@dots{}" | |
16708 (let ((height @dots{} | |
16709 @dots{})) | |
16710 @end group | |
16711 | |
16712 @group | |
16713 (while numbers-list | |
16714 @var{insert-columns-and-reposition-point} | |
16715 (setq numbers-list (cdr numbers-list))))) | |
16716 @end group | |
16717 @end smallexample | |
16718 | |
16719 @noindent | |
16720 We need to fill in the slots of the template. | |
16721 | |
16722 Clearly, we can use the @code{(apply 'max numbers-list)} expression to | |
16723 determine the height of the graph. | |
16724 | |
16725 The @code{while} loop will cycle through the @code{numbers-list} one | |
16726 element at a time. As it is shortened by the @code{(setq numbers-list | |
16727 (cdr numbers-list))} expression, the @sc{car} of each instance of the | |
16728 list is the value of the argument for @code{column-of-graph}. | |
16729 | |
16730 At each cycle of the @code{while} loop, the @code{insert-rectangle} | |
16731 function inserts the list returned by @code{column-of-graph}. Since | |
16732 the @code{insert-rectangle} function moves point to the lower right of | |
16733 the inserted rectangle, we need to save the location of point at the | |
16734 time the rectangle is inserted, move back to that position after the | |
16735 rectangle is inserted, and then move horizontally to the next place | |
16736 from which @code{insert-rectangle} is called. | |
16737 | |
16738 If the inserted columns are one character wide, as they will be if | |
16739 single blanks and asterisks are used, the repositioning command is | |
16740 simply @code{(forward-char 1)}; however, the width of a column may be | |
16741 greater than one. This means that the repositioning command should be | |
16742 written @code{(forward-char symbol-width)}. The @code{symbol-width} | |
16743 itself is the length of a @code{graph-blank} and can be found using | |
16744 the expression @code{(length graph-blank)}. The best place to bind | |
16745 the @code{symbol-width} variable to the value of the width of graph | |
16746 column is in the varlist of the @code{let} expression. | |
16747 | |
16748 @need 1250 | |
16749 These considerations lead to the following function definition: | |
16750 | |
16751 @smallexample | |
16752 @group | |
16753 (defun graph-body-print (numbers-list) | |
16754 "Print a bar graph of the NUMBERS-LIST. | |
16755 The numbers-list consists of the Y-axis values." | |
16756 | |
16757 (let ((height (apply 'max numbers-list)) | |
16758 (symbol-width (length graph-blank)) | |
16759 from-position) | |
16760 @end group | |
16761 | |
16762 @group | |
16763 (while numbers-list | |
16764 (setq from-position (point)) | |
16765 (insert-rectangle | |
16766 (column-of-graph height (car numbers-list))) | |
16767 (goto-char from-position) | |
16768 (forward-char symbol-width) | |
16769 @end group | |
16770 @group | |
16771 ;; @r{Draw graph column by column.} | |
16772 (sit-for 0) | |
16773 (setq numbers-list (cdr numbers-list))) | |
16774 @end group | |
16775 @group | |
16776 ;; @r{Place point for X axis labels.} | |
16777 (forward-line height) | |
16778 (insert "\n") | |
16779 )) | |
16780 @end group | |
16781 @end smallexample | |
16782 | |
16783 @noindent | |
16784 The one unexpected expression in this function is the | |
16785 @w{@code{(sit-for 0)}} expression in the @code{while} loop. This | |
16786 expression makes the graph printing operation more interesting to | |
16787 watch than it would be otherwise. The expression causes Emacs to | |
16788 `sit' or do nothing for a zero length of time and then redraw the | |
16789 screen. Placed here, it causes Emacs to redraw the screen column by | |
16790 column. Without it, Emacs would not redraw the screen until the | |
16791 function exits. | |
16792 | |
16793 We can test @code{graph-body-print} with a short list of numbers. | |
16794 | |
16795 @enumerate | |
16796 @item | |
16797 Install @code{graph-symbol}, @code{graph-blank}, | |
16798 @code{column-of-graph}, which are in | |
16799 @iftex | |
16800 @ref{Readying a Graph, , Readying a Graph}, | |
16801 @end iftex | |
16802 @ifinfo | |
16803 @ref{Columns of a graph}, | |
16804 @end ifinfo | |
16805 and @code{graph-body-print}. | |
16806 | |
16807 @need 800 | |
16808 @item | |
16809 Copy the following expression: | |
16810 | |
16811 @smallexample | |
16812 (graph-body-print '(1 2 3 4 6 4 3 5 7 6 5 2 3)) | |
16813 @end smallexample | |
16814 | |
16815 @item | |
16816 Switch to the @file{*scratch*} buffer and place the cursor where you | |
16817 want the graph to start. | |
16818 | |
16819 @item | |
16820 Type @kbd{M-:} (@code{eval-expression}). | |
16821 | |
16822 @item | |
16823 Yank the @code{graph-body-print} expression into the minibuffer | |
16824 with @kbd{C-y} (@code{yank)}. | |
16825 | |
16826 @item | |
16827 Press @key{RET} to evaluate the @code{graph-body-print} expression. | |
16828 @end enumerate | |
16829 | |
16830 @need 800 | |
16831 Emacs will print a graph like this: | |
16832 | |
16833 @smallexample | |
16834 @group | |
16835 * | |
16836 * ** | |
16837 * **** | |
16838 *** **** | |
16839 ********* * | |
16840 ************ | |
16841 ************* | |
16842 @end group | |
16843 @end smallexample | |
16844 | |
16845 @node recursive-graph-body-print, Printed Axes, graph-body-print, Readying a Graph | |
16846 @section The @code{recursive-graph-body-print} Function | |
16847 @findex recursive-graph-body-print | |
16848 | |
16849 The @code{graph-body-print} function may also be written recursively. | |
16850 The recursive solution is divided into two parts: an outside `wrapper' | |
16851 that uses a @code{let} expression to determine the values of several | |
16852 variables that need only be found once, such as the maximum height of | |
16853 the graph, and an inside function that is called recursively to print | |
16854 the graph. | |
16855 | |
16856 @need 1250 | |
16857 The `wrapper' is uncomplicated: | |
16858 | |
16859 @smallexample | |
16860 @group | |
16861 (defun recursive-graph-body-print (numbers-list) | |
16862 "Print a bar graph of the NUMBERS-LIST. | |
16863 The numbers-list consists of the Y-axis values." | |
16864 (let ((height (apply 'max numbers-list)) | |
16865 (symbol-width (length graph-blank)) | |
16866 from-position) | |
16867 (recursive-graph-body-print-internal | |
16868 numbers-list | |
16869 height | |
16870 symbol-width))) | |
16871 @end group | |
16872 @end smallexample | |
16873 | |
16874 The recursive function is a little more difficult. It has four parts: | |
16875 the `do-again-test', the printing code, the recursive call, and the | |
16876 `next-step-expression'. The `do-again-test' is a @code{when} | |
16877 expression that determines whether the @code{numbers-list} contains | |
16878 any remaining elements; if it does, the function prints one column of | |
16879 the graph using the printing code and calls itself again. The | |
16880 function calls itself again according to the value produced by the | |
16881 `next-step-expression' which causes the call to act on a shorter | |
16882 version of the @code{numbers-list}. | |
16883 | |
16884 @smallexample | |
16885 @group | |
16886 (defun recursive-graph-body-print-internal | |
16887 (numbers-list height symbol-width) | |
16888 "Print a bar graph. | |
16889 Used within recursive-graph-body-print function." | |
16890 @end group | |
16891 | |
16892 @group | |
16893 (when numbers-list | |
16894 (setq from-position (point)) | |
16895 (insert-rectangle | |
16896 (column-of-graph height (car numbers-list))) | |
16897 @end group | |
16898 @group | |
16899 (goto-char from-position) | |
16900 (forward-char symbol-width) | |
16901 (sit-for 0) ; @r{Draw graph column by column.} | |
16902 (recursive-graph-body-print-internal | |
16903 (cdr numbers-list) height symbol-width))) | |
16904 @end group | |
16905 @end smallexample | |
16906 | |
16907 @need 1250 | |
16908 After installation, this expression can be tested; here is a sample: | |
16909 | |
16910 @smallexample | |
16911 (recursive-graph-body-print '(3 2 5 6 7 5 3 4 6 4 3 2 1)) | |
16912 @end smallexample | |
16913 | |
16914 @need 800 | |
16915 Here is what @code{recursive-graph-body-print} produces: | |
16916 | |
16917 @smallexample | |
16918 @group | |
16919 * | |
16920 ** * | |
16921 **** * | |
16922 **** *** | |
16923 * ********* | |
16924 ************ | |
16925 ************* | |
16926 @end group | |
16927 @end smallexample | |
16928 | |
16929 Either of these two functions, @code{graph-body-print} or | |
16930 @code{recursive-graph-body-print}, create the body of a graph. | |
16931 | |
16932 @node Printed Axes, Line Graph Exercise, recursive-graph-body-print, Readying a Graph | |
16933 @section Need for Printed Axes | |
16934 | |
16935 A graph needs printed axes, so you can orient yourself. For a do-once | |
16936 project, it may be reasonable to draw the axes by hand using Emacs' | |
16937 Picture mode; but a graph drawing function may be used more than once. | |
16938 | |
16939 For this reason, I have written enhancements to the basic | |
16940 @code{print-graph-body} function that automatically print labels for | |
16941 the horizontal and vertical axes. Since the label printing functions | |
16942 do not contain much new material, I have placed their description in | |
16943 an appendix. @xref{Full Graph, , A Graph with Labelled Axes}. | |
16944 | |
16945 @node Line Graph Exercise, , Printed Axes, Readying a Graph | |
16946 @section Exercise | |
16947 | |
16948 Write a line graph version of the graph printing functions. | |
16949 | |
16950 @node Emacs Initialization, Debugging, Readying a Graph, Top | |
16951 @chapter Your @file{.emacs} File | |
16952 @cindex @file{.emacs} file | |
16953 @cindex Customizing your @file{.emacs} file | |
16954 @cindex Initialization file | |
16955 | |
16956 ``You don't have to like Emacs to like it'' -- this seemingly | |
16957 paradoxical statement is the secret of GNU Emacs. The plain, `out of | |
16958 the box' Emacs is a generic tool. Most people who use it, customize | |
16959 it to suit themselves. | |
16960 | |
16961 GNU Emacs is mostly written in Emacs Lisp; this means that by writing | |
16962 expressions in Emacs Lisp you can change or extend Emacs. | |
16963 | |
16964 @menu | |
16965 * Default Configuration:: | |
16966 * Site-wide Init:: You can write site-wide init files. | |
16967 * defcustom:: Emacs will write code for you. | |
16968 * Beginning a .emacs File:: How to write a @code{.emacs file}. | |
16969 * Text and Auto-fill:: Automatically wrap lines. | |
16970 * Mail Aliases:: Use abbreviations for email addresses. | |
16971 * Indent Tabs Mode:: Don't use tabs with @TeX{} | |
16972 * Keybindings:: Create some personal keybindings. | |
16973 * Keymaps:: More about key binding. | |
16974 * Loading Files:: Load (i.e., evaluate) files automatically. | |
16975 * Autoload:: Make functions available. | |
16976 * Simple Extension:: Define a function; bind it to a key. | |
16977 * X11 Colors:: Colors in X. | |
16978 * Miscellaneous:: | |
16979 * Mode Line:: How to customize your mode line. | |
16980 @end menu | |
16981 | |
16982 @node Default Configuration, Site-wide Init, Emacs Initialization, Emacs Initialization | |
16983 @ifnottex | |
16984 @unnumberedsec Emacs' Default Configuration | |
16985 @end ifnottex | |
16986 | |
16987 There are those who appreciate Emacs' default configuration. After | |
16988 all, Emacs starts you in C mode when you edit a C file, starts you in | |
16989 Fortran mode when you edit a Fortran file, and starts you in | |
16990 Fundamental mode when you edit an unadorned file. This all makes | |
16991 sense, if you do not know who is going to use Emacs. Who knows what a | |
16992 person hopes to do with an unadorned file? Fundamental mode is the | |
16993 right default for such a file, just as C mode is the right default for | |
16994 editing C code. (Enough programming languages have syntaxes | |
16995 that enable them to share or nearly share features, so C mode is | |
16996 now provided by by CC mode, the `C Collection'.) | |
16997 | |
16998 But when you do know who is going to use Emacs---you, | |
16999 yourself---then it makes sense to customize Emacs. | |
17000 | |
17001 For example, I seldom want Fundamental mode when I edit an | |
17002 otherwise undistinguished file; I want Text mode. This is why I | |
17003 customize Emacs: so it suits me. | |
17004 | |
17005 You can customize and extend Emacs by writing or adapting a | |
17006 @file{~/.emacs} file. This is your personal initialization file; its | |
17007 contents, written in Emacs Lisp, tell Emacs what to do.@footnote{You | |
17008 may also add @file{.el} to @file{~/.emacs} and call it a | |
17009 @file{~/.emacs.el} file. In the past, you were forbidden to type the | |
17010 extra keystrokes that the name @file{~/.emacs.el} requires, but now | |
17011 you may. The new format is consistent with the Emacs Lisp file | |
17012 naming conventions; the old format saves typing.} | |
17013 | |
17014 A @file{~/.emacs} file contains Emacs Lisp code. You can write this | |
17015 code yourself; or you can use Emacs' @code{customize} feature to write | |
17016 the code for you. You can combine your own expressions and | |
17017 auto-written Customize expressions in your @file{.emacs} file. | |
17018 | |
17019 (I myself prefer to write my own expressions, except for those, | |
17020 particularly fonts, that I find easier to manipulate using the | |
17021 @code{customize} command. I combine the two methods.) | |
17022 | |
17023 Most of this chapter is about writing expressions yourself. It | |
17024 describes a simple @file{.emacs} file; for more information, see | |
17025 @ref{Init File, , The Init File, emacs, The GNU Emacs Manual}, and | |
17026 @ref{Init File, , The Init File, elisp, The GNU Emacs Lisp Reference | |
17027 Manual}. | |
17028 | |
17029 @node Site-wide Init, defcustom, Default Configuration, Emacs Initialization | |
17030 @section Site-wide Initialization Files | |
17031 | |
17032 @cindex @file{default.el} init file | |
17033 @cindex @file{site-init.el} init file | |
17034 @cindex @file{site-load.el} init file | |
17035 In addition to your personal initialization file, Emacs automatically | |
17036 loads various site-wide initialization files, if they exist. These | |
17037 have the same form as your @file{.emacs} file, but are loaded by | |
17038 everyone. | |
17039 | |
17040 Two site-wide initialization files, @file{site-load.el} and | |
17041 @file{site-init.el}, are loaded into Emacs and then `dumped' if a | |
17042 `dumped' version of Emacs is created, as is most common. (Dumped | |
17043 copies of Emacs load more quickly. However, once a file is loaded and | |
17044 dumped, a change to it does not lead to a change in Emacs unless you | |
17045 load it yourself or re-dump Emacs. @xref{Building Emacs, , Building | |
17046 Emacs, elisp, The GNU Emacs Lisp Reference Manual}, and the | |
17047 @file{INSTALL} file.) | |
17048 | |
17049 Three other site-wide initialization files are loaded automatically | |
17050 each time you start Emacs, if they exist. These are | |
17051 @file{site-start.el}, which is loaded @emph{before} your @file{.emacs} | |
17052 file, and @file{default.el}, and the terminal type file, which are both | |
17053 loaded @emph{after} your @file{.emacs} file. | |
17054 | |
17055 Settings and definitions in your @file{.emacs} file will overwrite | |
17056 conflicting settings and definitions in a @file{site-start.el} file, | |
17057 if it exists; but the settings and definitions in a @file{default.el} | |
17058 or terminal type file will overwrite those in your @file{.emacs} file. | |
17059 (You can prevent interference from a terminal type file by setting | |
17060 @code{term-file-prefix} to @code{nil}. @xref{Simple Extension, , A | |
17061 Simple Extension}.) | |
17062 | |
17063 @c Rewritten to avoid overfull hbox. | |
17064 The @file{INSTALL} file that comes in the distribution contains | |
17065 descriptions of the @file{site-init.el} and @file{site-load.el} files. | |
17066 | |
17067 The @file{loadup.el}, @file{startup.el}, and @file{loaddefs.el} files | |
17068 control loading. These files are in the @file{lisp} directory of the | |
17069 Emacs distribution and are worth perusing. | |
17070 | |
17071 The @file{loaddefs.el} file contains a good many suggestions as to | |
17072 what to put into your own @file{.emacs} file, or into a site-wide | |
17073 initialization file. | |
17074 | |
17075 @node defcustom, Beginning a .emacs File, Site-wide Init, Emacs Initialization | |
17076 @section Specifying Variables using @code{defcustom} | |
17077 @findex defcustom | |
17078 | |
17079 You can specify variables using @code{defcustom} so that you and | |
17080 others can then use Emacs' @code{customize} feature to set their | |
17081 values. (You cannot use @code{customize} to write function | |
17082 definitions; but you can write @code{defuns} in your @file{.emacs} | |
17083 file. Indeed, you can write any Lisp expression in your @file{.emacs} | |
17084 file.) | |
17085 | |
17086 The @code{customize} feature depends on the @code{defcustom} special | |
17087 form. Although you can use @code{defvar} or @code{setq} for variables | |
17088 that users set, the @code{defcustom} special form is designed for the | |
17089 job. | |
17090 | |
17091 You can use your knowledge of @code{defvar} for writing the | |
17092 first three arguments for @code{defcustom}. The first argument to | |
17093 @code{defcustom} is the name of the variable. The second argument is | |
17094 the variable's initial value, if any; and this value is set only if | |
17095 the value has not already been set. The third argument is the | |
17096 documentation. | |
17097 | |
17098 The fourth and subsequent arguments to @code{defcustom} specify types | |
17099 and options; these are not featured in @code{defvar}. (These | |
17100 arguments are optional.) | |
17101 | |
17102 Each of these arguments consists of a keyword followed by a value. | |
17103 Each keyword starts with the colon character @samp{:}. | |
17104 | |
17105 @need 1250 | |
17106 For example, the customizable user option variable | |
17107 @code{text-mode-hook} looks like this: | |
17108 | |
17109 @smallexample | |
17110 @group | |
17111 (defcustom text-mode-hook nil | |
17112 "Normal hook run when entering Text mode and many related modes." | |
17113 :type 'hook | |
17114 :options '(turn-on-auto-fill flyspell-mode) | |
17115 :group 'data) | |
17116 @end group | |
17117 @end smallexample | |
17118 | |
17119 @noindent | |
17120 The name of the variable is @code{text-mode-hook}; it has no default | |
17121 value; and its documentation string tells you what it does. | |
17122 | |
17123 The @code{:type} keyword tells Emacs the kind of data to which | |
17124 @code{text-mode-hook} should be set and how to display the value in a | |
17125 Customization buffer. | |
17126 | |
17127 The @code{:options} keyword specifies a suggested list of values for | |
17128 the variable. Usually, @code{:options} applies to a hook. | |
17129 The list is only a suggestion; it is not exclusive; a person who sets | |
17130 the variable may set it to other values; the list shown following the | |
17131 @code{:options} keyword is intended to offer convenient choices to a | |
17132 user. | |
17133 | |
17134 Finally, the @code{:group} keyword tells the Emacs Customization | |
17135 command in which group the variable is located. This tells where to | |
17136 find it. | |
17137 | |
17138 The @code{defcustom} function recognizes more than a dozen keywords. | |
17139 For more information, see @ref{Customization, , Writing Customization | |
17140 Definitions, elisp, The GNU Emacs Lisp Reference Manual}. | |
17141 | |
17142 Consider @code{text-mode-hook} as an example. | |
17143 | |
17144 There are two ways to customize this variable. You can use the | |
17145 customization command or write the appropriate expressions yourself. | |
17146 | |
17147 @need 800 | |
17148 Using the customization command, you can type: | |
17149 | |
17150 @smallexample | |
17151 M-x customize | |
17152 @end smallexample | |
17153 | |
17154 @noindent | |
17155 and find that the group for editing files of data is called `data'. | |
17156 Enter that group. Text Mode Hook is the first member. You can click | |
17157 on its various options, such as @code{turn-on-auto-fill}, to set the | |
17158 values. After you click on the button to | |
17159 | |
17160 @smallexample | |
17161 Save for Future Sessions | |
17162 @end smallexample | |
17163 | |
17164 @noindent | |
17165 Emacs will write an expression into your @file{.emacs} file. | |
17166 It will look like this: | |
17167 | |
17168 @smallexample | |
17169 @group | |
17170 (custom-set-variables | |
17171 ;; custom-set-variables was added by Custom. | |
17172 ;; If you edit it by hand, you could mess it up, so be careful. | |
17173 ;; Your init file should contain only one such instance. | |
17174 ;; If there is more than one, they won't work right. | |
17175 '(text-mode-hook (quote (turn-on-auto-fill text-mode-hook-identify)))) | |
17176 @end group | |
17177 @end smallexample | |
17178 | |
17179 @noindent | |
17180 (The @code{text-mode-hook-identify} function tells | |
17181 @code{toggle-text-mode-auto-fill} which buffers are in Text mode. | |
17182 It comes on automatically.) | |
17183 | |
17184 The @code{custom-set-variables} function works somewhat differently | |
17185 than a @code{setq}. While I have never learned the differences, I | |
17186 modify the @code{custom-set-variables} expressions in my @file{.emacs} | |
17187 file by hand: I make the changes in what appears to me to be a | |
17188 reasonable manner and have not had any problems. Others prefer to use | |
17189 the Customization command and let Emacs do the work for them. | |
17190 | |
17191 Another @code{custom-set-@dots{}} function is @code{custom-set-faces}. | |
17192 This function sets the various font faces. Over time, I have set a | |
17193 considerable number of faces. Some of the time, I re-set them using | |
17194 @code{customize}; other times, I simply edit the | |
17195 @code{custom-set-faces} expression in my @file{.emacs} file itself. | |
17196 | |
17197 The second way to customize your @code{text-mode-hook} is to set it | |
17198 yourself in your @file{.emacs} file using code that has nothing to do | |
17199 with the @code{custom-set-@dots{}} functions. | |
17200 | |
17201 @need 800 | |
17202 When you do this, and later use @code{customize}, you will see a | |
17203 message that says | |
17204 | |
17205 @smallexample | |
17206 CHANGED outside Customize; operating on it here may be unreliable. | |
17207 @end smallexample | |
17208 | |
17209 @need 800 | |
17210 This message is only a warning. If you click on the button to | |
17211 | |
17212 @smallexample | |
17213 Save for Future Sessions | |
17214 @end smallexample | |
17215 | |
17216 @noindent | |
17217 Emacs will write a @code{custom-set-@dots{}} expression near the end | |
17218 of your @file{.emacs} file that will be evaluated after your | |
17219 hand-written expression. It will, therefore, overrule your | |
17220 hand-written expression. No harm will be done. When you do this, | |
17221 however, be careful to remember which expression is active; if you | |
17222 forget, you may confuse yourself. | |
17223 | |
17224 So long as you remember where the values are set, you will have no | |
17225 trouble. In any event, the values are always set in your | |
17226 initialization file, which is usually called @file{.emacs}. | |
17227 | |
17228 I myself use @code{customize} for hardly anything. Mostly, I write | |
17229 expressions myself. | |
17230 | |
17231 @findex defsubst | |
17232 @findex defconst | |
17233 Incidentally, to be more complete concerning defines: @code{defsubst} | |
17234 defines an inline function. The syntax is just like that of | |
17235 @code{defun}. @code{defconst} defines a symbol as a constant. The | |
17236 intent is that neither programs nor users should ever change a value | |
17237 set by @code{defconst}. (You can change it; the value set is a | |
17238 variable; but please do not.) | |
17239 | |
17240 @node Beginning a .emacs File, Text and Auto-fill, defcustom, Emacs Initialization | |
17241 @section Beginning a @file{.emacs} File | |
17242 @cindex @file{.emacs} file, beginning of | |
17243 | |
17244 When you start Emacs, it loads your @file{.emacs} file unless you tell | |
17245 it not to by specifying @samp{-q} on the command line. (The | |
17246 @code{emacs -q} command gives you a plain, out-of-the-box Emacs.) | |
17247 | |
17248 A @file{.emacs} file contains Lisp expressions. Often, these are no | |
17249 more than expressions to set values; sometimes they are function | |
17250 definitions. | |
17251 | |
17252 @xref{Init File, , The Init File @file{~/.emacs}, emacs, The GNU Emacs | |
17253 Manual}, for a short description of initialization files. | |
17254 | |
17255 This chapter goes over some of the same ground, but is a walk among | |
17256 extracts from a complete, long-used @file{.emacs} file---my own. | |
17257 | |
17258 The first part of the file consists of comments: reminders to myself. | |
17259 By now, of course, I remember these things, but when I started, I did | |
17260 not. | |
17261 | |
17262 @need 1200 | |
17263 @smallexample | |
17264 @group | |
17265 ;;;; Bob's .emacs file | |
17266 ; Robert J. Chassell | |
17267 ; 26 September 1985 | |
17268 @end group | |
17269 @end smallexample | |
17270 | |
17271 @noindent | |
17272 Look at that date! I started this file a long time ago. I have been | |
17273 adding to it ever since. | |
17274 | |
17275 @smallexample | |
17276 @group | |
17277 ; Each section in this file is introduced by a | |
17278 ; line beginning with four semicolons; and each | |
17279 ; entry is introduced by a line beginning with | |
17280 ; three semicolons. | |
17281 @end group | |
17282 @end smallexample | |
17283 | |
17284 @noindent | |
17285 This describes the usual conventions for comments in Emacs Lisp. | |
17286 Everything on a line that follows a semicolon is a comment. Two, | |
17287 three, and four semicolons are used as subsection and section markers. | |
17288 (@xref{Comments, ,, elisp, The GNU Emacs Lisp Reference Manual}, for | |
17289 more about comments.) | |
17290 | |
17291 @smallexample | |
17292 @group | |
17293 ;;;; The Help Key | |
17294 ; Control-h is the help key; | |
17295 ; after typing control-h, type a letter to | |
17296 ; indicate the subject about which you want help. | |
17297 ; For an explanation of the help facility, | |
17298 ; type control-h two times in a row. | |
17299 @end group | |
17300 @end smallexample | |
17301 | |
17302 @noindent | |
17303 Just remember: type @kbd{C-h} two times for help. | |
17304 | |
17305 @smallexample | |
17306 @group | |
17307 ; To find out about any mode, type control-h m | |
17308 ; while in that mode. For example, to find out | |
17309 ; about mail mode, enter mail mode and then type | |
17310 ; control-h m. | |
17311 @end group | |
17312 @end smallexample | |
17313 | |
17314 @noindent | |
17315 `Mode help', as I call this, is very helpful. Usually, it tells you | |
17316 all you need to know. | |
17317 | |
17318 Of course, you don't need to include comments like these in your | |
17319 @file{.emacs} file. I included them in mine because I kept forgetting | |
17320 about Mode help or the conventions for comments---but I was able to | |
17321 remember to look here to remind myself. | |
17322 | |
17323 @node Text and Auto-fill, Mail Aliases, Beginning a .emacs File, Emacs Initialization | |
17324 @section Text and Auto Fill Mode | |
17325 | |
17326 Now we come to the part that `turns on' Text mode and | |
17327 Auto Fill mode. | |
17328 | |
17329 @smallexample | |
17330 @group | |
17331 ;;; Text mode and Auto Fill mode | |
17332 ; The next two lines put Emacs into Text mode | |
17333 ; and Auto Fill mode, and are for writers who | |
17334 ; want to start writing prose rather than code. | |
17335 (setq default-major-mode 'text-mode) | |
17336 (add-hook 'text-mode-hook 'turn-on-auto-fill) | |
17337 @end group | |
17338 @end smallexample | |
17339 | |
17340 Here is the first part of this @file{.emacs} file that does something | |
17341 besides remind a forgetful human! | |
17342 | |
17343 The first of the two lines in parentheses tells Emacs to turn on Text | |
17344 mode when you find a file, @emph{unless} that file should go into some | |
17345 other mode, such as C mode. | |
17346 | |
17347 @cindex Per-buffer, local variables list | |
17348 @cindex Local variables list, per-buffer, | |
17349 @cindex Automatic mode selection | |
17350 @cindex Mode selection, automatic | |
17351 When Emacs reads a file, it looks at the extension to the file name, | |
17352 if any. (The extension is the part that comes after a @samp{.}.) If | |
17353 the file ends with a @samp{.c} or @samp{.h} extension then Emacs turns | |
17354 on C mode. Also, Emacs looks at first nonblank line of the file; if | |
17355 the line says @w{@samp{-*- C -*-}}, Emacs turns on C mode. Emacs | |
17356 possesses a list of extensions and specifications that it uses | |
17357 automatically. In addition, Emacs looks near the last page for a | |
17358 per-buffer, ``local variables list'', if any. | |
17359 | |
17360 @ifinfo | |
17361 @xref{Choosing Modes, , How Major Modes are Chosen, emacs, The GNU | |
17362 Emacs Manual}. | |
17363 | |
17364 @xref{File Variables, , Local Variables in Files, emacs, The GNU Emacs | |
17365 Manual}. | |
17366 @end ifinfo | |
17367 @iftex | |
17368 See sections ``How Major Modes are Chosen'' and ``Local Variables in | |
17369 Files'' in @cite{The GNU Emacs Manual}. | |
17370 @end iftex | |
17371 | |
17372 Now, back to the @file{.emacs} file. | |
17373 | |
17374 @need 800 | |
17375 Here is the line again; how does it work? | |
17376 | |
17377 @cindex Text Mode turned on | |
17378 @smallexample | |
17379 (setq default-major-mode 'text-mode) | |
17380 @end smallexample | |
17381 | |
17382 @noindent | |
17383 This line is a short, but complete Emacs Lisp expression. | |
17384 | |
17385 We are already familiar with @code{setq}. It sets the following variable, | |
17386 @code{default-major-mode}, to the subsequent value, which is | |
17387 @code{text-mode}. The single quote mark before @code{text-mode} tells | |
17388 Emacs to deal directly with the @code{text-mode} variable, not with | |
17389 whatever it might stand for. @xref{set & setq, , Setting the Value of | |
17390 a Variable}, for a reminder of how @code{setq} works. The main point | |
17391 is that there is no difference between the procedure you use to set | |
17392 a value in your @file{.emacs} file and the procedure you use anywhere | |
17393 else in Emacs. | |
17394 | |
17395 @need 800 | |
17396 Here is the next line: | |
17397 | |
17398 @cindex Auto Fill mode turned on | |
17399 @findex add-hook | |
17400 @smallexample | |
17401 (add-hook 'text-mode-hook 'turn-on-auto-fill) | |
17402 @end smallexample | |
17403 | |
17404 @noindent | |
17405 In this line, the @code{add-hook} command adds | |
17406 @code{turn-on-auto-fill} to the variable. | |
17407 | |
17408 @code{turn-on-auto-fill} is the name of a program, that, you guessed | |
17409 it!, turns on Auto Fill mode. | |
17410 | |
17411 Every time Emacs turns on Text mode, Emacs runs the commands `hooked' | |
17412 onto Text mode. So every time Emacs turns on Text mode, Emacs also | |
17413 turns on Auto Fill mode. | |
17414 | |
17415 In brief, the first line causes Emacs to enter Text mode when you edit a | |
17416 file, unless the file name extension, a first non-blank line, or local | |
17417 variables to tell Emacs otherwise. | |
17418 | |
17419 Text mode among other actions, sets the syntax table to work | |
17420 conveniently for writers. In Text mode, Emacs considers an apostrophe | |
17421 as part of a word like a letter; but Emacs does not consider a period | |
17422 or a space as part of a word. Thus, @kbd{M-f} moves you over | |
17423 @samp{it's}. On the other hand, in C mode, @kbd{M-f} stops just after | |
17424 the @samp{t} of @samp{it's}. | |
17425 | |
17426 The second line causes Emacs to turn on Auto Fill mode when it turns | |
17427 on Text mode. In Auto Fill mode, Emacs automatically breaks a line | |
17428 that is too wide and brings the excessively wide part of the line down | |
17429 to the next line. Emacs breaks lines between words, not within them. | |
17430 | |
17431 When Auto Fill mode is turned off, lines continue to the right as you | |
17432 type them. Depending on how you set the value of | |
17433 @code{truncate-lines}, the words you type either disappear off the | |
17434 right side of the screen, or else are shown, in a rather ugly and | |
17435 unreadable manner, as a continuation line on the screen. | |
17436 | |
17437 @need 1250 | |
17438 In addition, in this part of my @file{.emacs} file, I tell the Emacs | |
17439 fill commands to insert two spaces after a colon: | |
17440 | |
17441 @smallexample | |
17442 (setq colon-double-space t) | |
17443 @end smallexample | |
17444 | |
17445 @node Mail Aliases, Indent Tabs Mode, Text and Auto-fill, Emacs Initialization | |
17446 @section Mail Aliases | |
17447 | |
17448 Here is a @code{setq} that `turns on' mail aliases, along with more | |
17449 reminders. | |
17450 | |
17451 @smallexample | |
17452 @group | |
17453 ;;; Mail mode | |
17454 ; To enter mail mode, type `C-x m' | |
17455 ; To enter RMAIL (for reading mail), | |
17456 ; type `M-x rmail' | |
17457 (setq mail-aliases t) | |
17458 @end group | |
17459 @end smallexample | |
17460 | |
17461 @cindex Mail aliases | |
17462 @noindent | |
17463 This @code{setq} command sets the value of the variable | |
17464 @code{mail-aliases} to @code{t}. Since @code{t} means true, the line | |
17465 says, in effect, ``Yes, use mail aliases.'' | |
17466 | |
17467 Mail aliases are convenient short names for long email addresses or | |
17468 for lists of email addresses. The file where you keep your `aliases' | |
17469 is @file{~/.mailrc}. You write an alias like this: | |
17470 | |
17471 @smallexample | |
17472 alias geo george@@foobar.wiz.edu | |
17473 @end smallexample | |
17474 | |
17475 @noindent | |
17476 When you write a message to George, address it to @samp{geo}; the | |
17477 mailer will automatically expand @samp{geo} to the full address. | |
17478 | |
17479 @node Indent Tabs Mode, Keybindings, Mail Aliases, Emacs Initialization | |
17480 @section Indent Tabs Mode | |
17481 @cindex Tabs, preventing | |
17482 @findex indent-tabs-mode | |
17483 | |
17484 By default, Emacs inserts tabs in place of multiple spaces when it | |
17485 formats a region. (For example, you might indent many lines of text | |
17486 all at once with the @code{indent-region} command.) Tabs look fine on | |
17487 a terminal or with ordinary printing, but they produce badly indented | |
17488 output when you use @TeX{} or Texinfo since @TeX{} ignores tabs. | |
17489 | |
17490 @need 1250 | |
17491 The following turns off Indent Tabs mode: | |
17492 | |
17493 @smallexample | |
17494 @group | |
17495 ;;; Prevent Extraneous Tabs | |
17496 (setq-default indent-tabs-mode nil) | |
17497 @end group | |
17498 @end smallexample | |
17499 | |
17500 Note that this line uses @code{setq-default} rather than the | |
17501 @code{setq} command that we have seen before. The @code{setq-default} | |
17502 command sets values only in buffers that do not have their own local | |
17503 values for the variable. | |
17504 | |
17505 @ifinfo | |
17506 @xref{Just Spaces, , Tabs vs. Spaces, emacs, The GNU Emacs Manual}. | |
17507 | |
17508 @xref{File Variables, , Local Variables in Files, emacs, The GNU Emacs | |
17509 Manual}. | |
17510 @end ifinfo | |
17511 @iftex | |
17512 See sections ``Tabs vs.@: Spaces'' and ``Local Variables in | |
17513 Files'' in @cite{The GNU Emacs Manual}. | |
17514 @end iftex | |
17515 | |
17516 @need 1700 | |
17517 @node Keybindings, Keymaps, Indent Tabs Mode, Emacs Initialization | |
17518 @section Some Keybindings | |
17519 | |
17520 Now for some personal keybindings: | |
17521 | |
17522 @smallexample | |
17523 @group | |
17524 ;;; Compare windows | |
17525 (global-set-key "\C-cw" 'compare-windows) | |
17526 @end group | |
17527 @end smallexample | |
17528 | |
17529 @findex compare-windows | |
17530 @code{compare-windows} is a nifty command that compares the text in | |
17531 your current window with text in the next window. It makes the | |
17532 comparison by starting at point in each window, moving over text in | |
17533 each window as far as they match. I use this command all the time. | |
17534 | |
17535 This also shows how to set a key globally, for all modes. | |
17536 | |
17537 @cindex Setting a key globally | |
17538 @cindex Global set key | |
17539 @cindex Key setting globally | |
17540 @findex global-set-key | |
17541 The command is @code{global-set-key}. It is followed by the | |
17542 keybinding. In a @file{.emacs} file, the keybinding is written as | |
17543 shown: @code{\C-c} stands for `control-c', which means `press the | |
17544 control key and the @key{c} key at the same time'. The @code{w} means | |
17545 `press the @key{w} key'. The keybinding is surrounded by double | |
17546 quotation marks. In documentation, you would write this as | |
17547 @w{@kbd{C-c w}}. (If you were binding a @key{META} key, such as | |
17548 @kbd{M-c}, rather than a @key{CTRL} key, you would write | |
17549 @w{@code{\M-c}} in your @file{.emacs} file. @xref{Init Rebinding, , | |
17550 Rebinding Keys in Your Init File, emacs, The GNU Emacs Manual}, for | |
17551 details.) | |
17552 | |
17553 The command invoked by the keys is @code{compare-windows}. Note that | |
17554 @code{compare-windows} is preceded by a single quote; otherwise, Emacs | |
17555 would first try to evaluate the symbol to determine its value. | |
17556 | |
17557 These three things, the double quotation marks, the backslash before | |
17558 the @samp{C}, and the single quote mark are necessary parts of | |
17559 keybinding that I tend to forget. Fortunately, I have come to | |
17560 remember that I should look at my existing @file{.emacs} file, and | |
17561 adapt what is there. | |
17562 | |
17563 As for the keybinding itself: @kbd{C-c w}. This combines the prefix | |
17564 key, @kbd{C-c}, with a single character, in this case, @kbd{w}. This | |
17565 set of keys, @kbd{C-c} followed by a single character, is strictly | |
17566 reserved for individuals' own use. (I call these `own' keys, since | |
17567 these are for my own use.) You should always be able to create such a | |
17568 keybinding for your own use without stomping on someone else's | |
17569 keybinding. If you ever write an extension to Emacs, please avoid | |
17570 taking any of these keys for public use. Create a key like @kbd{C-c | |
17571 C-w} instead. Otherwise, we will run out of `own' keys. | |
17572 | |
17573 @need 1250 | |
17574 Here is another keybinding, with a comment: | |
17575 | |
17576 @smallexample | |
17577 @group | |
17578 ;;; Keybinding for `occur' | |
17579 ; I use occur a lot, so let's bind it to a key: | |
17580 (global-set-key "\C-co" 'occur) | |
17581 @end group | |
17582 @end smallexample | |
17583 | |
17584 @findex occur | |
17585 The @code{occur} command shows all the lines in the current buffer | |
17586 that contain a match for a regular expression. Matching lines are | |
17587 shown in a buffer called @file{*Occur*}. That buffer serves as a menu | |
17588 to jump to occurrences. | |
17589 | |
17590 @findex global-unset-key | |
17591 @cindex Unbinding key | |
17592 @cindex Key unbinding | |
17593 @need 1250 | |
17594 Here is how to unbind a key, so it does not | |
17595 work: | |
17596 | |
17597 @smallexample | |
17598 @group | |
17599 ;;; Unbind `C-x f' | |
17600 (global-unset-key "\C-xf") | |
17601 @end group | |
17602 @end smallexample | |
17603 | |
17604 There is a reason for this unbinding: I found I inadvertently typed | |
17605 @w{@kbd{C-x f}} when I meant to type @kbd{C-x C-f}. Rather than find a | |
17606 file, as I intended, I accidentally set the width for filled text, | |
17607 almost always to a width I did not want. Since I hardly ever reset my | |
17608 default width, I simply unbound the key. | |
17609 | |
17610 @findex list-buffers, @r{rebound} | |
17611 @findex buffer-menu, @r{bound to key} | |
17612 @need 1250 | |
17613 The following rebinds an existing key: | |
17614 | |
17615 @smallexample | |
17616 @group | |
17617 ;;; Rebind `C-x C-b' for `buffer-menu' | |
17618 (global-set-key "\C-x\C-b" 'buffer-menu) | |
17619 @end group | |
17620 @end smallexample | |
17621 | |
17622 By default, @kbd{C-x C-b} runs the | |
17623 @code{list-buffers} command. This command lists | |
17624 your buffers in @emph{another} window. Since I | |
17625 almost always want to do something in that | |
17626 window, I prefer the @code{buffer-menu} | |
17627 command, which not only lists the buffers, | |
17628 but moves point into that window. | |
17629 | |
17630 @node Keymaps, Loading Files, Keybindings, Emacs Initialization | |
17631 @section Keymaps | |
17632 @cindex Keymaps | |
17633 @cindex Rebinding keys | |
17634 | |
17635 Emacs uses @dfn{keymaps} to record which keys call which commands. | |
17636 When you use @code{global-set-key} to set the keybinding for a single | |
17637 command in all parts of Emacs, you are specifying the keybinding in | |
17638 @code{current-global-map}. | |
17639 | |
17640 Specific modes, such as C mode or Text mode, have their own keymaps; | |
17641 the mode-specific keymaps override the global map that is shared by | |
17642 all buffers. | |
17643 | |
17644 The @code{global-set-key} function binds, or rebinds, the global | |
17645 keymap. For example, the following binds the key @kbd{C-x C-b} to the | |
17646 function @code{buffer-menu}: | |
17647 | |
17648 @smallexample | |
17649 (global-set-key "\C-x\C-b" 'buffer-menu) | |
17650 @end smallexample | |
17651 | |
17652 Mode-specific keymaps are bound using the @code{define-key} function, | |
17653 which takes a specific keymap as an argument, as well as the key and | |
17654 the command. For example, my @file{.emacs} file contains the | |
17655 following expression to bind the @code{texinfo-insert-@@group} command | |
17656 to @kbd{C-c C-c g}: | |
17657 | |
17658 @smallexample | |
17659 @group | |
17660 (define-key texinfo-mode-map "\C-c\C-cg" 'texinfo-insert-@@group) | |
17661 @end group | |
17662 @end smallexample | |
17663 | |
17664 @noindent | |
17665 The @code{texinfo-insert-@@group} function itself is a little extension | |
17666 to Texinfo mode that inserts @samp{@@group} into a Texinfo file. I | |
17667 use this command all the time and prefer to type the three strokes | |
17668 @kbd{C-c C-c g} rather than the six strokes @kbd{@@ g r o u p}. | |
17669 (@samp{@@group} and its matching @samp{@@end group} are commands that | |
17670 keep all enclosed text together on one page; many multi-line examples | |
17671 in this book are surrounded by @samp{@@group @dots{} @@end group}.) | |
17672 | |
17673 @need 1250 | |
17674 Here is the @code{texinfo-insert-@@group} function definition: | |
17675 | |
17676 @smallexample | |
17677 @group | |
17678 (defun texinfo-insert-@@group () | |
17679 "Insert the string @@group in a Texinfo buffer." | |
17680 (interactive) | |
17681 (beginning-of-line) | |
17682 (insert "@@group\n")) | |
17683 @end group | |
17684 @end smallexample | |
17685 | |
17686 (Of course, I could have used Abbrev mode to save typing, rather than | |
17687 write a function to insert a word; but I prefer key strokes consistent | |
17688 with other Texinfo mode key bindings.) | |
17689 | |
17690 You will see numerous @code{define-key} expressions in | |
17691 @file{loaddefs.el} as well as in the various mode libraries, such as | |
17692 @file{cc-mode.el} and @file{lisp-mode.el}. | |
17693 | |
17694 @xref{Key Bindings, , Customizing Key Bindings, emacs, The GNU Emacs | |
17695 Manual}, and @ref{Keymaps, , Keymaps, elisp, The GNU Emacs Lisp | |
17696 Reference Manual}, for more information about keymaps. | |
17697 | |
17698 @node Loading Files, Autoload, Keymaps, Emacs Initialization | |
17699 @section Loading Files | |
17700 @cindex Loading files | |
17701 @c findex load | |
17702 | |
17703 Many people in the GNU Emacs community have written extensions to | |
17704 Emacs. As time goes by, these extensions are often included in new | |
17705 releases. For example, the Calendar and Diary packages are now part | |
17706 of the standard GNU Emacs, as is Calc. | |
17707 | |
17708 You can use a @code{load} command to evaluate a complete file and | |
17709 thereby install all the functions and variables in the file into Emacs. | |
17710 For example: | |
17711 | |
17712 @c (auto-compression-mode t) | |
17713 | |
17714 @smallexample | |
17715 (load "~/emacs/slowsplit") | |
17716 @end smallexample | |
17717 | |
17718 This evaluates, i.e.@: loads, the @file{slowsplit.el} file or if it | |
17719 exists, the faster, byte compiled @file{slowsplit.elc} file from the | |
17720 @file{emacs} sub-directory of your home directory. The file contains | |
17721 the function @code{split-window-quietly}, which John Robinson wrote in | |
17722 1989. | |
17723 | |
17724 The @code{split-window-quietly} function splits a window with the | |
17725 minimum of redisplay. I installed it in 1989 because it worked well | |
17726 with the slow 1200 baud terminals I was then using. Nowadays, I only | |
17727 occasionally come across such a slow connection, but I continue to use | |
17728 the function because I like the way it leaves the bottom half of a | |
17729 buffer in the lower of the new windows and the top half in the upper | |
17730 window. | |
17731 | |
17732 @need 1250 | |
17733 To replace the key binding for the default | |
17734 @code{split-window-vertically}, you must also unset that key and bind | |
17735 the keys to @code{split-window-quietly}, like this: | |
17736 | |
17737 @smallexample | |
17738 @group | |
17739 (global-unset-key "\C-x2") | |
17740 (global-set-key "\C-x2" 'split-window-quietly) | |
17741 @end group | |
17742 @end smallexample | |
17743 | |
17744 @vindex load-path | |
17745 If you load many extensions, as I do, then instead of specifying the | |
17746 exact location of the extension file, as shown above, you can specify | |
17747 that directory as part of Emacs' @code{load-path}. Then, when Emacs | |
17748 loads a file, it will search that directory as well as its default | |
17749 list of directories. (The default list is specified in @file{paths.h} | |
17750 when Emacs is built.) | |
17751 | |
17752 @need 1250 | |
17753 The following command adds your @file{~/emacs} directory to the | |
17754 existing load path: | |
17755 | |
17756 @smallexample | |
17757 @group | |
17758 ;;; Emacs Load Path | |
17759 (setq load-path (cons "~/emacs" load-path)) | |
17760 @end group | |
17761 @end smallexample | |
17762 | |
17763 Incidentally, @code{load-library} is an interactive interface to the | |
17764 @code{load} function. The complete function looks like this: | |
17765 | |
17766 @findex load-library | |
17767 @smallexample | |
17768 @group | |
17769 (defun load-library (library) | |
17770 "Load the library named LIBRARY. | |
17771 This is an interface to the function `load'." | |
17772 (interactive | |
17773 (list (completing-read "Load library: " | |
17774 'locate-file-completion | |
17775 (cons load-path (get-load-suffixes))))) | |
17776 (load library)) | |
17777 @end group | |
17778 @end smallexample | |
17779 | |
17780 The name of the function, @code{load-library}, comes from the use of | |
17781 `library' as a conventional synonym for `file'. The source for the | |
17782 @code{load-library} command is in the @file{files.el} library. | |
17783 | |
17784 Another interactive command that does a slightly different job is | |
17785 @code{load-file}. @xref{Lisp Libraries, , Libraries of Lisp Code for | |
17786 Emacs, emacs, The GNU Emacs Manual}, for information on the | |
17787 distinction between @code{load-library} and this command. | |
17788 | |
17789 @node Autoload, Simple Extension, Loading Files, Emacs Initialization | |
17790 @section Autoloading | |
17791 @findex autoload | |
17792 | |
17793 Instead of installing a function by loading the file that contains it, | |
17794 or by evaluating the function definition, you can make the function | |
17795 available but not actually install it until it is first called. This | |
17796 is called @dfn{autoloading}. | |
17797 | |
17798 When you execute an autoloaded function, Emacs automatically evaluates | |
17799 the file that contains the definition, and then calls the function. | |
17800 | |
17801 Emacs starts quicker with autoloaded functions, since their libraries | |
17802 are not loaded right away; but you need to wait a moment when you | |
17803 first use such a function, while its containing file is evaluated. | |
17804 | |
17805 Rarely used functions are frequently autoloaded. The | |
17806 @file{loaddefs.el} library contains hundreds of autoloaded functions, | |
17807 from @code{bookmark-set} to @code{wordstar-mode}. Of course, you may | |
17808 come to use a `rare' function frequently. When you do, you should | |
17809 load that function's file with a @code{load} expression in your | |
17810 @file{.emacs} file. | |
17811 | |
17812 In my @file{.emacs} file, I load 14 libraries that contain functions | |
17813 that would otherwise be autoloaded. (Actually, it would have been | |
17814 better to include these files in my `dumped' Emacs, but I forgot. | |
17815 @xref{Building Emacs, , Building Emacs, elisp, The GNU Emacs Lisp | |
17816 Reference Manual}, and the @file{INSTALL} file for more about | |
17817 dumping.) | |
17818 | |
17819 You may also want to include autoloaded expressions in your @file{.emacs} | |
17820 file. @code{autoload} is a built-in function that takes up to five | |
17821 arguments, the final three of which are optional. The first argument | |
17822 is the name of the function to be autoloaded; the second is the name | |
17823 of the file to be loaded. The third argument is documentation for the | |
17824 function, and the fourth tells whether the function can be called | |
17825 interactively. The fifth argument tells what type of | |
17826 object---@code{autoload} can handle a keymap or macro as well as a | |
17827 function (the default is a function). | |
17828 | |
17829 @need 800 | |
17830 Here is a typical example: | |
17831 | |
17832 @smallexample | |
17833 @group | |
17834 (autoload 'html-helper-mode | |
17835 "html-helper-mode" "Edit HTML documents" t) | |
17836 @end group | |
17837 @end smallexample | |
17838 | |
17839 @noindent | |
17840 (@code{html-helper-mode} is an older alternative to @code{html-mode}, | |
17841 which is a standard part of the distribution.) | |
17842 | |
17843 @noindent | |
17844 This expression autoloads the @code{html-helper-mode} function. It | |
17845 takes it from the @file{html-helper-mode.el} file (or from the byte | |
17846 compiled file @file{html-helper-mode.elc}, if it exists.) The file | |
17847 must be located in a directory specified by @code{load-path}. The | |
17848 documentation says that this is a mode to help you edit documents | |
17849 written in the HyperText Markup Language. You can call this mode | |
17850 interactively by typing @kbd{M-x html-helper-mode}. (You need to | |
17851 duplicate the function's regular documentation in the autoload | |
17852 expression because the regular function is not yet loaded, so its | |
17853 documentation is not available.) | |
17854 | |
17855 @xref{Autoload, , Autoload, elisp, The GNU Emacs Lisp Reference | |
17856 Manual}, for more information. | |
17857 | |
17858 @node Simple Extension, X11 Colors, Autoload, Emacs Initialization | |
17859 @section A Simple Extension: @code{line-to-top-of-window} | |
17860 @findex line-to-top-of-window | |
17861 @cindex Simple extension in @file{.emacs} file | |
17862 | |
17863 Here is a simple extension to Emacs that moves the line point is on to | |
17864 the top of the window. I use this all the time, to make text easier | |
17865 to read. | |
17866 | |
17867 You can put the following code into a separate file and then load it | |
17868 from your @file{.emacs} file, or you can include it within your | |
17869 @file{.emacs} file. | |
17870 | |
17871 @need 1250 | |
17872 Here is the definition: | |
17873 | |
17874 @smallexample | |
17875 @group | |
17876 ;;; Line to top of window; | |
17877 ;;; replace three keystroke sequence C-u 0 C-l | |
17878 (defun line-to-top-of-window () | |
17879 "Move the line point is on to top of window." | |
17880 (interactive) | |
17881 (recenter 0)) | |
17882 @end group | |
17883 @end smallexample | |
17884 | |
17885 @need 1250 | |
17886 Now for the keybinding. | |
17887 | |
17888 Nowadays, function keys as well as mouse button events and | |
17889 non-@sc{ascii} characters are written within square brackets, without | |
17890 quotation marks. (In Emacs version 18 and before, you had to write | |
17891 different function key bindings for each different make of terminal.) | |
17892 | |
17893 I bind @code{line-to-top-of-window} to my @key{F6} function key like | |
17894 this: | |
17895 | |
17896 @smallexample | |
17897 (global-set-key [f6] 'line-to-top-of-window) | |
17898 @end smallexample | |
17899 | |
17900 For more information, see @ref{Init Rebinding, , Rebinding Keys in | |
17901 Your Init File, emacs, The GNU Emacs Manual}. | |
17902 | |
17903 @cindex Conditional 'twixt two versions of Emacs | |
17904 @cindex Version of Emacs, choosing | |
17905 @cindex Emacs version, choosing | |
17906 If you run two versions of GNU Emacs, such as versions 21 and 22, and | |
17907 use one @file{.emacs} file, you can select which code to evaluate with | |
17908 the following conditional: | |
17909 | |
17910 @smallexample | |
17911 @group | |
17912 (cond | |
17913 (= 21 emacs-major-version) | |
17914 ;; evaluate version 21 code | |
17915 ( @dots{} )) | |
17916 (= 22 emacs-major-version) | |
17917 ;; evaluate version 22 code | |
17918 ( @dots{} ))) | |
17919 @end group | |
17920 @end smallexample | |
17921 | |
17922 For example, in contrast to version 20, more recent versions blink | |
17923 their cursors by default. I hate such blinking, as well as other | |
17924 features, so I placed the following in my @file{.emacs} | |
17925 file@footnote{When I start instances of Emacs that do not load my | |
17926 @file{.emacs} file or any site file, I also turn off blinking: | |
17927 | |
17928 @smallexample | |
17929 emacs -q --no-site-file -eval '(blink-cursor-mode nil)' | |
17930 | |
17931 @exdent Or nowadays, using an even more sophisticated set of options, | |
17932 | |
17933 emacs -Q - D | |
17934 @end smallexample | |
17935 }: | |
17936 | |
17937 @smallexample | |
17938 @group | |
17939 (when (or (= 21 emacs-major-version) | |
17940 (= 22 emacs-major-version)) | |
17941 (blink-cursor-mode 0) | |
17942 ;; Insert newline when you press `C-n' (next-line) | |
17943 ;; at the end of the buffer | |
17944 (setq next-line-add-newlines t) | |
17945 @end group | |
17946 @group | |
17947 ;; Turn on image viewing | |
17948 (auto-image-file-mode t) | |
17949 @end group | |
17950 @group | |
17951 ;; Turn on menu bar (this bar has text) | |
17952 ;; (Use numeric argument to turn on) | |
17953 (menu-bar-mode 1) | |
17954 @end group | |
17955 @group | |
17956 ;; Turn off tool bar (this bar has icons) | |
17957 ;; (Use numeric argument to turn on) | |
17958 (tool-bar-mode nil) | |
17959 @end group | |
17960 @group | |
17961 ;; Turn off tooltip mode for tool bar | |
17962 ;; (This mode causes icon explanations to pop up) | |
17963 ;; (Use numeric argument to turn on) | |
17964 (tooltip-mode nil) | |
17965 ;; If tooltips turned on, make tips appear promptly | |
17966 (setq tooltip-delay 0.1) ; default is 0.7 second | |
17967 ) | |
17968 @end group | |
17969 @end smallexample | |
17970 | |
17971 @need 1250 | |
17972 Alternatively, since @code{blink-cursor-mode} has existed since Emacs | |
17973 version 21 and is likely to continue, you could write | |
17974 | |
17975 @smallexample | |
17976 @group | |
17977 (when (>= emacs-major-version 21) | |
17978 (blink-cursor-mode 0) | |
17979 @end group | |
17980 @end smallexample | |
17981 | |
17982 @noindent | |
17983 and add other expressions, too. | |
17984 | |
17985 | |
17986 @node X11 Colors, Miscellaneous, Simple Extension, Emacs Initialization | |
17987 @section X11 Colors | |
17988 | |
17989 You can specify colors when you use Emacs with the MIT X Windowing | |
17990 system. | |
17991 | |
17992 I dislike the default colors and specify my own. | |
17993 | |
17994 @need 1250 | |
17995 Here are the expressions in my @file{.emacs} | |
17996 file that set values: | |
17997 | |
17998 @smallexample | |
17999 @group | |
18000 ;; Set cursor color | |
18001 (set-cursor-color "white") | |
18002 | |
18003 ;; Set mouse color | |
18004 (set-mouse-color "white") | |
18005 | |
18006 ;; Set foreground and background | |
18007 (set-foreground-color "white") | |
18008 (set-background-color "darkblue") | |
18009 @end group | |
18010 | |
18011 @group | |
18012 ;;; Set highlighting colors for isearch and drag | |
18013 (set-face-foreground 'highlight "white") | |
18014 (set-face-background 'highlight "blue") | |
18015 @end group | |
18016 | |
18017 @group | |
18018 (set-face-foreground 'region "cyan") | |
18019 (set-face-background 'region "blue") | |
18020 @end group | |
18021 | |
18022 @group | |
18023 (set-face-foreground 'secondary-selection "skyblue") | |
18024 (set-face-background 'secondary-selection "darkblue") | |
18025 @end group | |
18026 | |
18027 @group | |
18028 ;; Set calendar highlighting colors | |
18029 (setq calendar-load-hook | |
18030 '(lambda () | |
18031 (set-face-foreground 'diary-face "skyblue") | |
18032 (set-face-background 'holiday-face "slate blue") | |
18033 (set-face-foreground 'holiday-face "white"))) | |
18034 @end group | |
18035 @end smallexample | |
18036 | |
18037 The various shades of blue soothe my eye and prevent me from seeing | |
18038 the screen flicker. | |
18039 | |
18040 Alternatively, I could have set my specifications in various X | |
18041 initialization files. For example, I could set the foreground, | |
18042 background, cursor, and pointer (i.e., mouse) colors in my | |
18043 @file{~/.Xresources} file like this: | |
18044 | |
18045 @smallexample | |
18046 @group | |
18047 Emacs*foreground: white | |
18048 Emacs*background: darkblue | |
18049 Emacs*cursorColor: white | |
18050 Emacs*pointerColor: white | |
18051 @end group | |
18052 @end smallexample | |
18053 | |
18054 In any event, since it is not part of Emacs, I set the root color of | |
18055 my X window in my @file{~/.xinitrc} file, like this@footnote{I also | |
18056 run more modern window managers, such as Enlightenment, Gnome, or KDE; | |
18057 in those cases, I often specify an image rather than a plain color.}: | |
18058 | |
18059 @smallexample | |
18060 xsetroot -solid Navy -fg white & | |
18061 @end smallexample | |
18062 | |
18063 @need 1700 | |
18064 @node Miscellaneous, Mode Line, X11 Colors, Emacs Initialization | |
18065 @section Miscellaneous Settings for a @file{.emacs} File | |
18066 | |
18067 @need 1250 | |
18068 Here are a few miscellaneous settings: | |
18069 @sp 1 | |
18070 | |
18071 @itemize @minus | |
18072 @item | |
18073 Set the shape and color of the mouse cursor: | |
18074 | |
18075 @smallexample | |
18076 @group | |
18077 ; Cursor shapes are defined in | |
18078 ; `/usr/include/X11/cursorfont.h'; | |
18079 ; for example, the `target' cursor is number 128; | |
18080 ; the `top_left_arrow' cursor is number 132. | |
18081 @end group | |
18082 | |
18083 @group | |
18084 (let ((mpointer (x-get-resource "*mpointer" | |
18085 "*emacs*mpointer"))) | |
18086 ;; If you have not set your mouse pointer | |
18087 ;; then set it, otherwise leave as is: | |
18088 (if (eq mpointer nil) | |
18089 (setq mpointer "132")) ; top_left_arrow | |
18090 @end group | |
18091 @group | |
18092 (setq x-pointer-shape (string-to-int mpointer)) | |
18093 (set-mouse-color "white")) | |
18094 @end group | |
18095 @end smallexample | |
18096 | |
18097 @item | |
18098 Or you can set the values of a variety of features in an alist, like | |
18099 this: | |
18100 | |
18101 @smallexample | |
18102 @group | |
18103 (setq-default | |
18104 default-frame-alist | |
18105 '((cursor-color . "white") | |
18106 (mouse-color . "white") | |
18107 (foreground-color . "white") | |
18108 (background-color . "DodgerBlue4") | |
18109 ;; (cursor-type . bar) | |
18110 (cursor-type . box) | |
18111 @end group | |
18112 @group | |
18113 (tool-bar-lines . 0) | |
18114 (menu-bar-lines . 1) | |
18115 (width . 80) | |
18116 (height . 58) | |
18117 (font . | |
18118 "-Misc-Fixed-Medium-R-Normal--20-200-75-75-C-100-ISO8859-1") | |
18119 )) | |
18120 @end group | |
18121 @end smallexample | |
18122 | |
18123 @item | |
18124 Convert @kbd{@key{CTRL}-h} into @key{DEL} and @key{DEL} | |
18125 into @kbd{@key{CTRL}-h}.@* | |
18126 (Some older keyboards needed this, although I have not seen the | |
18127 problem recently.) | |
18128 | |
18129 @smallexample | |
18130 @group | |
18131 ;; Translate `C-h' to <DEL>. | |
18132 ; (keyboard-translate ?\C-h ?\C-?) | |
18133 | |
18134 ;; Translate <DEL> to `C-h'. | |
18135 (keyboard-translate ?\C-? ?\C-h) | |
18136 @end group | |
18137 @end smallexample | |
18138 | |
18139 @item Turn off a blinking cursor! | |
18140 | |
18141 @smallexample | |
18142 @group | |
18143 (if (fboundp 'blink-cursor-mode) | |
18144 (blink-cursor-mode -1)) | |
18145 @end group | |
18146 @end smallexample | |
18147 | |
18148 @noindent | |
18149 or start GNU Emacs with the command @code{emacs -nbc}. | |
18150 | |
18151 @need 1250 | |
18152 @item When using `grep'@* | |
18153 @samp{-i}@w{ } Ignore case distinctions@* | |
18154 @samp{-n}@w{ } Prefix each line of output with line number@* | |
18155 @samp{-H}@w{ } Print the filename for each match.@* | |
18156 @samp{-e}@w{ } Protect patterns beginning with a hyphen character, @samp{-} | |
18157 | |
18158 @smallexample | |
18159 (setq grep-command "grep -i -nH -e ") | |
18160 @end smallexample | |
18161 | |
18162 @ignore | |
18163 @c Evidently, no longer needed in GNU Emacs 22 | |
18164 | |
18165 item Automatically uncompress compressed files when visiting them | |
18166 | |
18167 smallexample | |
18168 (load "uncompress") | |
18169 end smallexample | |
18170 | |
18171 @end ignore | |
18172 | |
18173 @item Find an existing buffer, even if it has a different name@* | |
18174 This avoids problems with symbolic links. | |
18175 | |
18176 @smallexample | |
18177 (setq find-file-existing-other-name t) | |
18178 @end smallexample | |
18179 | |
18180 @item Set your language environment and default input method | |
18181 | |
18182 @smallexample | |
18183 @group | |
18184 (set-language-environment "latin-1") | |
18185 ;; Remember you can enable or disable multilingual text input | |
18186 ;; with the @code{toggle-input-method'} (@kbd{C-\}) command | |
18187 (setq default-input-method "latin-1-prefix") | |
18188 @end group | |
18189 @end smallexample | |
18190 | |
18191 If you want to write with Chinese `GB' characters, set this instead: | |
18192 | |
18193 @smallexample | |
18194 @group | |
18195 (set-language-environment "Chinese-GB") | |
18196 (setq default-input-method "chinese-tonepy") | |
18197 @end group | |
18198 @end smallexample | |
18199 @end itemize | |
18200 | |
18201 @subsubheading Fixing Unpleasant Key Bindings | |
18202 @cindex Key bindings, fixing | |
18203 @cindex Bindings, key, fixing unpleasant | |
18204 | |
18205 Some systems bind keys unpleasantly. Sometimes, for example, the | |
18206 @key{CTRL} key appears in an awkward spot rather than at the far left | |
18207 of the home row. | |
18208 | |
18209 Usually, when people fix these sorts of keybindings, they do not | |
18210 change their @file{~/.emacs} file. Instead, they bind the proper keys | |
18211 on their consoles with the @code{loadkeys} or @code{install-keymap} | |
18212 commands in their boot script and then include @code{xmodmap} commands | |
18213 in their @file{.xinitrc} or @file{.Xsession} file for X Windows. | |
18214 | |
18215 @need 1250 | |
18216 @noindent | |
18217 For a boot script: | |
18218 | |
18219 @smallexample | |
18220 @group | |
18221 loadkeys /usr/share/keymaps/i386/qwerty/emacs2.kmap.gz | |
18222 @exdent or | |
18223 install-keymap emacs2 | |
18224 @end group | |
18225 @end smallexample | |
18226 | |
18227 @need 1250 | |
18228 @noindent | |
18229 For a @file{.xinitrc} or @file{.Xsession} file when the @key{Caps | |
18230 Lock} key is at the far left of the home row: | |
18231 | |
18232 @smallexample | |
18233 @group | |
18234 # Bind the key labeled `Caps Lock' to `Control' | |
18235 # (Such a broken user interface suggests that keyboard manufacturers | |
18236 # think that computers are typewriters from 1885.) | |
18237 | |
18238 xmodmap -e "clear Lock" | |
18239 xmodmap -e "add Control = Caps_Lock" | |
18240 @end group | |
18241 @end smallexample | |
18242 | |
18243 @need 1250 | |
18244 @noindent | |
18245 In a @file{.xinitrc} or @file{.Xsession} file, to convert an @key{ALT} | |
18246 key to a @key{META} key: | |
18247 | |
18248 @smallexample | |
18249 @group | |
18250 # Some ill designed keyboards have a key labeled ALT and no Meta | |
18251 xmodmap -e "keysym Alt_L = Meta_L Alt_L" | |
18252 @end group | |
18253 @end smallexample | |
18254 | |
18255 @need 1700 | |
18256 @node Mode Line, , Miscellaneous, Emacs Initialization | |
18257 @section A Modified Mode Line | |
18258 @vindex default-mode-line-format | |
18259 @cindex Mode line format | |
18260 | |
18261 Finally, a feature I really like: a modified mode line. | |
18262 | |
18263 When I work over a network, I forget which machine I am using. Also, | |
18264 I tend to I lose track of where I am, and which line point is on. | |
18265 | |
18266 So I reset my mode line to look like this: | |
18267 | |
18268 @smallexample | |
18269 -:-- foo.texi rattlesnake:/home/bob/ Line 1 (Texinfo Fill) Top | |
18270 @end smallexample | |
18271 | |
18272 I am visiting a file called @file{foo.texi}, on my machine | |
18273 @file{rattlesnake} in my @file{/home/bob} buffer. I am on line 1, in | |
18274 Texinfo mode, and am at the top of the buffer. | |
18275 | |
18276 @need 1200 | |
18277 My @file{.emacs} file has a section that looks like this: | |
18278 | |
18279 @smallexample | |
18280 @group | |
18281 ;; Set a Mode Line that tells me which machine, which directory, | |
18282 ;; and which line I am on, plus the other customary information. | |
18283 (setq default-mode-line-format | |
18284 (quote | |
18285 (#("-" 0 1 | |
18286 (help-echo | |
18287 "mouse-1: select window, mouse-2: delete others ...")) | |
18288 mode-line-mule-info | |
18289 mode-line-modified | |
18290 mode-line-frame-identification | |
18291 " " | |
18292 @end group | |
18293 @group | |
18294 mode-line-buffer-identification | |
18295 " " | |
18296 (:eval (substring | |
18297 (system-name) 0 (string-match "\\..+" (system-name)))) | |
18298 ":" | |
18299 default-directory | |
18300 #(" " 0 1 | |
18301 (help-echo | |
18302 "mouse-1: select window, mouse-2: delete others ...")) | |
18303 (line-number-mode " Line %l ") | |
18304 global-mode-string | |
18305 @end group | |
18306 @group | |
18307 #(" %[(" 0 6 | |
18308 (help-echo | |
18309 "mouse-1: select window, mouse-2: delete others ...")) | |
18310 (:eval (mode-line-mode-name)) | |
18311 mode-line-process | |
18312 minor-mode-alist | |
18313 #("%n" 0 2 (help-echo "mouse-2: widen" local-map (keymap ...))) | |
18314 ")%] " | |
18315 (-3 . "%P") | |
18316 ;; "-%-" | |
18317 ))) | |
18318 @end group | |
18319 @end smallexample | |
18320 | |
18321 @noindent | |
18322 Here, I redefine the default mode line. Most of the parts are from | |
18323 the original; but I make a few changes. I set the @emph{default} mode | |
18324 line format so as to permit various modes, such as Info, to override | |
18325 it. | |
18326 | |
18327 Many elements in the list are self-explanatory: | |
18328 @code{mode-line-modified} is a variable that tells whether the buffer | |
18329 has been modified, @code{mode-name} tells the name of the mode, and so | |
18330 on. However, the format looks complicated because of two features we | |
18331 have not discussed. | |
18332 | |
18333 @cindex Properties, in mode line example | |
18334 The first string in the mode line is a dash or hyphen, @samp{-}. In | |
18335 the old days, it would have been specified simply as @code{"-"}. But | |
18336 nowadays, Emacs can add properties to a string, such as highlighting | |
18337 or, as in this case, a help feature. If you place your mouse cursor | |
18338 over the hyphen, some help information appears (By default, you must | |
18339 wait seven-tenths of a second before the information appears. You can | |
18340 change that timing by changing the value of @code{tooltip-delay}.) | |
18341 | |
18342 @need 1000 | |
18343 The new string format has a special syntax: | |
18344 | |
18345 @smallexample | |
18346 #("-" 0 1 (help-echo "mouse-1: select window, ...")) | |
18347 @end smallexample | |
18348 | |
18349 @noindent | |
18350 The @code{#(} begins a list. The first element of the list is the | |
18351 string itself, just one @samp{-}. The second and third | |
18352 elements specify the range over which the fourth element applies. A | |
18353 range starts @emph{after} a character, so a zero means the range | |
18354 starts just before the first character; a 1 means that the range ends | |
18355 just after the first character. The third element is the property for | |
18356 the range. It consists of a property list, a | |
18357 property name, in this case, @samp{help-echo}, followed by a value, in this | |
18358 case, a string. The second, third, and fourth elements of this new | |
18359 string format can be repeated. | |
18360 | |
18361 @xref{Text Properties, , Text Properties, elisp, The GNU Emacs Lisp | |
18362 Reference Manual}, and see @ref{Mode Line Format, , Mode Line Format, | |
18363 elisp, The GNU Emacs Lisp Reference Manual}, for more information. | |
18364 | |
18365 @code{mode-line-buffer-identification} | |
18366 displays the current buffer name. It is a list | |
18367 beginning @code{(#("%12b" 0 4 @dots{}}. | |
18368 The @code{#(} begins the list. | |
18369 | |
18370 The @samp{"%12b"} displays the current buffer name, using the | |
18371 @code{buffer-name} function with which we are familiar; the `12' | |
18372 specifies the maximum number of characters that will be displayed. | |
18373 When a name has fewer characters, whitespace is added to fill out to | |
18374 this number. (Buffer names can and often should be longer than 12 | |
18375 characters; this length works well in a typical 80 column wide | |
18376 window.) | |
18377 | |
18378 @code{:eval} says to evaluate the following form and use the result as | |
18379 a string to display. In this case, the expression displays the first | |
18380 component of the full system name. The end of the first component is | |
18381 a @samp{.} (`period'), so I use the @code{string-match} function to | |
18382 tell me the length of the first component. The substring from the | |
18383 zeroth character to that length is the name of the machine. | |
18384 | |
18385 @need 1250 | |
18386 This is the expression: | |
18387 | |
18388 @smallexample | |
18389 @group | |
18390 (:eval (substring | |
18391 (system-name) 0 (string-match "\\..+" (system-name)))) | |
18392 @end group | |
18393 @end smallexample | |
18394 | |
18395 @samp{%[} and @samp{%]} cause a pair of square brackets | |
18396 to appear for each recursive editing level. @samp{%n} says `Narrow' | |
18397 when narrowing is in effect. @samp{%P} tells you the percentage of | |
18398 the buffer that is above the bottom of the window, or `Top', `Bottom', | |
18399 or `All'. (A lower case @samp{p} tell you the percentage above the | |
18400 @emph{top} of the window.) @samp{%-} inserts enough dashes to fill | |
18401 out the line. | |
18402 | |
18403 Remember, ``You don't have to like Emacs to like it'' --- your own | |
18404 Emacs can have different colors, different commands, and different | |
18405 keys than a default Emacs. | |
18406 | |
18407 On the other hand, if you want to bring up a plain `out of the box' | |
18408 Emacs, with no customization, type: | |
18409 | |
18410 @smallexample | |
18411 emacs -q | |
18412 @end smallexample | |
18413 | |
18414 @noindent | |
18415 This will start an Emacs that does @emph{not} load your | |
18416 @file{~/.emacs} initialization file. A plain, default Emacs. Nothing | |
18417 more. | |
18418 | |
18419 @node Debugging, Conclusion, Emacs Initialization, Top | |
18420 @chapter Debugging | |
18421 @cindex debugging | |
18422 | |
18423 GNU Emacs has two debuggers, @code{debug} and @code{edebug}. The | |
18424 first is built into the internals of Emacs and is always with you; | |
18425 the second requires that you instrument a function before you can use it. | |
18426 | |
18427 Both debuggers are described extensively in @ref{Debugging, , | |
18428 Debugging Lisp Programs, elisp, The GNU Emacs Lisp Reference Manual}. | |
18429 In this chapter, I will walk through a short example of each. | |
18430 | |
18431 @menu | |
18432 * debug:: How to use the built-in debugger. | |
18433 * debug-on-entry:: Start debugging when you call a function. | |
18434 * debug-on-quit:: Start debugging when you quit with @kbd{C-g}. | |
18435 * edebug:: How to use Edebug, a source level debugger. | |
18436 * Debugging Exercises:: | |
18437 @end menu | |
18438 | |
18439 @node debug, debug-on-entry, Debugging, Debugging | |
18440 @section @code{debug} | |
18441 @findex debug | |
18442 | |
18443 Suppose you have written a function definition that is intended to | |
18444 return the sum of the numbers 1 through a given number. (This is the | |
18445 @code{triangle} function discussed earlier. @xref{Decrementing | |
18446 Example, , Example with Decrementing Counter}, for a discussion.) | |
18447 @c xref{Decrementing Loop,, Loop with a Decrementing Counter}, for a discussion.) | |
18448 | |
18449 However, your function definition has a bug. You have mistyped | |
18450 @samp{1=} for @samp{1-}. Here is the broken definition: | |
18451 | |
18452 @findex triangle-bugged | |
18453 @smallexample | |
18454 @group | |
18455 (defun triangle-bugged (number) | |
18456 "Return sum of numbers 1 through NUMBER inclusive." | |
18457 (let ((total 0)) | |
18458 (while (> number 0) | |
18459 (setq total (+ total number)) | |
18460 (setq number (1= number))) ; @r{Error here.} | |
18461 total)) | |
18462 @end group | |
18463 @end smallexample | |
18464 | |
18465 If you are reading this in Info, you can evaluate this definition in | |
18466 the normal fashion. You will see @code{triangle-bugged} appear in the | |
18467 echo area. | |
18468 | |
18469 @need 1250 | |
18470 Now evaluate the @code{triangle-bugged} function with an | |
18471 argument of 4: | |
18472 | |
18473 @smallexample | |
18474 (triangle-bugged 4) | |
18475 @end smallexample | |
18476 | |
18477 @noindent | |
18478 In a recent GNU Emacs, you will create and enter a @file{*Backtrace*} | |
18479 buffer that says: | |
18480 | |
18481 @noindent | |
18482 @smallexample | |
18483 @group | |
18484 ---------- Buffer: *Backtrace* ---------- | |
18485 Debugger entered--Lisp error: (void-function 1=) | |
18486 (1= number) | |
18487 (setq number (1= number)) | |
18488 (while (> number 0) (setq total (+ total number)) | |
18489 (setq number (1= number))) | |
18490 (let ((total 0)) (while (> number 0) (setq total ...) | |
18491 (setq number ...)) total) | |
18492 triangle-bugged(4) | |
18493 @end group | |
18494 @group | |
18495 eval((triangle-bugged 4)) | |
18496 eval-last-sexp-1(nil) | |
18497 eval-last-sexp(nil) | |
18498 call-interactively(eval-last-sexp) | |
18499 ---------- Buffer: *Backtrace* ---------- | |
18500 @end group | |
18501 @end smallexample | |
18502 | |
18503 @noindent | |
18504 (I have reformatted this example slightly; the debugger does not fold | |
18505 long lines. As usual, you can quit the debugger by typing @kbd{q} in | |
18506 the @file{*Backtrace*} buffer.) | |
18507 | |
18508 In practice, for a bug as simple as this, the `Lisp error' line will | |
18509 tell you what you need to know to correct the definition. The | |
18510 function @code{1=} is `void'. | |
18511 | |
18512 @ignore | |
18513 @need 800 | |
18514 In GNU Emacs 20 and before, you will see: | |
18515 | |
18516 @smallexample | |
18517 Symbol's function definition is void:@: 1= | |
18518 @end smallexample | |
18519 | |
18520 @noindent | |
18521 which has the same meaning as the @file{*Backtrace*} buffer line in | |
18522 version 21. | |
18523 @end ignore | |
18524 | |
18525 However, suppose you are not quite certain what is going on? | |
18526 You can read the complete backtrace. | |
18527 | |
18528 In this case, you need to run a recent GNU Emacs, which automatically | |
18529 starts the debugger that puts you in the @file{*Backtrace*} buffer; or | |
18530 else, you need to start the debugger manually as described below. | |
18531 | |
18532 Read the @file{*Backtrace*} buffer from the bottom up; it tells you | |
18533 what Emacs did that led to the error. Emacs made an interactive call | |
18534 to @kbd{C-x C-e} (@code{eval-last-sexp}), which led to the evaluation | |
18535 of the @code{triangle-bugged} expression. Each line above tells you | |
18536 what the Lisp interpreter evaluated next. | |
18537 | |
18538 @need 1250 | |
18539 The third line from the top of the buffer is | |
18540 | |
18541 @smallexample | |
18542 (setq number (1= number)) | |
18543 @end smallexample | |
18544 | |
18545 @noindent | |
18546 Emacs tried to evaluate this expression; in order to do so, it tried | |
18547 to evaluate the inner expression shown on the second line from the | |
18548 top: | |
18549 | |
18550 @smallexample | |
18551 (1= number) | |
18552 @end smallexample | |
18553 | |
18554 @need 1250 | |
18555 @noindent | |
18556 This is where the error occurred; as the top line says: | |
18557 | |
18558 @smallexample | |
18559 Debugger entered--Lisp error: (void-function 1=) | |
18560 @end smallexample | |
18561 | |
18562 @noindent | |
18563 You can correct the mistake, re-evaluate the function definition, and | |
18564 then run your test again. | |
18565 | |
18566 @node debug-on-entry, debug-on-quit, debug, Debugging | |
18567 @section @code{debug-on-entry} | |
18568 @findex debug-on-entry | |
18569 | |
18570 A recent GNU Emacs starts the debugger automatically when your | |
18571 function has an error. | |
18572 | |
18573 @ignore | |
18574 GNU Emacs version 20 and before did not; it simply | |
18575 presented you with an error message. You had to start the debugger | |
18576 manually. | |
18577 @end ignore | |
18578 | |
18579 Incidentally, you can start the debugger manually for all versions of | |
18580 Emacs; the advantage is that the debugger runs even if you do not have | |
18581 a bug in your code. Sometimes your code will be free of bugs! | |
18582 | |
18583 You can enter the debugger when you call the function by calling | |
18584 @code{debug-on-entry}. | |
18585 | |
18586 @need 1250 | |
18587 @noindent | |
18588 Type: | |
18589 | |
18590 @smallexample | |
18591 M-x debug-on-entry RET triangle-bugged RET | |
18592 @end smallexample | |
18593 | |
18594 @need 1250 | |
18595 @noindent | |
18596 Now, evaluate the following: | |
18597 | |
18598 @smallexample | |
18599 (triangle-bugged 5) | |
18600 @end smallexample | |
18601 | |
18602 @noindent | |
18603 All versions of Emacs will create a @file{*Backtrace*} buffer and tell | |
18604 you that it is beginning to evaluate the @code{triangle-bugged} | |
18605 function: | |
18606 | |
18607 @smallexample | |
18608 @group | |
18609 ---------- Buffer: *Backtrace* ---------- | |
18610 Debugger entered--entering a function: | |
18611 * triangle-bugged(5) | |
18612 eval((triangle-bugged 5)) | |
18613 @end group | |
18614 @group | |
18615 eval-last-sexp-1(nil) | |
18616 eval-last-sexp(nil) | |
18617 call-interactively(eval-last-sexp) | |
18618 ---------- Buffer: *Backtrace* ---------- | |
18619 @end group | |
18620 @end smallexample | |
18621 | |
18622 In the @file{*Backtrace*} buffer, type @kbd{d}. Emacs will evaluate | |
18623 the first expression in @code{triangle-bugged}; the buffer will look | |
18624 like this: | |
18625 | |
18626 @smallexample | |
18627 @group | |
18628 ---------- Buffer: *Backtrace* ---------- | |
18629 Debugger entered--beginning evaluation of function call form: | |
18630 * (let ((total 0)) (while (> number 0) (setq total ...) | |
18631 (setq number ...)) total) | |
18632 * triangle-bugged(5) | |
18633 eval((triangle-bugged 5)) | |
18634 @end group | |
18635 @group | |
18636 eval-last-sexp-1(nil) | |
18637 eval-last-sexp(nil) | |
18638 call-interactively(eval-last-sexp) | |
18639 ---------- Buffer: *Backtrace* ---------- | |
18640 @end group | |
18641 @end smallexample | |
18642 | |
18643 @noindent | |
18644 Now, type @kbd{d} again, eight times, slowly. Each time you type | |
18645 @kbd{d}, Emacs will evaluate another expression in the function | |
18646 definition. | |
18647 | |
18648 @need 1750 | |
18649 Eventually, the buffer will look like this: | |
18650 | |
18651 @smallexample | |
18652 @group | |
18653 ---------- Buffer: *Backtrace* ---------- | |
18654 Debugger entered--beginning evaluation of function call form: | |
18655 * (setq number (1= number)) | |
18656 * (while (> number 0) (setq total (+ total number)) | |
18657 (setq number (1= number))) | |
18658 @group | |
18659 @end group | |
18660 * (let ((total 0)) (while (> number 0) (setq total ...) | |
18661 (setq number ...)) total) | |
18662 * triangle-bugged(5) | |
18663 eval((triangle-bugged 5)) | |
18664 @group | |
18665 @end group | |
18666 eval-last-sexp-1(nil) | |
18667 eval-last-sexp(nil) | |
18668 call-interactively(eval-last-sexp) | |
18669 ---------- Buffer: *Backtrace* ---------- | |
18670 @end group | |
18671 @end smallexample | |
18672 | |
18673 @need 1500 | |
18674 @noindent | |
18675 Finally, after you type @kbd{d} two more times, Emacs will reach the | |
18676 error, and the top two lines of the @file{*Backtrace*} buffer will look | |
18677 like this: | |
18678 | |
18679 @smallexample | |
18680 @group | |
18681 ---------- Buffer: *Backtrace* ---------- | |
18682 Debugger entered--Lisp error: (void-function 1=) | |
18683 * (1= number) | |
18684 @dots{} | |
18685 ---------- Buffer: *Backtrace* ---------- | |
18686 @end group | |
18687 @end smallexample | |
18688 | |
18689 By typing @kbd{d}, you were able to step through the function. | |
18690 | |
18691 You can quit a @file{*Backtrace*} buffer by typing @kbd{q} in it; this | |
18692 quits the trace, but does not cancel @code{debug-on-entry}. | |
18693 | |
18694 @findex cancel-debug-on-entry | |
18695 To cancel the effect of @code{debug-on-entry}, call | |
18696 @code{cancel-debug-on-entry} and the name of the function, like this: | |
18697 | |
18698 @smallexample | |
18699 M-x cancel-debug-on-entry RET triangle-bugged RET | |
18700 @end smallexample | |
18701 | |
18702 @noindent | |
18703 (If you are reading this in Info, cancel @code{debug-on-entry} now.) | |
18704 | |
18705 @node debug-on-quit, edebug, debug-on-entry, Debugging | |
18706 @section @code{debug-on-quit} and @code{(debug)} | |
18707 | |
18708 In addition to setting @code{debug-on-error} or calling @code{debug-on-entry}, | |
18709 there are two other ways to start @code{debug}. | |
18710 | |
18711 @findex debug-on-quit | |
18712 You can start @code{debug} whenever you type @kbd{C-g} | |
18713 (@code{keyboard-quit}) by setting the variable @code{debug-on-quit} to | |
18714 @code{t}. This is useful for debugging infinite loops. | |
18715 | |
18716 @need 1500 | |
18717 @cindex @code{(debug)} in code | |
18718 Or, you can insert a line that says @code{(debug)} into your code | |
18719 where you want the debugger to start, like this: | |
18720 | |
18721 @smallexample | |
18722 @group | |
18723 (defun triangle-bugged (number) | |
18724 "Return sum of numbers 1 through NUMBER inclusive." | |
18725 (let ((total 0)) | |
18726 (while (> number 0) | |
18727 (setq total (+ total number)) | |
18728 (debug) ; @r{Start debugger.} | |
18729 (setq number (1= number))) ; @r{Error here.} | |
18730 total)) | |
18731 @end group | |
18732 @end smallexample | |
18733 | |
18734 The @code{debug} function is described in detail in @ref{Debugger, , | |
18735 The Lisp Debugger, elisp, The GNU Emacs Lisp Reference Manual}. | |
18736 | |
18737 @node edebug, Debugging Exercises, debug-on-quit, Debugging | |
18738 @section The @code{edebug} Source Level Debugger | |
18739 @cindex Source level debugger | |
18740 @findex edebug | |
18741 | |
18742 Edebug is a source level debugger. Edebug normally displays the | |
18743 source of the code you are debugging, with an arrow at the left that | |
18744 shows which line you are currently executing. | |
18745 | |
18746 You can walk through the execution of a function, line by line, or run | |
18747 quickly until reaching a @dfn{breakpoint} where execution stops. | |
18748 | |
18749 Edebug is described in @ref{edebug, , Edebug, elisp, The GNU Emacs | |
18750 Lisp Reference Manual}. | |
18751 | |
18752 @need 1250 | |
18753 Here is a bugged function definition for @code{triangle-recursively}. | |
18754 @xref{Recursive triangle function, , Recursion in place of a counter}, | |
18755 for a review of it. | |
18756 | |
18757 @smallexample | |
18758 @group | |
18759 (defun triangle-recursively-bugged (number) | |
18760 "Return sum of numbers 1 through NUMBER inclusive. | |
18761 Uses recursion." | |
18762 (if (= number 1) | |
18763 1 | |
18764 (+ number | |
18765 (triangle-recursively-bugged | |
18766 (1= number))))) ; @r{Error here.} | |
18767 @end group | |
18768 @end smallexample | |
18769 | |
18770 @noindent | |
18771 Normally, you would install this definition by positioning your cursor | |
18772 after the function's closing parenthesis and typing @kbd{C-x C-e} | |
18773 (@code{eval-last-sexp}) or else by positioning your cursor within the | |
18774 definition and typing @kbd{C-M-x} (@code{eval-defun}). (By default, | |
18775 the @code{eval-defun} command works only in Emacs Lisp mode or in Lisp | |
18776 Interactive mode.) | |
18777 | |
18778 @need 1500 | |
18779 However, to prepare this function definition for Edebug, you must | |
18780 first @dfn{instrument} the code using a different command. You can do | |
18781 this by positioning your cursor within or just after the definition | |
18782 and typing | |
18783 | |
18784 @smallexample | |
18785 M-x edebug-defun RET | |
18786 @end smallexample | |
18787 | |
18788 @noindent | |
18789 This will cause Emacs to load Edebug automatically if it is not | |
18790 already loaded, and properly instrument the function. | |
18791 | |
18792 After instrumenting the function, place your cursor after the | |
18793 following expression and type @kbd{C-x C-e} (@code{eval-last-sexp}): | |
18794 | |
18795 @smallexample | |
18796 (triangle-recursively-bugged 3) | |
18797 @end smallexample | |
18798 | |
18799 @noindent | |
18800 You will be jumped back to the source for | |
18801 @code{triangle-recursively-bugged} and the cursor positioned at the | |
18802 beginning of the @code{if} line of the function. Also, you will see | |
18803 an arrowhead at the left hand side of that line. The arrowhead marks | |
18804 the line where the function is executing. (In the following examples, | |
18805 we show the arrowhead with @samp{=>}; in a windowing system, you may | |
18806 see the arrowhead as a solid triangle in the window `fringe'.) | |
18807 | |
18808 @smallexample | |
18809 =>@point{}(if (= number 1) | |
18810 @end smallexample | |
18811 | |
18812 @noindent | |
18813 @iftex | |
18814 In the example, the location of point is displayed with a star, | |
18815 @samp{@point{}} (in Info, it is displayed as @samp{-!-}). | |
18816 @end iftex | |
18817 @ifnottex | |
18818 In the example, the location of point is displayed as @samp{@point{}} | |
18819 (in a printed book, it is displayed with a five pointed star). | |
18820 @end ifnottex | |
18821 | |
18822 If you now press @key{SPC}, point will move to the next expression to | |
18823 be executed; the line will look like this: | |
18824 | |
18825 @smallexample | |
18826 =>(if @point{}(= number 1) | |
18827 @end smallexample | |
18828 | |
18829 @noindent | |
18830 As you continue to press @key{SPC}, point will move from expression to | |
18831 expression. At the same time, whenever an expression returns a value, | |
18832 that value will be displayed in the echo area. For example, after you | |
18833 move point past @code{number}, you will see the following: | |
18834 | |
18835 @smallexample | |
18836 Result: 3 (#o3, #x3, ?\C-c) | |
18837 @end smallexample | |
18838 | |
18839 @noindent | |
18840 This means the value of @code{number} is 3, which is octal three, | |
18841 hexadecimal three, and @sc{ascii} `control-c' (the third letter of the | |
18842 alphabet, in case you need to know this information). | |
18843 | |
18844 You can continue moving through the code until you reach the line with | |
18845 the error. Before evaluation, that line looks like this: | |
18846 | |
18847 @smallexample | |
18848 => @point{}(1= number))))) ; @r{Error here.} | |
18849 @end smallexample | |
18850 | |
18851 @need 1250 | |
18852 @noindent | |
18853 When you press @key{SPC} once again, you will produce an error message | |
18854 that says: | |
18855 | |
18856 @smallexample | |
18857 Symbol's function definition is void:@: 1= | |
18858 @end smallexample | |
18859 | |
18860 @noindent | |
18861 This is the bug. | |
18862 | |
18863 Press @kbd{q} to quit Edebug. | |
18864 | |
18865 To remove instrumentation from a function definition, simply | |
18866 re-evaluate it with a command that does not instrument it. | |
18867 For example, you could place your cursor after the definition's | |
18868 closing parenthesis and type @kbd{C-x C-e}. | |
18869 | |
18870 Edebug does a great deal more than walk with you through a function. | |
18871 You can set it so it races through on its own, stopping only at an | |
18872 error or at specified stopping points; you can cause it to display the | |
18873 changing values of various expressions; you can find out how many | |
18874 times a function is called, and more. | |
18875 | |
18876 Edebug is described in @ref{edebug, , Edebug, elisp, The GNU Emacs | |
18877 Lisp Reference Manual}. | |
18878 | |
18879 @need 1500 | |
18880 @node Debugging Exercises, , edebug, Debugging | |
18881 @section Debugging Exercises | |
18882 | |
18883 @itemize @bullet | |
18884 @item | |
18885 Install the @code{count-words-region} function and then cause it to | |
18886 enter the built-in debugger when you call it. Run the command on a | |
18887 region containing two words. You will need to press @kbd{d} a | |
18888 remarkable number of times. On your system, is a `hook' called after | |
18889 the command finishes? (For information on hooks, see @ref{Command | |
18890 Overview, , Command Loop Overview, elisp, The GNU Emacs Lisp Reference | |
18891 Manual}.) | |
18892 | |
18893 @item | |
18894 Copy @code{count-words-region} into the @file{*scratch*} buffer, | |
18895 instrument the function for Edebug, and walk through its execution. | |
18896 The function does not need to have a bug, although you can introduce | |
18897 one if you wish. If the function lacks a bug, the walk-through | |
18898 completes without problems. | |
18899 | |
18900 @item | |
18901 While running Edebug, type @kbd{?} to see a list of all the Edebug commands. | |
18902 (The @code{global-edebug-prefix} is usually @kbd{C-x X}, i.e.@: | |
18903 @kbd{@key{CTRL}-x} followed by an upper case @kbd{X}; use this prefix | |
18904 for commands made outside of the Edebug debugging buffer.) | |
18905 | |
18906 @item | |
18907 In the Edebug debugging buffer, use the @kbd{p} | |
18908 (@code{edebug-bounce-point}) command to see where in the region the | |
18909 @code{count-words-region} is working. | |
18910 | |
18911 @item | |
18912 Move point to some spot further down the function and then type the | |
18913 @kbd{h} (@code{edebug-goto-here}) command to jump to that location. | |
18914 | |
18915 @item | |
18916 Use the @kbd{t} (@code{edebug-trace-mode}) command to cause Edebug to | |
18917 walk through the function on its own; use an upper case @kbd{T} for | |
18918 @code{edebug-Trace-fast-mode}. | |
18919 | |
18920 @item | |
18921 Set a breakpoint, then run Edebug in Trace mode until it reaches the | |
18922 stopping point. | |
18923 @end itemize | |
18924 | |
18925 @node Conclusion, the-the, Debugging, Top | |
18926 @chapter Conclusion | |
18927 | |
18928 We have now reached the end of this Introduction. You have now | |
18929 learned enough about programming in Emacs Lisp to set values, to write | |
18930 simple @file{.emacs} files for yourself and your friends, and write | |
18931 simple customizations and extensions to Emacs. | |
18932 | |
18933 This is a place to stop. Or, if you wish, you can now go onward, and | |
18934 teach yourself. | |
18935 | |
18936 You have learned some of the basic nuts and bolts of programming. But | |
18937 only some. There are a great many more brackets and hinges that are | |
18938 easy to use that we have not touched. | |
18939 | |
18940 A path you can follow right now lies among the sources to GNU Emacs | |
18941 and in | |
18942 @ifnotinfo | |
18943 @cite{The GNU Emacs Lisp Reference Manual}. | |
18944 @end ifnotinfo | |
18945 @ifinfo | |
18946 @ref{Top, , The GNU Emacs Lisp Reference Manual, elisp, The GNU | |
18947 Emacs Lisp Reference Manual}. | |
18948 @end ifinfo | |
18949 | |
18950 The Emacs Lisp sources are an adventure. When you read the sources and | |
18951 come across a function or expression that is unfamiliar, you need to | |
18952 figure out or find out what it does. | |
18953 | |
18954 Go to the Reference Manual. It is a thorough, complete, and fairly | |
18955 easy-to-read description of Emacs Lisp. It is written not only for | |
18956 experts, but for people who know what you know. (The @cite{Reference | |
18957 Manual} comes with the standard GNU Emacs distribution. Like this | |
18958 introduction, it comes as a Texinfo source file, so you can read it | |
18959 on-line and as a typeset, printed book.) | |
18960 | |
18961 Go to the other on-line help that is part of GNU Emacs: the on-line | |
18962 documentation for all functions and variables, and @code{find-tags}, | |
18963 the program that takes you to sources. | |
18964 | |
18965 Here is an example of how I explore the sources. Because of its name, | |
18966 @file{simple.el} is the file I looked at first, a long time ago. As | |
18967 it happens some of the functions in @file{simple.el} are complicated, | |
18968 or at least look complicated at first sight. The @code{open-line} | |
18969 function, for example, looks complicated. | |
18970 | |
18971 You may want to walk through this function slowly, as we did with the | |
18972 @code{forward-sentence} function. (@xref{forward-sentence, The | |
18973 @code{forward-sentence} function}.) Or you may want to skip that | |
18974 function and look at another, such as @code{split-line}. You don't | |
18975 need to read all the functions. According to | |
18976 @code{count-words-in-defun}, the @code{split-line} function contains | |
18977 102 words and symbols. | |
18978 | |
18979 Even though it is short, @code{split-line} contains expressions | |
18980 we have not studied: @code{skip-chars-forward}, @code{indent-to}, | |
18981 @code{current-column} and @code{insert-and-inherit}. | |
18982 | |
18983 Consider the @code{skip-chars-forward} function. (It is part of the | |
18984 function definition for @code{back-to-indentation}, which is shown in | |
18985 @ref{Review, , Review}.) | |
18986 | |
18987 In GNU Emacs, you can find out more about @code{skip-chars-forward} by | |
18988 typing @kbd{C-h f} (@code{describe-function}) and the name of the | |
18989 function. This gives you the function documentation. | |
18990 | |
18991 You may be able to guess what is done by a well named function such as | |
18992 @code{indent-to}; or you can look it up, too. Incidentally, the | |
18993 @code{describe-function} function itself is in @file{help.el}; it is | |
18994 one of those long, but decipherable functions. You can look up | |
18995 @code{describe-function} using the @kbd{C-h f} command! | |
18996 | |
18997 In this instance, since the code is Lisp, the @file{*Help*} buffer | |
18998 contains the name of the library containing the function's source. | |
18999 You can put point over the name of the library and press the RET key, | |
19000 which in this situation is bound to @code{help-follow}, and be taken | |
19001 directly to the source, in the same way as @kbd{M-.} | |
19002 (@code{find-tag}). | |
19003 | |
19004 The definition for @code{describe-function} illustrates how to | |
19005 customize the @code{interactive} expression without using the standard | |
19006 character codes; and it shows how to create a temporary buffer. | |
19007 | |
19008 (The @code{indent-to} function is written in C rather than Emacs Lisp; | |
19009 it is a `built-in' function. @code{help-follow} takes you to its | |
19010 source as does @code{find-tag}, when properly set up.) | |
19011 | |
19012 You can look at a function's source using @code{find-tag}, which is | |
19013 bound to @kbd{M-.} Finally, you can find out what the Reference | |
19014 Manual has to say by visiting the manual in Info, and typing @kbd{i} | |
19015 (@code{Info-index}) and the name of the function, or by looking up the | |
19016 function in the index to a printed copy of the manual. | |
19017 | |
19018 Similarly, you can find out what is meant by | |
19019 @code{insert-and-inherit}. | |
19020 | |
19021 Other interesting source files include @file{paragraphs.el}, | |
19022 @file{loaddefs.el}, and @file{loadup.el}. The @file{paragraphs.el} | |
19023 file includes short, easily understood functions as well as longer | |
19024 ones. The @file{loaddefs.el} file contains the many standard | |
19025 autoloads and many keymaps. I have never looked at it all; only at | |
19026 parts. @file{loadup.el} is the file that loads the standard parts of | |
19027 Emacs; it tells you a great deal about how Emacs is built. | |
19028 (@xref{Building Emacs, , Building Emacs, elisp, The GNU Emacs Lisp | |
19029 Reference Manual}, for more about building.) | |
19030 | |
19031 As I said, you have learned some nuts and bolts; however, and very | |
19032 importantly, we have hardly touched major aspects of programming; I | |
19033 have said nothing about how to sort information, except to use the | |
19034 predefined @code{sort} function; I have said nothing about how to store | |
19035 information, except to use variables and lists; I have said nothing | |
19036 about how to write programs that write programs. These are topics for | |
19037 another, and different kind of book, a different kind of learning. | |
19038 | |
19039 What you have done is learn enough for much practical work with GNU | |
19040 Emacs. What you have done is get started. This is the end of a | |
19041 beginning. | |
19042 | |
19043 @c ================ Appendix ================ | |
19044 | |
19045 @node the-the, Kill Ring, Conclusion, Top | |
19046 @appendix The @code{the-the} Function | |
19047 @findex the-the | |
19048 @cindex Duplicated words function | |
19049 @cindex Words, duplicated | |
19050 | |
19051 Sometimes when you you write text, you duplicate words---as with ``you | |
19052 you'' near the beginning of this sentence. I find that most | |
19053 frequently, I duplicate ``the''; hence, I call the function for | |
19054 detecting duplicated words, @code{the-the}. | |
19055 | |
19056 @need 1250 | |
19057 As a first step, you could use the following regular expression to | |
19058 search for duplicates: | |
19059 | |
19060 @smallexample | |
19061 \\(\\w+[ \t\n]+\\)\\1 | |
19062 @end smallexample | |
19063 | |
19064 @noindent | |
19065 This regexp matches one or more word-constituent characters followed | |
19066 by one or more spaces, tabs, or newlines. However, it does not detect | |
19067 duplicated words on different lines, since the ending of the first | |
19068 word, the end of the line, is different from the ending of the second | |
19069 word, a space. (For more information about regular expressions, see | |
19070 @ref{Regexp Search, , Regular Expression Searches}, as well as | |
19071 @ref{Regexps, , Syntax of Regular Expressions, emacs, The GNU Emacs | |
19072 Manual}, and @ref{Regular Expressions, , Regular Expressions, elisp, | |
19073 The GNU Emacs Lisp Reference Manual}.) | |
19074 | |
19075 You might try searching just for duplicated word-constituent | |
19076 characters but that does not work since the pattern detects doubles | |
19077 such as the two occurrences of `th' in `with the'. | |
19078 | |
19079 Another possible regexp searches for word-constituent characters | |
19080 followed by non-word-constituent characters, reduplicated. Here, | |
19081 @w{@samp{\\w+}} matches one or more word-constituent characters and | |
19082 @w{@samp{\\W*}} matches zero or more non-word-constituent characters. | |
19083 | |
19084 @smallexample | |
19085 \\(\\(\\w+\\)\\W*\\)\\1 | |
19086 @end smallexample | |
19087 | |
19088 @noindent | |
19089 Again, not useful. | |
19090 | |
19091 Here is the pattern that I use. It is not perfect, but good enough. | |
19092 @w{@samp{\\b}} matches the empty string, provided it is at the beginning | |
19093 or end of a word; @w{@samp{[^@@ \n\t]+}} matches one or more occurrences of | |
19094 any characters that are @emph{not} an @@-sign, space, newline, or tab. | |
19095 | |
19096 @smallexample | |
19097 \\b\\([^@@ \n\t]+\\)[ \n\t]+\\1\\b | |
19098 @end smallexample | |
19099 | |
19100 One can write more complicated expressions, but I found that this | |
19101 expression is good enough, so I use it. | |
19102 | |
19103 Here is the @code{the-the} function, as I include it in my | |
19104 @file{.emacs} file, along with a handy global key binding: | |
19105 | |
19106 @smallexample | |
19107 @group | |
19108 (defun the-the () | |
19109 "Search forward for for a duplicated word." | |
19110 (interactive) | |
19111 (message "Searching for for duplicated words ...") | |
19112 (push-mark) | |
19113 @end group | |
19114 @group | |
19115 ;; This regexp is not perfect | |
19116 ;; but is fairly good over all: | |
19117 (if (re-search-forward | |
19118 "\\b\\([^@@ \n\t]+\\)[ \n\t]+\\1\\b" nil 'move) | |
19119 (message "Found duplicated word.") | |
19120 (message "End of buffer"))) | |
19121 @end group | |
19122 | |
19123 @group | |
19124 ;; Bind `the-the' to C-c \ | |
19125 (global-set-key "\C-c\\" 'the-the) | |
19126 @end group | |
19127 @end smallexample | |
19128 | |
19129 @sp 1 | |
19130 Here is test text: | |
19131 | |
19132 @smallexample | |
19133 @group | |
19134 one two two three four five | |
19135 five six seven | |
19136 @end group | |
19137 @end smallexample | |
19138 | |
19139 You can substitute the other regular expressions shown above in the | |
19140 function definition and try each of them on this list. | |
19141 | |
19142 @node Kill Ring, Full Graph, the-the, Top | |
19143 @appendix Handling the Kill Ring | |
19144 @cindex Kill ring handling | |
19145 @cindex Handling the kill ring | |
19146 @cindex Ring, making a list like a | |
19147 | |
19148 The kill ring is a list that is transformed into a ring by the | |
19149 workings of the @code{current-kill} function. The @code{yank} and | |
19150 @code{yank-pop} commands use the @code{current-kill} function. | |
19151 | |
19152 This appendix describes the @code{current-kill} function as well as | |
19153 both the @code{yank} and the @code{yank-pop} commands, but first, | |
19154 consider the workings of the kill ring. | |
19155 | |
19156 @menu | |
19157 * What the Kill Ring Does:: | |
19158 * current-kill:: | |
19159 * yank:: Paste a copy of a clipped element. | |
19160 * yank-pop:: Insert element pointed to. | |
19161 * ring file:: | |
19162 @end menu | |
19163 | |
19164 @node What the Kill Ring Does, current-kill, Kill Ring, Kill Ring | |
19165 @ifnottex | |
19166 @unnumberedsec What the Kill Ring Does | |
19167 @end ifnottex | |
19168 | |
19169 @need 1250 | |
19170 The kill ring has a default maximum length of sixty items; this number | |
19171 is too large for an explanation. Instead, set it to four. Please | |
19172 evaluate the following: | |
19173 | |
19174 @smallexample | |
19175 @group | |
19176 (setq old-kill-ring-max kill-ring-max) | |
19177 (setq kill-ring-max 4) | |
19178 @end group | |
19179 @end smallexample | |
19180 | |
19181 @noindent | |
19182 Then, please copy each line of the following indented example into the | |
19183 kill ring. You may kill each line with @kbd{C-k} or mark it and copy | |
19184 it with @kbd{M-w}. | |
19185 | |
19186 @noindent | |
19187 (In a read-only buffer, such as the @file{*info*} buffer, the kill | |
19188 command, @kbd{C-k} (@code{kill-line}), will not remove the text, | |
19189 merely copy it to the kill ring. However, your machine may beep at | |
19190 you. Alternatively, for silence, you may copy the region of each line | |
19191 with the @kbd{M-w} (@code{kill-ring-save}) command. You must mark | |
19192 each line for this command to succeed, but it does not matter at which | |
19193 end you put point or mark.) | |
19194 | |
19195 @need 1250 | |
19196 @noindent | |
19197 Please invoke the calls in order, so that five elements attempt to | |
19198 fill the kill ring: | |
19199 | |
19200 @smallexample | |
19201 @group | |
19202 first some text | |
19203 second piece of text | |
19204 third line | |
19205 fourth line of text | |
19206 fifth bit of text | |
19207 @end group | |
19208 @end smallexample | |
19209 | |
19210 @need 1250 | |
19211 @noindent | |
19212 Then find the value of @code{kill-ring} by evaluating | |
19213 | |
19214 @smallexample | |
19215 kill-ring | |
19216 @end smallexample | |
19217 | |
19218 @need 800 | |
19219 @noindent | |
19220 It is: | |
19221 | |
19222 @smallexample | |
19223 @group | |
19224 ("fifth bit of text" "fourth line of text" | |
19225 "third line" "second piece of text") | |
19226 @end group | |
19227 @end smallexample | |
19228 | |
19229 @noindent | |
19230 The first element, @samp{first some text}, was dropped. | |
19231 | |
19232 @need 1250 | |
19233 To return to the old value for the length of the kill ring, evaluate: | |
19234 | |
19235 @smallexample | |
19236 (setq kill-ring-max old-kill-ring-max) | |
19237 @end smallexample | |
19238 | |
19239 @node current-kill, yank, What the Kill Ring Does, Kill Ring | |
19240 @comment node-name, next, previous, up | |
19241 @appendixsec The @code{current-kill} Function | |
19242 @findex current-kill | |
19243 | |
19244 The @code{current-kill} function changes the element in the kill ring | |
19245 to which @code{kill-ring-yank-pointer} points. (Also, the | |
19246 @code{kill-new} function sets @code{kill-ring-yank-pointer} to point | |
19247 to the latest element of the the kill ring. The @code{kill-new} | |
19248 function is used directly or indirectly by @code{kill-append}, | |
19249 @code{copy-region-as-kill}, @code{kill-ring-save}, @code{kill-line}, | |
19250 and @code{kill-region}.) | |
19251 | |
19252 @menu | |
19253 * Code for current-kill:: | |
19254 * Understanding current-kill:: | |
19255 @end menu | |
19256 | |
19257 @node Code for current-kill, Understanding current-kill, current-kill, current-kill | |
19258 @ifnottex | |
19259 @unnumberedsubsec The code for @code{current-kill} | |
19260 @end ifnottex | |
19261 | |
19262 | |
19263 @need 1500 | |
19264 The @code{current-kill} function is used by @code{yank} and by | |
19265 @code{yank-pop}. Here is the code for @code{current-kill}: | |
19266 | |
19267 @smallexample | |
19268 @group | |
19269 (defun current-kill (n &optional do-not-move) | |
19270 "Rotate the yanking point by N places, and then return that kill. | |
19271 If N is zero, `interprogram-paste-function' is set, and calling it | |
19272 returns a string, then that string is added to the front of the | |
19273 kill ring and returned as the latest kill. | |
19274 @end group | |
19275 @group | |
19276 If optional arg DO-NOT-MOVE is non-nil, then don't actually move the | |
19277 yanking point; just return the Nth kill forward." | |
19278 (let ((interprogram-paste (and (= n 0) | |
19279 interprogram-paste-function | |
19280 (funcall interprogram-paste-function)))) | |
19281 @end group | |
19282 @group | |
19283 (if interprogram-paste | |
19284 (progn | |
19285 ;; Disable the interprogram cut function when we add the new | |
19286 ;; text to the kill ring, so Emacs doesn't try to own the | |
19287 ;; selection, with identical text. | |
19288 (let ((interprogram-cut-function nil)) | |
19289 (kill-new interprogram-paste)) | |
19290 interprogram-paste) | |
19291 @end group | |
19292 @group | |
19293 (or kill-ring (error "Kill ring is empty")) | |
19294 (let ((ARGth-kill-element | |
19295 (nthcdr (mod (- n (length kill-ring-yank-pointer)) | |
19296 (length kill-ring)) | |
19297 kill-ring))) | |
19298 (or do-not-move | |
19299 (setq kill-ring-yank-pointer ARGth-kill-element)) | |
19300 (car ARGth-kill-element))))) | |
19301 @end group | |
19302 @end smallexample | |
19303 | |
19304 Remember also that the @code{kill-new} function sets | |
19305 @code{kill-ring-yank-pointer} to the latest element of the the kill | |
19306 ring, which means that all the functions that call it set the value | |
19307 indirectly: @code{kill-append}, @code{copy-region-as-kill}, | |
19308 @code{kill-ring-save}, @code{kill-line}, and @code{kill-region}. | |
19309 | |
19310 @need 1500 | |
19311 Here is the line in @code{kill-new}, which is explained in | |
19312 @ref{kill-new function, , The @code{kill-new} function}. | |
19313 | |
19314 @smallexample | |
19315 (setq kill-ring-yank-pointer kill-ring) | |
19316 @end smallexample | |
19317 | |
19318 @node Understanding current-kill, , Code for current-kill, current-kill | |
19319 @ifnottex | |
19320 @unnumberedsubsec @code{current-kill} in Outline | |
19321 @end ifnottex | |
19322 | |
19323 The @code{current-kill} function looks complex, but as usual, it can | |
19324 be understood by taking it apart piece by piece. First look at it in | |
19325 skeletal form: | |
19326 | |
19327 @smallexample | |
19328 @group | |
19329 (defun current-kill (n &optional do-not-move) | |
19330 "Rotate the yanking point by N places, and then return that kill." | |
19331 (let @var{varlist} | |
19332 @var{body}@dots{}) | |
19333 @end group | |
19334 @end smallexample | |
19335 | |
19336 This function takes two arguments, one of which is optional. It has a | |
19337 documentation string. It is @emph{not} interactive. | |
19338 | |
19339 @menu | |
19340 * Body of current-kill:: | |
19341 * Digression concerning error:: How to mislead humans, but not computers. | |
19342 * Determining the Element:: | |
19343 @end menu | |
19344 | |
19345 @node Body of current-kill, Digression concerning error, Understanding current-kill, Understanding current-kill | |
19346 @ifnottex | |
19347 @unnumberedsubsubsec The Body of @code{current-kill} | |
19348 @end ifnottex | |
19349 | |
19350 The body of the function definition is a @code{let} expression, which | |
19351 itself has a body as well as a @var{varlist}. | |
19352 | |
19353 The @code{let} expression declares a variable that will be only usable | |
19354 within the bounds of this function. This variable is called | |
19355 @code{interprogram-paste} and is for copying to another program. It | |
19356 is not for copying within this instance of GNU Emacs. Most window | |
19357 systems provide a facility for interprogram pasting. Sadly, that | |
19358 facility usually provides only for the last element. Most windowing | |
19359 systems have not adopted a ring of many possibilities, even though | |
19360 Emacs has provided it for decades. | |
19361 | |
19362 The @code{if} expression has two parts, one if there exists | |
19363 @code{interprogram-paste} and one if not. | |
19364 | |
19365 @need 2000 | |
19366 Let us consider the `if not' or else-part of the @code{current-kill} | |
19367 function. (The then-part uses the the @code{kill-new} function, which | |
19368 we have already described. @xref{kill-new function, , The | |
19369 @code{kill-new} function}.) | |
19370 | |
19371 @smallexample | |
19372 @group | |
19373 (or kill-ring (error "Kill ring is empty")) | |
19374 (let ((ARGth-kill-element | |
19375 (nthcdr (mod (- n (length kill-ring-yank-pointer)) | |
19376 (length kill-ring)) | |
19377 kill-ring))) | |
19378 (or do-not-move | |
19379 (setq kill-ring-yank-pointer ARGth-kill-element)) | |
19380 (car ARGth-kill-element)) | |
19381 @end group | |
19382 @end smallexample | |
19383 | |
19384 @noindent | |
19385 The code first checks whether the kill ring has content; otherwise it | |
19386 signals an error. | |
19387 | |
19388 @need 1000 | |
19389 Note that the @code{or} expression is very similar to testing length | |
19390 with an @code{if}: | |
19391 | |
19392 @findex zerop | |
19393 @findex error | |
19394 @smallexample | |
19395 @group | |
19396 (if (zerop (length kill-ring)) ; @r{if-part} | |
19397 (error "Kill ring is empty")) ; @r{then-part} | |
19398 ;; No else-part | |
19399 @end group | |
19400 @end smallexample | |
19401 | |
19402 @noindent | |
19403 If there is not anything in the kill ring, its length must be zero and | |
19404 an error message sent to the user: @samp{Kill ring is empty}. The | |
19405 @code{current-kill} function uses an @code{or} expression which is | |
19406 simpler. But an @code{if} expression reminds us what goes on. | |
19407 | |
19408 This @code{if} expression uses the function @code{zerop} which returns | |
19409 true if the value it is testing is zero. When @code{zerop} tests | |
19410 true, the then-part of the @code{if} is evaluated. The then-part is a | |
19411 list starting with the function @code{error}, which is a function that | |
19412 is similar to the @code{message} function | |
19413 (@pxref{message, , The @code{message} Function}) in that | |
19414 it prints a one-line message in the echo area. However, in addition | |
19415 to printing a message, @code{error} also stops evaluation of the | |
19416 function within which it is embedded. This means that the rest of the | |
19417 function will not be evaluated if the length of the kill ring is zero. | |
19418 | |
19419 Then the @code{current-kill} function selects the element to return. | |
19420 The selection depends on the number of places that @code{current-kill} | |
19421 rotates and on where @code{kill-ring-yank-pointer} points. | |
19422 | |
19423 Next, either the optional @code{do-not-move} argument is true or the | |
19424 current value of @code{kill-ring-yank-pointer} is set to point to the | |
19425 list. Finally, another expression returns the first element of the | |
19426 list even if the @code{do-not-move} argument is true. | |
19427 | |
19428 @node Digression concerning error, Determining the Element, Body of current-kill, Understanding current-kill | |
19429 @ifnottex | |
19430 @unnumberedsubsubsec Digression about the word `error' | |
19431 @end ifnottex | |
19432 | |
19433 In my opinion, it is slightly misleading, at least to humans, to use | |
19434 the term `error' as the name of the @code{error} function. A better | |
19435 term would be `cancel'. Strictly speaking, of course, you cannot | |
19436 point to, much less rotate a pointer to a list that has no length, so | |
19437 from the point of view of the computer, the word `error' is correct. | |
19438 But a human expects to attempt this sort of thing, if only to find out | |
19439 whether the kill ring is full or empty. This is an act of | |
19440 exploration. | |
19441 | |
19442 From the human point of view, the act of exploration and discovery is | |
19443 not necessarily an error, and therefore should not be labelled as one, | |
19444 even in the bowels of a computer. As it is, the code in Emacs implies | |
19445 that a human who is acting virtuously, by exploring his or her | |
19446 environment, is making an error. This is bad. Even though the computer | |
19447 takes the same steps as it does when there is an `error', a term such as | |
19448 `cancel' would have a clearer connotation. | |
19449 | |
19450 @node Determining the Element, , Digression concerning error, Understanding current-kill | |
19451 @ifnottex | |
19452 @unnumberedsubsubsec Determining the Element | |
19453 @end ifnottex | |
19454 | |
19455 Among other actions, the else-part of the @code{if} expression sets | |
19456 the value of @code{kill-ring-yank-pointer} to | |
19457 @code{ARGth-kill-element} when the kill ring has something in it and | |
19458 the value of @code{do-not-move} is @code{nil}. | |
19459 | |
19460 @need 800 | |
19461 The code looks like this: | |
19462 | |
19463 @smallexample | |
19464 @group | |
19465 (nthcdr (mod (- n (length kill-ring-yank-pointer)) | |
19466 (length kill-ring)) | |
19467 kill-ring))) | |
19468 @end group | |
19469 @end smallexample | |
19470 | |
19471 This needs some examination. Unless it is not supposed to move the | |
19472 pointer, the @code{current-kill} function changes where | |
19473 @code{kill-ring-yank-pointer} points. | |
19474 That is what the | |
19475 @w{@code{(setq kill-ring-yank-pointer ARGth-kill-element))}} | |
19476 expression does. Also, clearly, @code{ARGth-kill-element} is being | |
19477 set to be equal to some @sc{cdr} of the kill ring, using the | |
19478 @code{nthcdr} function that is described in an earlier section. | |
19479 (@xref{copy-region-as-kill}.) How does it do this? | |
19480 | |
19481 As we have seen before (@pxref{nthcdr}), the @code{nthcdr} function | |
19482 works by repeatedly taking the @sc{cdr} of a list---it takes the | |
19483 @sc{cdr} of the @sc{cdr} of the @sc{cdr} @dots{} | |
19484 | |
19485 @need 800 | |
19486 The two following expressions produce the same result: | |
19487 | |
19488 @smallexample | |
19489 @group | |
19490 (setq kill-ring-yank-pointer (cdr kill-ring)) | |
19491 | |
19492 (setq kill-ring-yank-pointer (nthcdr 1 kill-ring)) | |
19493 @end group | |
19494 @end smallexample | |
19495 | |
19496 However, the @code{nthcdr} expression is more complicated. It uses | |
19497 the @code{mod} function to determine which @sc{cdr} to select. | |
19498 | |
19499 (You will remember to look at inner functions first; indeed, we will | |
19500 have to go inside the @code{mod}.) | |
19501 | |
19502 The @code{mod} function returns the value of its first argument modulo | |
19503 the second; that is to say, it returns the remainder after dividing | |
19504 the first argument by the second. The value returned has the same | |
19505 sign as the second argument. | |
19506 | |
19507 @need 800 | |
19508 Thus, | |
19509 | |
19510 @smallexample | |
19511 @group | |
19512 (mod 12 4) | |
19513 @result{} 0 ;; @r{because there is no remainder} | |
19514 (mod 13 4) | |
19515 @result{} 1 | |
19516 @end group | |
19517 @end smallexample | |
19518 | |
19519 @need 1250 | |
19520 In this case, the first argument is often smaller than the second. | |
19521 That is fine. | |
19522 | |
19523 @smallexample | |
19524 @group | |
19525 (mod 0 4) | |
19526 @result{} 0 | |
19527 (mod 1 4) | |
19528 @result{} 1 | |
19529 @end group | |
19530 @end smallexample | |
19531 | |
19532 We can guess what the @code{-} function does. It is like @code{+} but | |
19533 subtracts instead of adds; the @code{-} function subtracts its second | |
19534 argument from its first. Also, we already know what the @code{length} | |
19535 function does (@pxref{length}). It returns the length of a list. | |
19536 | |
19537 And @code{n} is the name of the required argument to the | |
19538 @code{current-kill} function. | |
19539 | |
19540 @need 1250 | |
19541 So when the first argument to @code{nthcdr} is zero, the @code{nthcdr} | |
19542 expression returns the whole list, as you can see by evaluating the | |
19543 following: | |
19544 | |
19545 @smallexample | |
19546 @group | |
19547 ;; kill-ring-yank-pointer @r{and} kill-ring @r{have a length of four} | |
19548 ;; @r{and} (mod (- 0 4) 4) @result{} 0 | |
19549 (nthcdr (mod (- 0 4) 4) | |
19550 '("fourth line of text" | |
19551 "third line" | |
19552 "second piece of text" | |
19553 "first some text")) | |
19554 @end group | |
19555 @end smallexample | |
19556 | |
19557 @need 1250 | |
19558 When the first argument to the @code{current-kill} function is one, | |
19559 the @code{nthcdr} expression returns the list without its first | |
19560 element. | |
19561 | |
19562 @smallexample | |
19563 @group | |
19564 (nthcdr (mod (- 1 4) 4) | |
19565 '("fourth line of text" | |
19566 "third line" | |
19567 "second piece of text" | |
19568 "first some text")) | |
19569 @end group | |
19570 @end smallexample | |
19571 | |
19572 @cindex @samp{global variable} defined | |
19573 @cindex @samp{variable, global}, defined | |
19574 Incidentally, both @code{kill-ring} and @code{kill-ring-yank-pointer} | |
19575 are @dfn{global variables}. That means that any expression in Emacs | |
19576 Lisp can access them. They are not like the local variables set by | |
19577 @code{let} or like the symbols in an argument list. | |
19578 Local variables can only be accessed | |
19579 within the @code{let} that defines them or the function that specifies | |
19580 them in an argument list (and within expressions called by them). | |
19581 | |
19582 @ignore | |
19583 @c texi2dvi fails when the name of the section is within ifnottex ... | |
19584 (@xref{Prevent confusion, , @code{let} Prevents Confusion}, and | |
19585 @ref{defun, , The @code{defun} Special Form}.) | |
19586 @end ignore | |
19587 | |
19588 @node yank, yank-pop, current-kill, Kill Ring | |
19589 @comment node-name, next, previous, up | |
19590 @appendixsec @code{yank} | |
19591 @findex yank | |
19592 | |
19593 After learning about @code{current-kill}, the code for the | |
19594 @code{yank} function is almost easy. | |
19595 | |
19596 The @code{yank} function does not use the | |
19597 @code{kill-ring-yank-pointer} variable directly. It calls | |
19598 @code{insert-for-yank} which calls @code{current-kill} which sets the | |
19599 @code{kill-ring-yank-pointer} variable. | |
19600 | |
19601 @need 1250 | |
19602 The code looks like this: | |
19603 | |
19604 @c in GNU Emacs 22 | |
19605 @smallexample | |
19606 @group | |
19607 (defun yank (&optional arg) | |
19608 "Reinsert (\"paste\") the last stretch of killed text. | |
19609 More precisely, reinsert the stretch of killed text most recently | |
19610 killed OR yanked. Put point at end, and set mark at beginning. | |
19611 With just \\[universal-argument] as argument, same but put point at | |
19612 beginning (and mark at end). With argument N, reinsert the Nth most | |
19613 recently killed stretch of killed text. | |
19614 | |
19615 When this command inserts killed text into the buffer, it honors | |
19616 `yank-excluded-properties' and `yank-handler' as described in the | |
19617 doc string for `insert-for-yank-1', which see. | |
19618 | |
19619 See also the command \\[yank-pop]." | |
19620 @end group | |
19621 @group | |
19622 (interactive "*P") | |
19623 (setq yank-window-start (window-start)) | |
19624 ;; If we don't get all the way thru, make last-command indicate that | |
19625 ;; for the following command. | |
19626 (setq this-command t) | |
19627 (push-mark (point)) | |
19628 @end group | |
19629 @group | |
19630 (insert-for-yank (current-kill (cond | |
19631 ((listp arg) 0) | |
19632 ((eq arg '-) -2) | |
19633 (t (1- arg))))) | |
19634 (if (consp arg) | |
19635 ;; This is like exchange-point-and-mark, | |
19636 ;; but doesn't activate the mark. | |
19637 ;; It is cleaner to avoid activation, even though the command | |
19638 ;; loop would deactivate the mark because we inserted text. | |
19639 (goto-char (prog1 (mark t) | |
19640 (set-marker (mark-marker) (point) (current-buffer))))) | |
19641 @end group | |
19642 @group | |
19643 ;; If we do get all the way thru, make this-command indicate that. | |
19644 (if (eq this-command t) | |
19645 (setq this-command 'yank)) | |
19646 nil) | |
19647 @end group | |
19648 @end smallexample | |
19649 | |
19650 The key expression is @code{insert-for-yank}, which inserts the string | |
19651 returned by @code{current-kill}, but removes some text properties from | |
19652 it. | |
19653 | |
19654 However, before getting to that expression, the function sets the value | |
19655 of @code{yank-window-start} to the position returned by the | |
19656 @code{(window-start)} expression, the position at which the display | |
19657 currently starts. The @code{yank} function also sets | |
19658 @code{this-command} and pushes the mark. | |
19659 | |
19660 After it yanks the appropriate element, if the optional argument is a | |
19661 @sc{cons} rather than a number or nothing, it puts point at beginning | |
19662 of the yanked text and mark at its end. | |
19663 | |
19664 (The @code{prog1} function is like @code{progn} but returns the value | |
19665 of its first argument rather than the value of its last argument. Its | |
19666 first argument is forced to return the buffer's mark as an integer. | |
19667 You can see the documentation for these functions by placing point | |
19668 over them in this buffer and then typing @kbd{C-h f} | |
19669 (@code{describe-function}) followed by a @kbd{RET}; the default is the | |
19670 function.) | |
19671 | |
19672 The last part of the function tells what to do when it succeeds. | |
19673 | |
19674 @node yank-pop, ring file, yank, Kill Ring | |
19675 @comment node-name, next, previous, up | |
19676 @appendixsec @code{yank-pop} | |
19677 @findex yank-pop | |
19678 | |
19679 After understanding @code{yank} and @code{current-kill}, you know how | |
19680 to approach the @code{yank-pop} function. Leaving out the | |
19681 documentation to save space, it looks like this: | |
19682 | |
19683 @c GNU Emacs 22 | |
19684 @smallexample | |
19685 @group | |
19686 (defun yank-pop (&optional arg) | |
19687 "@dots{}" | |
19688 (interactive "*p") | |
19689 (if (not (eq last-command 'yank)) | |
19690 (error "Previous command was not a yank")) | |
19691 @end group | |
19692 @group | |
19693 (setq this-command 'yank) | |
19694 (unless arg (setq arg 1)) | |
19695 (let ((inhibit-read-only t) | |
19696 (before (< (point) (mark t)))) | |
19697 @end group | |
19698 @group | |
19699 (if before | |
19700 (funcall (or yank-undo-function 'delete-region) (point) (mark t)) | |
19701 (funcall (or yank-undo-function 'delete-region) (mark t) (point))) | |
19702 (setq yank-undo-function nil) | |
19703 @end group | |
19704 @group | |
19705 (set-marker (mark-marker) (point) (current-buffer)) | |
19706 (insert-for-yank (current-kill arg)) | |
19707 ;; Set the window start back where it was in the yank command, | |
19708 ;; if possible. | |
19709 (set-window-start (selected-window) yank-window-start t) | |
19710 @end group | |
19711 @group | |
19712 (if before | |
19713 ;; This is like exchange-point-and-mark, | |
19714 ;; but doesn't activate the mark. | |
19715 ;; It is cleaner to avoid activation, even though the command | |
19716 ;; loop would deactivate the mark because we inserted text. | |
19717 (goto-char (prog1 (mark t) | |
19718 (set-marker (mark-marker) | |
19719 (point) | |
19720 (current-buffer)))))) | |
19721 nil) | |
19722 @end group | |
19723 @end smallexample | |
19724 | |
19725 The function is interactive with a small @samp{p} so the prefix | |
19726 argument is processed and passed to the function. The command can | |
19727 only be used after a previous yank; otherwise an error message is | |
19728 sent. This check uses the variable @code{last-command} which is set | |
19729 by @code{yank} and is discussed elsewhere. | |
19730 (@xref{copy-region-as-kill}.) | |
19731 | |
19732 The @code{let} clause sets the variable @code{before} to true or false | |
19733 depending whether point is before or after mark and then the region | |
19734 between point and mark is deleted. This is the region that was just | |
19735 inserted by the previous yank and it is this text that will be | |
19736 replaced. | |
19737 | |
19738 @code{funcall} calls its first argument as a function, passing | |
19739 remaining arguments to it. The first argument is whatever the | |
19740 @code{or} expression returns. The two remaining arguments are the | |
19741 positions of point and mark set by the preceding @code{yank} command. | |
19742 | |
19743 There is more, but that is the hardest part. | |
19744 | |
19745 @node ring file, , yank-pop, Kill Ring | |
19746 @comment node-name, next, previous, up | |
19747 @appendixsec The @file{ring.el} File | |
19748 @cindex @file{ring.el} file | |
19749 | |
19750 Interestingly, GNU Emacs posses a file called @file{ring.el} that | |
19751 provides many of the features we just discussed. But functions such | |
19752 as @code{kill-ring-yank-pointer} do not use this library, possibly | |
19753 because they were written earlier. | |
19754 | |
19755 @node Full Graph, Free Software and Free Manuals, Kill Ring, Top | |
19756 @appendix A Graph with Labelled Axes | |
19757 | |
19758 Printed axes help you understand a graph. They convey scale. In an | |
19759 earlier chapter (@pxref{Readying a Graph, , Readying a Graph}), we | |
19760 wrote the code to print the body of a graph. Here we write the code | |
19761 for printing and labelling vertical and horizontal axes, along with the | |
19762 body itself. | |
19763 | |
19764 @menu | |
19765 * Labelled Example:: | |
19766 * print-graph Varlist:: @code{let} expression in @code{print-graph}. | |
19767 * print-Y-axis:: Print a label for the vertical axis. | |
19768 * print-X-axis:: Print a horizontal label. | |
19769 * Print Whole Graph:: The function to print a complete graph. | |
19770 @end menu | |
19771 | |
19772 @node Labelled Example, print-graph Varlist, Full Graph, Full Graph | |
19773 @ifnottex | |
19774 @unnumberedsec Labelled Example Graph | |
19775 @end ifnottex | |
19776 | |
19777 Since insertions fill a buffer to the right and below point, the new | |
19778 graph printing function should first print the Y or vertical axis, | |
19779 then the body of the graph, and finally the X or horizontal axis. | |
19780 This sequence lays out for us the contents of the function: | |
19781 | |
19782 @enumerate | |
19783 @item | |
19784 Set up code. | |
19785 | |
19786 @item | |
19787 Print Y axis. | |
19788 | |
19789 @item | |
19790 Print body of graph. | |
19791 | |
19792 @item | |
19793 Print X axis. | |
19794 @end enumerate | |
19795 | |
19796 @need 800 | |
19797 Here is an example of how a finished graph should look: | |
19798 | |
19799 @smallexample | |
19800 @group | |
19801 10 - | |
19802 * | |
19803 * * | |
19804 * ** | |
19805 * *** | |
19806 5 - * ******* | |
19807 * *** ******* | |
19808 ************* | |
19809 *************** | |
19810 1 - **************** | |
19811 | | | | | |
19812 1 5 10 15 | |
19813 @end group | |
19814 @end smallexample | |
19815 | |
19816 @noindent | |
19817 In this graph, both the vertical and the horizontal axes are labelled | |
19818 with numbers. However, in some graphs, the horizontal axis is time | |
19819 and would be better labelled with months, like this: | |
19820 | |
19821 @smallexample | |
19822 @group | |
19823 5 - * | |
19824 * ** * | |
19825 ******* | |
19826 ********** ** | |
19827 1 - ************** | |
19828 | ^ | | |
19829 Jan June Jan | |
19830 @end group | |
19831 @end smallexample | |
19832 | |
19833 Indeed, with a little thought, we can easily come up with a variety of | |
19834 vertical and horizontal labelling schemes. Our task could become | |
19835 complicated. But complications breed confusion. Rather than permit | |
19836 this, it is better choose a simple labelling scheme for our first | |
19837 effort, and to modify or replace it later. | |
19838 | |
19839 @need 1200 | |
19840 These considerations suggest the following outline for the | |
19841 @code{print-graph} function: | |
19842 | |
19843 @smallexample | |
19844 @group | |
19845 (defun print-graph (numbers-list) | |
19846 "@var{documentation}@dots{}" | |
19847 (let ((height @dots{} | |
19848 @dots{})) | |
19849 @end group | |
19850 @group | |
19851 (print-Y-axis height @dots{} ) | |
19852 (graph-body-print numbers-list) | |
19853 (print-X-axis @dots{} ))) | |
19854 @end group | |
19855 @end smallexample | |
19856 | |
19857 We can work on each part of the @code{print-graph} function definition | |
19858 in turn. | |
19859 | |
19860 @node print-graph Varlist, print-Y-axis, Labelled Example, Full Graph | |
19861 @comment node-name, next, previous, up | |
19862 @appendixsec The @code{print-graph} Varlist | |
19863 @cindex @code{print-graph} varlist | |
19864 | |
19865 In writing the @code{print-graph} function, the first task is to write | |
19866 the varlist in the @code{let} expression. (We will leave aside for the | |
19867 moment any thoughts about making the function interactive or about the | |
19868 contents of its documentation string.) | |
19869 | |
19870 The varlist should set several values. Clearly, the top of the label | |
19871 for the vertical axis must be at least the height of the graph, which | |
19872 means that we must obtain this information here. Note that the | |
19873 @code{print-graph-body} function also requires this information. There | |
19874 is no reason to calculate the height of the graph in two different | |
19875 places, so we should change @code{print-graph-body} from the way we | |
19876 defined it earlier to take advantage of the calculation. | |
19877 | |
19878 Similarly, both the function for printing the X axis labels and the | |
19879 @code{print-graph-body} function need to learn the value of the width of | |
19880 each symbol. We can perform the calculation here and change the | |
19881 definition for @code{print-graph-body} from the way we defined it in the | |
19882 previous chapter. | |
19883 | |
19884 The length of the label for the horizontal axis must be at least as long | |
19885 as the graph. However, this information is used only in the function | |
19886 that prints the horizontal axis, so it does not need to be calculated here. | |
19887 | |
19888 These thoughts lead us directly to the following form for the varlist | |
19889 in the @code{let} for @code{print-graph}: | |
19890 | |
19891 @smallexample | |
19892 @group | |
19893 (let ((height (apply 'max numbers-list)) ; @r{First version.} | |
19894 (symbol-width (length graph-blank))) | |
19895 @end group | |
19896 @end smallexample | |
19897 | |
19898 @noindent | |
19899 As we shall see, this expression is not quite right. | |
19900 | |
19901 @need 2000 | |
19902 @node print-Y-axis, print-X-axis, print-graph Varlist, Full Graph | |
19903 @comment node-name, next, previous, up | |
19904 @appendixsec The @code{print-Y-axis} Function | |
19905 @cindex Axis, print vertical | |
19906 @cindex Y axis printing | |
19907 @cindex Vertical axis printing | |
19908 @cindex Print vertical axis | |
19909 | |
19910 The job of the @code{print-Y-axis} function is to print a label for | |
19911 the vertical axis that looks like this: | |
19912 | |
19913 @smallexample | |
19914 @group | |
19915 10 - | |
19916 | |
19917 | |
19918 | |
19919 | |
19920 5 - | |
19921 | |
19922 | |
19923 | |
19924 1 - | |
19925 @end group | |
19926 @end smallexample | |
19927 | |
19928 @noindent | |
19929 The function should be passed the height of the graph, and then should | |
19930 construct and insert the appropriate numbers and marks. | |
19931 | |
19932 @menu | |
19933 * print-Y-axis in Detail:: | |
19934 * Height of label:: What height for the Y axis? | |
19935 * Compute a Remainder:: How to compute the remainder of a division. | |
19936 * Y Axis Element:: Construct a line for the Y axis. | |
19937 * Y-axis-column:: Generate a list of Y axis labels. | |
19938 * print-Y-axis Penultimate:: A not quite final version. | |
19939 @end menu | |
19940 | |
19941 @node print-Y-axis in Detail, Height of label, print-Y-axis, print-Y-axis | |
19942 @ifnottex | |
19943 @unnumberedsubsec The @code{print-Y-axis} Function in Detail | |
19944 @end ifnottex | |
19945 | |
19946 It is easy enough to see in the figure what the Y axis label should | |
19947 look like; but to say in words, and then to write a function | |
19948 definition to do the job is another matter. It is not quite true to | |
19949 say that we want a number and a tic every five lines: there are only | |
19950 three lines between the @samp{1} and the @samp{5} (lines 2, 3, and 4), | |
19951 but four lines between the @samp{5} and the @samp{10} (lines 6, 7, 8, | |
19952 and 9). It is better to say that we want a number and a tic mark on | |
19953 the base line (number 1) and then that we want a number and a tic on | |
19954 the fifth line from the bottom and on every line that is a multiple of | |
19955 five. | |
19956 | |
19957 @node Height of label, Compute a Remainder, print-Y-axis in Detail, print-Y-axis | |
19958 @ifnottex | |
19959 @unnumberedsubsec What height should the label be? | |
19960 @end ifnottex | |
19961 | |
19962 The next issue is what height the label should be? Suppose the maximum | |
19963 height of tallest column of the graph is seven. Should the highest | |
19964 label on the Y axis be @samp{5 -}, and should the graph stick up above | |
19965 the label? Or should the highest label be @samp{7 -}, and mark the peak | |
19966 of the graph? Or should the highest label be @code{10 -}, which is a | |
19967 multiple of five, and be higher than the topmost value of the graph? | |
19968 | |
19969 The latter form is preferred. Most graphs are drawn within rectangles | |
19970 whose sides are an integral number of steps long---5, 10, 15, and so | |
19971 on for a step distance of five. But as soon as we decide to use a | |
19972 step height for the vertical axis, we discover that the simple | |
19973 expression in the varlist for computing the height is wrong. The | |
19974 expression is @code{(apply 'max numbers-list)}. This returns the | |
19975 precise height, not the maximum height plus whatever is necessary to | |
19976 round up to the nearest multiple of five. A more complex expression | |
19977 is required. | |
19978 | |
19979 As usual in cases like this, a complex problem becomes simpler if it is | |
19980 divided into several smaller problems. | |
19981 | |
19982 First, consider the case when the highest value of the graph is an | |
19983 integral multiple of five---when it is 5, 10, 15, or some higher | |
19984 multiple of five. We can use this value as the Y axis height. | |
19985 | |
19986 A fairly simply way to determine whether a number is a multiple of | |
19987 five is to divide it by five and see if the division results in a | |
19988 remainder. If there is no remainder, the number is a multiple of | |
19989 five. Thus, seven divided by five has a remainder of two, and seven | |
19990 is not an integral multiple of five. Put in slightly different | |
19991 language, more reminiscent of the classroom, five goes into seven | |
19992 once, with a remainder of two. However, five goes into ten twice, | |
19993 with no remainder: ten is an integral multiple of five. | |
19994 | |
19995 @node Compute a Remainder, Y Axis Element, Height of label, print-Y-axis | |
19996 @appendixsubsec Side Trip: Compute a Remainder | |
19997 | |
19998 @findex % @r{(remainder function)} | |
19999 @cindex Remainder function, @code{%} | |
20000 In Lisp, the function for computing a remainder is @code{%}. The | |
20001 function returns the remainder of its first argument divided by its | |
20002 second argument. As it happens, @code{%} is a function in Emacs Lisp | |
20003 that you cannot discover using @code{apropos}: you find nothing if you | |
20004 type @kbd{M-x apropos @key{RET} remainder @key{RET}}. The only way to | |
20005 learn of the existence of @code{%} is to read about it in a book such | |
20006 as this or in the Emacs Lisp sources. | |
20007 | |
20008 You can try the @code{%} function by evaluating the following two | |
20009 expressions: | |
20010 | |
20011 @smallexample | |
20012 @group | |
20013 (% 7 5) | |
20014 | |
20015 (% 10 5) | |
20016 @end group | |
20017 @end smallexample | |
20018 | |
20019 @noindent | |
20020 The first expression returns 2 and the second expression returns 0. | |
20021 | |
20022 To test whether the returned value is zero or some other number, we | |
20023 can use the @code{zerop} function. This function returns @code{t} if | |
20024 its argument, which must be a number, is zero. | |
20025 | |
20026 @smallexample | |
20027 @group | |
20028 (zerop (% 7 5)) | |
20029 @result{} nil | |
20030 | |
20031 (zerop (% 10 5)) | |
20032 @result{} t | |
20033 @end group | |
20034 @end smallexample | |
20035 | |
20036 Thus, the following expression will return @code{t} if the height | |
20037 of the graph is evenly divisible by five: | |
20038 | |
20039 @smallexample | |
20040 (zerop (% height 5)) | |
20041 @end smallexample | |
20042 | |
20043 @noindent | |
20044 (The value of @code{height}, of course, can be found from @code{(apply | |
20045 'max numbers-list)}.) | |
20046 | |
20047 On the other hand, if the value of @code{height} is not a multiple of | |
20048 five, we want to reset the value to the next higher multiple of five. | |
20049 This is straightforward arithmetic using functions with which we are | |
20050 already familiar. First, we divide the value of @code{height} by five | |
20051 to determine how many times five goes into the number. Thus, five | |
20052 goes into twelve twice. If we add one to this quotient and multiply by | |
20053 five, we will obtain the value of the next multiple of five that is | |
20054 larger than the height. Five goes into twelve twice. Add one to two, | |
20055 and multiply by five; the result is fifteen, which is the next multiple | |
20056 of five that is higher than twelve. The Lisp expression for this is: | |
20057 | |
20058 @smallexample | |
20059 (* (1+ (/ height 5)) 5) | |
20060 @end smallexample | |
20061 | |
20062 @noindent | |
20063 For example, if you evaluate the following, the result is 15: | |
20064 | |
20065 @smallexample | |
20066 (* (1+ (/ 12 5)) 5) | |
20067 @end smallexample | |
20068 | |
20069 All through this discussion, we have been using `five' as the value | |
20070 for spacing labels on the Y axis; but we may want to use some other | |
20071 value. For generality, we should replace `five' with a variable to | |
20072 which we can assign a value. The best name I can think of for this | |
20073 variable is @code{Y-axis-label-spacing}. | |
20074 | |
20075 @need 1250 | |
20076 Using this term, and an @code{if} expression, we produce the | |
20077 following: | |
20078 | |
20079 @smallexample | |
20080 @group | |
20081 (if (zerop (% height Y-axis-label-spacing)) | |
20082 height | |
20083 ;; @r{else} | |
20084 (* (1+ (/ height Y-axis-label-spacing)) | |
20085 Y-axis-label-spacing)) | |
20086 @end group | |
20087 @end smallexample | |
20088 | |
20089 @noindent | |
20090 This expression returns the value of @code{height} itself if the height | |
20091 is an even multiple of the value of the @code{Y-axis-label-spacing} or | |
20092 else it computes and returns a value of @code{height} that is equal to | |
20093 the next higher multiple of the value of the @code{Y-axis-label-spacing}. | |
20094 | |
20095 We can now include this expression in the @code{let} expression of the | |
20096 @code{print-graph} function (after first setting the value of | |
20097 @code{Y-axis-label-spacing}): | |
20098 @vindex Y-axis-label-spacing | |
20099 | |
20100 @smallexample | |
20101 @group | |
20102 (defvar Y-axis-label-spacing 5 | |
20103 "Number of lines from one Y axis label to next.") | |
20104 @end group | |
20105 | |
20106 @group | |
20107 @dots{} | |
20108 (let* ((height (apply 'max numbers-list)) | |
20109 (height-of-top-line | |
20110 (if (zerop (% height Y-axis-label-spacing)) | |
20111 height | |
20112 @end group | |
20113 @group | |
20114 ;; @r{else} | |
20115 (* (1+ (/ height Y-axis-label-spacing)) | |
20116 Y-axis-label-spacing))) | |
20117 (symbol-width (length graph-blank)))) | |
20118 @dots{} | |
20119 @end group | |
20120 @end smallexample | |
20121 | |
20122 @noindent | |
20123 (Note use of the @code{let*} function: the initial value of height is | |
20124 computed once by the @code{(apply 'max numbers-list)} expression and | |
20125 then the resulting value of @code{height} is used to compute its | |
20126 final value. @xref{fwd-para let, , The @code{let*} expression}, for | |
20127 more about @code{let*}.) | |
20128 | |
20129 @node Y Axis Element, Y-axis-column, Compute a Remainder, print-Y-axis | |
20130 @appendixsubsec Construct a Y Axis Element | |
20131 | |
20132 When we print the vertical axis, we want to insert strings such as | |
20133 @w{@samp{5 -}} and @w{@samp{10 - }} every five lines. | |
20134 Moreover, we want the numbers and dashes to line up, so shorter | |
20135 numbers must be padded with leading spaces. If some of the strings | |
20136 use two digit numbers, the strings with single digit numbers must | |
20137 include a leading blank space before the number. | |
20138 | |
20139 @findex number-to-string | |
20140 To figure out the length of the number, the @code{length} function is | |
20141 used. But the @code{length} function works only with a string, not with | |
20142 a number. So the number has to be converted from being a number to | |
20143 being a string. This is done with the @code{number-to-string} function. | |
20144 For example, | |
20145 | |
20146 @smallexample | |
20147 @group | |
20148 (length (number-to-string 35)) | |
20149 @result{} 2 | |
20150 | |
20151 (length (number-to-string 100)) | |
20152 @result{} 3 | |
20153 @end group | |
20154 @end smallexample | |
20155 | |
20156 @noindent | |
20157 (@code{number-to-string} is also called @code{int-to-string}; you will | |
20158 see this alternative name in various sources.) | |
20159 | |
20160 In addition, in each label, each number is followed by a string such | |
20161 as @w{@samp{ - }}, which we will call the @code{Y-axis-tic} marker. | |
20162 This variable is defined with @code{defvar}: | |
20163 | |
20164 @vindex Y-axis-tic | |
20165 @smallexample | |
20166 @group | |
20167 (defvar Y-axis-tic " - " | |
20168 "String that follows number in a Y axis label.") | |
20169 @end group | |
20170 @end smallexample | |
20171 | |
20172 The length of the Y label is the sum of the length of the Y axis tic | |
20173 mark and the length of the number of the top of the graph. | |
20174 | |
20175 @smallexample | |
20176 (length (concat (number-to-string height) Y-axis-tic))) | |
20177 @end smallexample | |
20178 | |
20179 This value will be calculated by the @code{print-graph} function in | |
20180 its varlist as @code{full-Y-label-width} and passed on. (Note that we | |
20181 did not think to include this in the varlist when we first proposed it.) | |
20182 | |
20183 To make a complete vertical axis label, a tic mark is concatenated | |
20184 with a number; and the two together may be preceded by one or more | |
20185 spaces depending on how long the number is. The label consists of | |
20186 three parts: the (optional) leading spaces, the number, and the tic | |
20187 mark. The function is passed the value of the number for the specific | |
20188 row, and the value of the width of the top line, which is calculated | |
20189 (just once) by @code{print-graph}. | |
20190 | |
20191 @smallexample | |
20192 @group | |
20193 (defun Y-axis-element (number full-Y-label-width) | |
20194 "Construct a NUMBERed label element. | |
20195 A numbered element looks like this ` 5 - ', | |
20196 and is padded as needed so all line up with | |
20197 the element for the largest number." | |
20198 @end group | |
20199 @group | |
20200 (let* ((leading-spaces | |
20201 (- full-Y-label-width | |
20202 (length | |
20203 (concat (number-to-string number) | |
20204 Y-axis-tic))))) | |
20205 @end group | |
20206 @group | |
20207 (concat | |
20208 (make-string leading-spaces ? ) | |
20209 (number-to-string number) | |
20210 Y-axis-tic))) | |
20211 @end group | |
20212 @end smallexample | |
20213 | |
20214 The @code{Y-axis-element} function concatenates together the leading | |
20215 spaces, if any; the number, as a string; and the tic mark. | |
20216 | |
20217 To figure out how many leading spaces the label will need, the | |
20218 function subtracts the actual length of the label---the length of the | |
20219 number plus the length of the tic mark---from the desired label width. | |
20220 | |
20221 @findex make-string | |
20222 Blank spaces are inserted using the @code{make-string} function. This | |
20223 function takes two arguments: the first tells it how long the string | |
20224 will be and the second is a symbol for the character to insert, in a | |
20225 special format. The format is a question mark followed by a blank | |
20226 space, like this, @samp{? }. @xref{Character Type, , Character Type, | |
20227 elisp, The GNU Emacs Lisp Reference Manual}, for a description of the | |
20228 syntax for characters. (Of course, you might want to replace the | |
20229 blank space by some other character @dots{} You know what to do.) | |
20230 | |
20231 The @code{number-to-string} function is used in the concatenation | |
20232 expression, to convert the number to a string that is concatenated | |
20233 with the leading spaces and the tic mark. | |
20234 | |
20235 @node Y-axis-column, print-Y-axis Penultimate, Y Axis Element, print-Y-axis | |
20236 @appendixsubsec Create a Y Axis Column | |
20237 | |
20238 The preceding functions provide all the tools needed to construct a | |
20239 function that generates a list of numbered and blank strings to insert | |
20240 as the label for the vertical axis: | |
20241 | |
20242 @findex Y-axis-column | |
20243 @smallexample | |
20244 @group | |
20245 (defun Y-axis-column (height width-of-label) | |
20246 "Construct list of Y axis labels and blank strings. | |
20247 For HEIGHT of line above base and WIDTH-OF-LABEL." | |
20248 (let (Y-axis) | |
20249 @group | |
20250 @end group | |
20251 (while (> height 1) | |
20252 (if (zerop (% height Y-axis-label-spacing)) | |
20253 ;; @r{Insert label.} | |
20254 (setq Y-axis | |
20255 (cons | |
20256 (Y-axis-element height width-of-label) | |
20257 Y-axis)) | |
20258 @group | |
20259 @end group | |
20260 ;; @r{Else, insert blanks.} | |
20261 (setq Y-axis | |
20262 (cons | |
20263 (make-string width-of-label ? ) | |
20264 Y-axis))) | |
20265 (setq height (1- height))) | |
20266 ;; @r{Insert base line.} | |
20267 (setq Y-axis | |
20268 (cons (Y-axis-element 1 width-of-label) Y-axis)) | |
20269 (nreverse Y-axis))) | |
20270 @end group | |
20271 @end smallexample | |
20272 | |
20273 In this function, we start with the value of @code{height} and | |
20274 repetitively subtract one from its value. After each subtraction, we | |
20275 test to see whether the value is an integral multiple of the | |
20276 @code{Y-axis-label-spacing}. If it is, we construct a numbered label | |
20277 using the @code{Y-axis-element} function; if not, we construct a | |
20278 blank label using the @code{make-string} function. The base line | |
20279 consists of the number one followed by a tic mark. | |
20280 | |
20281 @need 2000 | |
20282 @node print-Y-axis Penultimate, , Y-axis-column, print-Y-axis | |
20283 @appendixsubsec The Not Quite Final Version of @code{print-Y-axis} | |
20284 | |
20285 The list constructed by the @code{Y-axis-column} function is passed to | |
20286 the @code{print-Y-axis} function, which inserts the list as a column. | |
20287 | |
20288 @findex print-Y-axis | |
20289 @smallexample | |
20290 @group | |
20291 (defun print-Y-axis (height full-Y-label-width) | |
20292 "Insert Y axis using HEIGHT and FULL-Y-LABEL-WIDTH. | |
20293 Height must be the maximum height of the graph. | |
20294 Full width is the width of the highest label element." | |
20295 ;; Value of height and full-Y-label-width | |
20296 ;; are passed by `print-graph'. | |
20297 @end group | |
20298 @group | |
20299 (let ((start (point))) | |
20300 (insert-rectangle | |
20301 (Y-axis-column height full-Y-label-width)) | |
20302 ;; @r{Place point ready for inserting graph.} | |
20303 (goto-char start) | |
20304 ;; @r{Move point forward by value of} full-Y-label-width | |
20305 (forward-char full-Y-label-width))) | |
20306 @end group | |
20307 @end smallexample | |
20308 | |
20309 The @code{print-Y-axis} uses the @code{insert-rectangle} function to | |
20310 insert the Y axis labels created by the @code{Y-axis-column} function. | |
20311 In addition, it places point at the correct position for printing the body of | |
20312 the graph. | |
20313 | |
20314 You can test @code{print-Y-axis}: | |
20315 | |
20316 @enumerate | |
20317 @item | |
20318 Install | |
20319 | |
20320 @smallexample | |
20321 @group | |
20322 Y-axis-label-spacing | |
20323 Y-axis-tic | |
20324 Y-axis-element | |
20325 Y-axis-column | |
20326 print-Y-axis | |
20327 @end group | |
20328 @end smallexample | |
20329 | |
20330 @item | |
20331 Copy the following expression: | |
20332 | |
20333 @smallexample | |
20334 (print-Y-axis 12 5) | |
20335 @end smallexample | |
20336 | |
20337 @item | |
20338 Switch to the @file{*scratch*} buffer and place the cursor where you | |
20339 want the axis labels to start. | |
20340 | |
20341 @item | |
20342 Type @kbd{M-:} (@code{eval-expression}). | |
20343 | |
20344 @item | |
20345 Yank the @code{graph-body-print} expression into the minibuffer | |
20346 with @kbd{C-y} (@code{yank)}. | |
20347 | |
20348 @item | |
20349 Press @key{RET} to evaluate the expression. | |
20350 @end enumerate | |
20351 | |
20352 Emacs will print labels vertically, the top one being @w{@samp{10 -@w{ | |
20353 }}}. (The @code{print-graph} function will pass the value of | |
20354 @code{height-of-top-line}, which in this case will end up as 15, | |
20355 thereby getting rid of what might appear as a bug.) | |
20356 | |
20357 @need 2000 | |
20358 @node print-X-axis, Print Whole Graph, print-Y-axis, Full Graph | |
20359 @appendixsec The @code{print-X-axis} Function | |
20360 @cindex Axis, print horizontal | |
20361 @cindex X axis printing | |
20362 @cindex Print horizontal axis | |
20363 @cindex Horizontal axis printing | |
20364 | |
20365 X axis labels are much like Y axis labels, except that the ticks are on a | |
20366 line above the numbers. Labels should look like this: | |
20367 | |
20368 @smallexample | |
20369 @group | |
20370 | | | | | |
20371 1 5 10 15 | |
20372 @end group | |
20373 @end smallexample | |
20374 | |
20375 The first tic is under the first column of the graph and is preceded by | |
20376 several blank spaces. These spaces provide room in rows above for the Y | |
20377 axis labels. The second, third, fourth, and subsequent ticks are all | |
20378 spaced equally, according to the value of @code{X-axis-label-spacing}. | |
20379 | |
20380 The second row of the X axis consists of numbers, preceded by several | |
20381 blank spaces and also separated according to the value of the variable | |
20382 @code{X-axis-label-spacing}. | |
20383 | |
20384 The value of the variable @code{X-axis-label-spacing} should itself be | |
20385 measured in units of @code{symbol-width}, since you may want to change | |
20386 the width of the symbols that you are using to print the body of the | |
20387 graph without changing the ways the graph is labelled. | |
20388 | |
20389 @menu | |
20390 * Similarities differences:: Much like @code{print-Y-axis}, but not exactly. | |
20391 * X Axis Tic Marks:: Create tic marks for the horizontal axis. | |
20392 @end menu | |
20393 | |
20394 @node Similarities differences, X Axis Tic Marks, print-X-axis, print-X-axis | |
20395 @ifnottex | |
20396 @unnumberedsubsec Similarities and differences | |
20397 @end ifnottex | |
20398 | |
20399 The @code{print-X-axis} function is constructed in more or less the | |
20400 same fashion as the @code{print-Y-axis} function except that it has | |
20401 two lines: the line of tic marks and the numbers. We will write a | |
20402 separate function to print each line and then combine them within the | |
20403 @code{print-X-axis} function. | |
20404 | |
20405 This is a three step process: | |
20406 | |
20407 @enumerate | |
20408 @item | |
20409 Write a function to print the X axis tic marks, @code{print-X-axis-tic-line}. | |
20410 | |
20411 @item | |
20412 Write a function to print the X numbers, @code{print-X-axis-numbered-line}. | |
20413 | |
20414 @item | |
20415 Write a function to print both lines, the @code{print-X-axis} function, | |
20416 using @code{print-X-axis-tic-line} and | |
20417 @code{print-X-axis-numbered-line}. | |
20418 @end enumerate | |
20419 | |
20420 @node X Axis Tic Marks, , Similarities differences, print-X-axis | |
20421 @appendixsubsec X Axis Tic Marks | |
20422 | |
20423 The first function should print the X axis tic marks. We must specify | |
20424 the tic marks themselves and their spacing: | |
20425 | |
20426 @smallexample | |
20427 @group | |
20428 (defvar X-axis-label-spacing | |
20429 (if (boundp 'graph-blank) | |
20430 (* 5 (length graph-blank)) 5) | |
20431 "Number of units from one X axis label to next.") | |
20432 @end group | |
20433 @end smallexample | |
20434 | |
20435 @noindent | |
20436 (Note that the value of @code{graph-blank} is set by another | |
20437 @code{defvar}. The @code{boundp} predicate checks whether it has | |
20438 already been set; @code{boundp} returns @code{nil} if it has not. If | |
20439 @code{graph-blank} were unbound and we did not use this conditional | |
20440 construction, in a recent GNU Emacs, we would enter the debugger and | |
20441 see an error message saying @samp{@w{Debugger entered--Lisp error:} | |
20442 @w{(void-variable graph-blank)}}.) | |
20443 | |
20444 @need 1200 | |
20445 Here is the @code{defvar} for @code{X-axis-tic-symbol}: | |
20446 | |
20447 @smallexample | |
20448 @group | |
20449 (defvar X-axis-tic-symbol "|" | |
20450 "String to insert to point to a column in X axis.") | |
20451 @end group | |
20452 @end smallexample | |
20453 | |
20454 @need 1250 | |
20455 The goal is to make a line that looks like this: | |
20456 | |
20457 @smallexample | |
20458 | | | | | |
20459 @end smallexample | |
20460 | |
20461 The first tic is indented so that it is under the first column, which is | |
20462 indented to provide space for the Y axis labels. | |
20463 | |
20464 A tic element consists of the blank spaces that stretch from one tic to | |
20465 the next plus a tic symbol. The number of blanks is determined by the | |
20466 width of the tic symbol and the @code{X-axis-label-spacing}. | |
20467 | |
20468 @need 1250 | |
20469 The code looks like this: | |
20470 | |
20471 @smallexample | |
20472 @group | |
20473 ;;; X-axis-tic-element | |
20474 @dots{} | |
20475 (concat | |
20476 (make-string | |
20477 ;; @r{Make a string of blanks.} | |
20478 (- (* symbol-width X-axis-label-spacing) | |
20479 (length X-axis-tic-symbol)) | |
20480 ? ) | |
20481 ;; @r{Concatenate blanks with tic symbol.} | |
20482 X-axis-tic-symbol) | |
20483 @dots{} | |
20484 @end group | |
20485 @end smallexample | |
20486 | |
20487 Next, we determine how many blanks are needed to indent the first tic | |
20488 mark to the first column of the graph. This uses the value of | |
20489 @code{full-Y-label-width} passed it by the @code{print-graph} function. | |
20490 | |
20491 @need 1250 | |
20492 The code to make @code{X-axis-leading-spaces} | |
20493 looks like this: | |
20494 | |
20495 @smallexample | |
20496 @group | |
20497 ;; X-axis-leading-spaces | |
20498 @dots{} | |
20499 (make-string full-Y-label-width ? ) | |
20500 @dots{} | |
20501 @end group | |
20502 @end smallexample | |
20503 | |
20504 We also need to determine the length of the horizontal axis, which is | |
20505 the length of the numbers list, and the number of ticks in the horizontal | |
20506 axis: | |
20507 | |
20508 @smallexample | |
20509 @group | |
20510 ;; X-length | |
20511 @dots{} | |
20512 (length numbers-list) | |
20513 @end group | |
20514 | |
20515 @group | |
20516 ;; tic-width | |
20517 @dots{} | |
20518 (* symbol-width X-axis-label-spacing) | |
20519 @end group | |
20520 | |
20521 @group | |
20522 ;; number-of-X-ticks | |
20523 (if (zerop (% (X-length tic-width))) | |
20524 (/ (X-length tic-width)) | |
20525 (1+ (/ (X-length tic-width)))) | |
20526 @end group | |
20527 @end smallexample | |
20528 | |
20529 @need 1250 | |
20530 All this leads us directly to the function for printing the X axis tic line: | |
20531 | |
20532 @findex print-X-axis-tic-line | |
20533 @smallexample | |
20534 @group | |
20535 (defun print-X-axis-tic-line | |
20536 (number-of-X-tics X-axis-leading-spaces X-axis-tic-element) | |
20537 "Print ticks for X axis." | |
20538 (insert X-axis-leading-spaces) | |
20539 (insert X-axis-tic-symbol) ; @r{Under first column.} | |
20540 @end group | |
20541 @group | |
20542 ;; @r{Insert second tic in the right spot.} | |
20543 (insert (concat | |
20544 (make-string | |
20545 (- (* symbol-width X-axis-label-spacing) | |
20546 ;; @r{Insert white space up to second tic symbol.} | |
20547 (* 2 (length X-axis-tic-symbol))) | |
20548 ? ) | |
20549 X-axis-tic-symbol)) | |
20550 @end group | |
20551 @group | |
20552 ;; @r{Insert remaining ticks.} | |
20553 (while (> number-of-X-tics 1) | |
20554 (insert X-axis-tic-element) | |
20555 (setq number-of-X-tics (1- number-of-X-tics)))) | |
20556 @end group | |
20557 @end smallexample | |
20558 | |
20559 The line of numbers is equally straightforward: | |
20560 | |
20561 @need 1250 | |
20562 First, we create a numbered element with blank spaces before each number: | |
20563 | |
20564 @findex X-axis-element | |
20565 @smallexample | |
20566 @group | |
20567 (defun X-axis-element (number) | |
20568 "Construct a numbered X axis element." | |
20569 (let ((leading-spaces | |
20570 (- (* symbol-width X-axis-label-spacing) | |
20571 (length (number-to-string number))))) | |
20572 (concat (make-string leading-spaces ? ) | |
20573 (number-to-string number)))) | |
20574 @end group | |
20575 @end smallexample | |
20576 | |
20577 Next, we create the function to print the numbered line, starting with | |
20578 the number ``1'' under the first column: | |
20579 | |
20580 @findex print-X-axis-numbered-line | |
20581 @smallexample | |
20582 @group | |
20583 (defun print-X-axis-numbered-line | |
20584 (number-of-X-tics X-axis-leading-spaces) | |
20585 "Print line of X-axis numbers" | |
20586 (let ((number X-axis-label-spacing)) | |
20587 (insert X-axis-leading-spaces) | |
20588 (insert "1") | |
20589 @end group | |
20590 @group | |
20591 (insert (concat | |
20592 (make-string | |
20593 ;; @r{Insert white space up to next number.} | |
20594 (- (* symbol-width X-axis-label-spacing) 2) | |
20595 ? ) | |
20596 (number-to-string number))) | |
20597 @end group | |
20598 @group | |
20599 ;; @r{Insert remaining numbers.} | |
20600 (setq number (+ number X-axis-label-spacing)) | |
20601 (while (> number-of-X-tics 1) | |
20602 (insert (X-axis-element number)) | |
20603 (setq number (+ number X-axis-label-spacing)) | |
20604 (setq number-of-X-tics (1- number-of-X-tics))))) | |
20605 @end group | |
20606 @end smallexample | |
20607 | |
20608 Finally, we need to write the @code{print-X-axis} that uses | |
20609 @code{print-X-axis-tic-line} and | |
20610 @code{print-X-axis-numbered-line}. | |
20611 | |
20612 The function must determine the local values of the variables used by both | |
20613 @code{print-X-axis-tic-line} and @code{print-X-axis-numbered-line}, and | |
20614 then it must call them. Also, it must print the carriage return that | |
20615 separates the two lines. | |
20616 | |
20617 The function consists of a varlist that specifies five local variables, | |
20618 and calls to each of the two line printing functions: | |
20619 | |
20620 @findex print-X-axis | |
20621 @smallexample | |
20622 @group | |
20623 (defun print-X-axis (numbers-list) | |
20624 "Print X axis labels to length of NUMBERS-LIST." | |
20625 (let* ((leading-spaces | |
20626 (make-string full-Y-label-width ? )) | |
20627 @end group | |
20628 @group | |
20629 ;; symbol-width @r{is provided by} graph-body-print | |
20630 (tic-width (* symbol-width X-axis-label-spacing)) | |
20631 (X-length (length numbers-list)) | |
20632 @end group | |
20633 @group | |
20634 (X-tic | |
20635 (concat | |
20636 (make-string | |
20637 @end group | |
20638 @group | |
20639 ;; @r{Make a string of blanks.} | |
20640 (- (* symbol-width X-axis-label-spacing) | |
20641 (length X-axis-tic-symbol)) | |
20642 ? ) | |
20643 @end group | |
20644 @group | |
20645 ;; @r{Concatenate blanks with tic symbol.} | |
20646 X-axis-tic-symbol)) | |
20647 @end group | |
20648 @group | |
20649 (tic-number | |
20650 (if (zerop (% X-length tic-width)) | |
20651 (/ X-length tic-width) | |
20652 (1+ (/ X-length tic-width))))) | |
20653 @end group | |
20654 @group | |
20655 (print-X-axis-tic-line tic-number leading-spaces X-tic) | |
20656 (insert "\n") | |
20657 (print-X-axis-numbered-line tic-number leading-spaces))) | |
20658 @end group | |
20659 @end smallexample | |
20660 | |
20661 @need 1250 | |
20662 You can test @code{print-X-axis}: | |
20663 | |
20664 @enumerate | |
20665 @item | |
20666 Install @code{X-axis-tic-symbol}, @code{X-axis-label-spacing}, | |
20667 @code{print-X-axis-tic-line}, as well as @code{X-axis-element}, | |
20668 @code{print-X-axis-numbered-line}, and @code{print-X-axis}. | |
20669 | |
20670 @item | |
20671 Copy the following expression: | |
20672 | |
20673 @smallexample | |
20674 @group | |
20675 (progn | |
20676 (let ((full-Y-label-width 5) | |
20677 (symbol-width 1)) | |
20678 (print-X-axis | |
20679 '(1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16)))) | |
20680 @end group | |
20681 @end smallexample | |
20682 | |
20683 @item | |
20684 Switch to the @file{*scratch*} buffer and place the cursor where you | |
20685 want the axis labels to start. | |
20686 | |
20687 @item | |
20688 Type @kbd{M-:} (@code{eval-expression}). | |
20689 | |
20690 @item | |
20691 Yank the test expression into the minibuffer | |
20692 with @kbd{C-y} (@code{yank)}. | |
20693 | |
20694 @item | |
20695 Press @key{RET} to evaluate the expression. | |
20696 @end enumerate | |
20697 | |
20698 @need 1250 | |
20699 Emacs will print the horizontal axis like this: | |
20700 @sp 1 | |
20701 | |
20702 @smallexample | |
20703 @group | |
20704 | | | | | | |
20705 1 5 10 15 20 | |
20706 @end group | |
20707 @end smallexample | |
20708 | |
20709 @node Print Whole Graph, , print-X-axis, Full Graph | |
20710 @appendixsec Printing the Whole Graph | |
20711 @cindex Printing the whole graph | |
20712 @cindex Whole graph printing | |
20713 @cindex Graph, printing all | |
20714 | |
20715 Now we are nearly ready to print the whole graph. | |
20716 | |
20717 The function to print the graph with the proper labels follows the | |
20718 outline we created earlier (@pxref{Full Graph, , A Graph with Labelled | |
20719 Axes}), but with additions. | |
20720 | |
20721 @need 1250 | |
20722 Here is the outline: | |
20723 | |
20724 @smallexample | |
20725 @group | |
20726 (defun print-graph (numbers-list) | |
20727 "@var{documentation}@dots{}" | |
20728 (let ((height @dots{} | |
20729 @dots{})) | |
20730 @end group | |
20731 @group | |
20732 (print-Y-axis height @dots{} ) | |
20733 (graph-body-print numbers-list) | |
20734 (print-X-axis @dots{} ))) | |
20735 @end group | |
20736 @end smallexample | |
20737 | |
20738 @menu | |
20739 * The final version:: A few changes. | |
20740 * Test print-graph:: Run a short test. | |
20741 * Graphing words in defuns:: Executing the final code. | |
20742 * lambda:: How to write an anonymous function. | |
20743 * mapcar:: Apply a function to elements of a list. | |
20744 * Another Bug:: Yet another bug @dots{} most insidious. | |
20745 * Final printed graph:: The graph itself! | |
20746 @end menu | |
20747 | |
20748 @node The final version, Test print-graph, Print Whole Graph, Print Whole Graph | |
20749 @ifnottex | |
20750 @unnumberedsubsec Changes for the Final Version | |
20751 @end ifnottex | |
20752 | |
20753 The final version is different from what we planned in two ways: | |
20754 first, it contains additional values calculated once in the varlist; | |
20755 second, it carries an option to specify the labels' increment per row. | |
20756 This latter feature turns out to be essential; otherwise, a graph may | |
20757 have more rows than fit on a display or on a sheet of paper. | |
20758 | |
20759 @need 1500 | |
20760 This new feature requires a change to the @code{Y-axis-column} | |
20761 function, to add @code{vertical-step} to it. The function looks like | |
20762 this: | |
20763 | |
20764 @findex Y-axis-column @r{Final version.} | |
20765 @smallexample | |
20766 @group | |
20767 ;;; @r{Final version.} | |
20768 (defun Y-axis-column | |
20769 (height width-of-label &optional vertical-step) | |
20770 "Construct list of labels for Y axis. | |
20771 HEIGHT is maximum height of graph. | |
20772 WIDTH-OF-LABEL is maximum width of label. | |
20773 VERTICAL-STEP, an option, is a positive integer | |
20774 that specifies how much a Y axis label increments | |
20775 for each line. For example, a step of 5 means | |
20776 that each line is five units of the graph." | |
20777 @end group | |
20778 @group | |
20779 (let (Y-axis | |
20780 (number-per-line (or vertical-step 1))) | |
20781 (while (> height 1) | |
20782 (if (zerop (% height Y-axis-label-spacing)) | |
20783 @end group | |
20784 @group | |
20785 ;; @r{Insert label.} | |
20786 (setq Y-axis | |
20787 (cons | |
20788 (Y-axis-element | |
20789 (* height number-per-line) | |
20790 width-of-label) | |
20791 Y-axis)) | |
20792 @end group | |
20793 @group | |
20794 ;; @r{Else, insert blanks.} | |
20795 (setq Y-axis | |
20796 (cons | |
20797 (make-string width-of-label ? ) | |
20798 Y-axis))) | |
20799 (setq height (1- height))) | |
20800 @end group | |
20801 @group | |
20802 ;; @r{Insert base line.} | |
20803 (setq Y-axis (cons (Y-axis-element | |
20804 (or vertical-step 1) | |
20805 width-of-label) | |
20806 Y-axis)) | |
20807 (nreverse Y-axis))) | |
20808 @end group | |
20809 @end smallexample | |
20810 | |
20811 The values for the maximum height of graph and the width of a symbol | |
20812 are computed by @code{print-graph} in its @code{let} expression; so | |
20813 @code{graph-body-print} must be changed to accept them. | |
20814 | |
20815 @findex graph-body-print @r{Final version.} | |
20816 @smallexample | |
20817 @group | |
20818 ;;; @r{Final version.} | |
20819 (defun graph-body-print (numbers-list height symbol-width) | |
20820 "Print a bar graph of the NUMBERS-LIST. | |
20821 The numbers-list consists of the Y-axis values. | |
20822 HEIGHT is maximum height of graph. | |
20823 SYMBOL-WIDTH is number of each column." | |
20824 @end group | |
20825 @group | |
20826 (let (from-position) | |
20827 (while numbers-list | |
20828 (setq from-position (point)) | |
20829 (insert-rectangle | |
20830 (column-of-graph height (car numbers-list))) | |
20831 (goto-char from-position) | |
20832 (forward-char symbol-width) | |
20833 @end group | |
20834 @group | |
20835 ;; @r{Draw graph column by column.} | |
20836 (sit-for 0) | |
20837 (setq numbers-list (cdr numbers-list))) | |
20838 ;; @r{Place point for X axis labels.} | |
20839 (forward-line height) | |
20840 (insert "\n"))) | |
20841 @end group | |
20842 @end smallexample | |
20843 | |
20844 @need 1250 | |
20845 Finally, the code for the @code{print-graph} function: | |
20846 | |
20847 @findex print-graph @r{Final version.} | |
20848 @smallexample | |
20849 @group | |
20850 ;;; @r{Final version.} | |
20851 (defun print-graph | |
20852 (numbers-list &optional vertical-step) | |
20853 "Print labelled bar graph of the NUMBERS-LIST. | |
20854 The numbers-list consists of the Y-axis values. | |
20855 @end group | |
20856 | |
20857 @group | |
20858 Optionally, VERTICAL-STEP, a positive integer, | |
20859 specifies how much a Y axis label increments for | |
20860 each line. For example, a step of 5 means that | |
20861 each row is five units." | |
20862 @end group | |
20863 @group | |
20864 (let* ((symbol-width (length graph-blank)) | |
20865 ;; @code{height} @r{is both the largest number} | |
20866 ;; @r{and the number with the most digits.} | |
20867 (height (apply 'max numbers-list)) | |
20868 @end group | |
20869 @group | |
20870 (height-of-top-line | |
20871 (if (zerop (% height Y-axis-label-spacing)) | |
20872 height | |
20873 ;; @r{else} | |
20874 (* (1+ (/ height Y-axis-label-spacing)) | |
20875 Y-axis-label-spacing))) | |
20876 @end group | |
20877 @group | |
20878 (vertical-step (or vertical-step 1)) | |
20879 (full-Y-label-width | |
20880 (length | |
20881 @end group | |
20882 @group | |
20883 (concat | |
20884 (number-to-string | |
20885 (* height-of-top-line vertical-step)) | |
20886 Y-axis-tic)))) | |
20887 @end group | |
20888 | |
20889 @group | |
20890 (print-Y-axis | |
20891 height-of-top-line full-Y-label-width vertical-step) | |
20892 @end group | |
20893 @group | |
20894 (graph-body-print | |
20895 numbers-list height-of-top-line symbol-width) | |
20896 (print-X-axis numbers-list))) | |
20897 @end group | |
20898 @end smallexample | |
20899 | |
20900 @node Test print-graph, Graphing words in defuns, The final version, Print Whole Graph | |
20901 @appendixsubsec Testing @code{print-graph} | |
20902 | |
20903 @need 1250 | |
20904 We can test the @code{print-graph} function with a short list of numbers: | |
20905 | |
20906 @enumerate | |
20907 @item | |
20908 Install the final versions of @code{Y-axis-column}, | |
20909 @code{graph-body-print}, and @code{print-graph} (in addition to the | |
20910 rest of the code.) | |
20911 | |
20912 @item | |
20913 Copy the following expression: | |
20914 | |
20915 @smallexample | |
20916 (print-graph '(3 2 5 6 7 5 3 4 6 4 3 2 1)) | |
20917 @end smallexample | |
20918 | |
20919 @item | |
20920 Switch to the @file{*scratch*} buffer and place the cursor where you | |
20921 want the axis labels to start. | |
20922 | |
20923 @item | |
20924 Type @kbd{M-:} (@code{eval-expression}). | |
20925 | |
20926 @item | |
20927 Yank the test expression into the minibuffer | |
20928 with @kbd{C-y} (@code{yank)}. | |
20929 | |
20930 @item | |
20931 Press @key{RET} to evaluate the expression. | |
20932 @end enumerate | |
20933 | |
20934 @need 1250 | |
20935 Emacs will print a graph that looks like this: | |
20936 | |
20937 @smallexample | |
20938 @group | |
20939 10 - | |
20940 | |
20941 | |
20942 * | |
20943 ** * | |
20944 5 - **** * | |
20945 **** *** | |
20946 * ********* | |
20947 ************ | |
20948 1 - ************* | |
20949 | |
20950 | | | | | |
20951 1 5 10 15 | |
20952 @end group | |
20953 @end smallexample | |
20954 | |
20955 @need 1200 | |
20956 On the other hand, if you pass @code{print-graph} a | |
20957 @code{vertical-step} value of 2, by evaluating this expression: | |
20958 | |
20959 @smallexample | |
20960 (print-graph '(3 2 5 6 7 5 3 4 6 4 3 2 1) 2) | |
20961 @end smallexample | |
20962 | |
20963 @need 1250 | |
20964 @noindent | |
20965 The graph looks like this: | |
20966 | |
20967 @smallexample | |
20968 @group | |
20969 20 - | |
20970 | |
20971 | |
20972 * | |
20973 ** * | |
20974 10 - **** * | |
20975 **** *** | |
20976 * ********* | |
20977 ************ | |
20978 2 - ************* | |
20979 | |
20980 | | | | | |
20981 1 5 10 15 | |
20982 @end group | |
20983 @end smallexample | |
20984 | |
20985 @noindent | |
20986 (A question: is the `2' on the bottom of the vertical axis a bug or a | |
20987 feature? If you think it is a bug, and should be a `1' instead, (or | |
20988 even a `0'), you can modify the sources.) | |
20989 | |
20990 @node Graphing words in defuns, lambda, Test print-graph, Print Whole Graph | |
20991 @appendixsubsec Graphing Numbers of Words and Symbols | |
20992 | |
20993 Now for the graph for which all this code was written: a graph that | |
20994 shows how many function definitions contain fewer than 10 words and | |
20995 symbols, how many contain between 10 and 19 words and symbols, how | |
20996 many contain between 20 and 29 words and symbols, and so on. | |
20997 | |
20998 This is a multi-step process. First make sure you have loaded all the | |
20999 requisite code. | |
21000 | |
21001 @need 1500 | |
21002 It is a good idea to reset the value of @code{top-of-ranges} in case | |
21003 you have set it to some different value. You can evaluate the | |
21004 following: | |
21005 | |
21006 @smallexample | |
21007 @group | |
21008 (setq top-of-ranges | |
21009 '(10 20 30 40 50 | |
21010 60 70 80 90 100 | |
21011 110 120 130 140 150 | |
21012 160 170 180 190 200 | |
21013 210 220 230 240 250 | |
21014 260 270 280 290 300) | |
21015 @end group | |
21016 @end smallexample | |
21017 | |
21018 @noindent | |
21019 Next create a list of the number of words and symbols in each range. | |
21020 | |
21021 @need 1500 | |
21022 @noindent | |
21023 Evaluate the following: | |
21024 | |
21025 @smallexample | |
21026 @group | |
21027 (setq list-for-graph | |
21028 (defuns-per-range | |
21029 (sort | |
21030 (recursive-lengths-list-many-files | |
21031 (directory-files "/usr/local/emacs/lisp" | |
21032 t ".+el$")) | |
21033 '<) | |
21034 top-of-ranges)) | |
21035 @end group | |
21036 @end smallexample | |
21037 | |
21038 @noindent | |
21039 On my old machine, this took about an hour. It looked though 303 Lisp | |
21040 files in my copy of Emacs version 19.23. After all that computing, | |
21041 the @code{list-for-graph} had this value: | |
21042 | |
21043 @smallexample | |
21044 @group | |
21045 (537 1027 955 785 594 483 349 292 224 199 166 120 116 99 | |
21046 90 80 67 48 52 45 41 33 28 26 25 20 12 28 11 13 220) | |
21047 @end group | |
21048 @end smallexample | |
21049 | |
21050 @noindent | |
21051 This means that my copy of Emacs had 537 function definitions with | |
21052 fewer than 10 words or symbols in them, 1,027 function definitions | |
21053 with 10 to 19 words or symbols in them, 955 function definitions with | |
21054 20 to 29 words or symbols in them, and so on. | |
21055 | |
21056 Clearly, just by looking at this list we can see that most function | |
21057 definitions contain ten to thirty words and symbols. | |
21058 | |
21059 Now for printing. We do @emph{not} want to print a graph that is | |
21060 1,030 lines high @dots{} Instead, we should print a graph that is | |
21061 fewer than twenty-five lines high. A graph that height can be | |
21062 displayed on almost any monitor, and easily printed on a sheet of paper. | |
21063 | |
21064 This means that each value in @code{list-for-graph} must be reduced to | |
21065 one-fiftieth its present value. | |
21066 | |
21067 Here is a short function to do just that, using two functions we have | |
21068 not yet seen, @code{mapcar} and @code{lambda}. | |
21069 | |
21070 @smallexample | |
21071 @group | |
21072 (defun one-fiftieth (full-range) | |
21073 "Return list, each number one-fiftieth of previous." | |
21074 (mapcar '(lambda (arg) (/ arg 50)) full-range)) | |
21075 @end group | |
21076 @end smallexample | |
21077 | |
21078 @node lambda, mapcar, Graphing words in defuns, Print Whole Graph | |
21079 @appendixsubsec A @code{lambda} Expression: Useful Anonymity | |
21080 @cindex Anonymous function | |
21081 @findex lambda | |
21082 | |
21083 @code{lambda} is the symbol for an anonymous function, a function | |
21084 without a name. Every time you use an anonymous function, you need to | |
21085 include its whole body. | |
21086 | |
21087 @need 1250 | |
21088 @noindent | |
21089 Thus, | |
21090 | |
21091 @smallexample | |
21092 (lambda (arg) (/ arg 50)) | |
21093 @end smallexample | |
21094 | |
21095 @noindent | |
21096 is a function definition that says `return the value resulting from | |
21097 dividing whatever is passed to me as @code{arg} by 50'. | |
21098 | |
21099 @need 1200 | |
21100 Earlier, for example, we had a function @code{multiply-by-seven}; it | |
21101 multiplied its argument by 7. This function is similar, except it | |
21102 divides its argument by 50; and, it has no name. The anonymous | |
21103 equivalent of @code{multiply-by-seven} is: | |
21104 | |
21105 @smallexample | |
21106 (lambda (number) (* 7 number)) | |
21107 @end smallexample | |
21108 | |
21109 @noindent | |
21110 (@xref{defun, , The @code{defun} Special Form}.) | |
21111 | |
21112 @need 1250 | |
21113 @noindent | |
21114 If we want to multiply 3 by 7, we can write: | |
21115 | |
21116 @c !!! Clear print-postscript-figures if the computer formatting this | |
21117 @c document is too small and cannot handle all the diagrams and figures. | |
21118 @c clear print-postscript-figures | |
21119 @c set print-postscript-figures | |
21120 @c lambda example diagram #1 | |
21121 @ifnottex | |
21122 @smallexample | |
21123 @group | |
21124 (multiply-by-seven 3) | |
21125 \_______________/ ^ | |
21126 | | | |
21127 function argument | |
21128 @end group | |
21129 @end smallexample | |
21130 @end ifnottex | |
21131 @ifset print-postscript-figures | |
21132 @sp 1 | |
21133 @tex | |
21134 @center @image{lambda-1} | |
21135 %%%% old method of including an image | |
21136 % \input /usr/local/lib/tex/inputs/psfig.tex | |
21137 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/lambda-1.eps}} | |
21138 % \catcode`\@=0 % | |
21139 @end tex | |
21140 @sp 1 | |
21141 @end ifset | |
21142 @ifclear print-postscript-figures | |
21143 @iftex | |
21144 @smallexample | |
21145 @group | |
21146 (multiply-by-seven 3) | |
21147 \_______________/ ^ | |
21148 | | | |
21149 function argument | |
21150 @end group | |
21151 @end smallexample | |
21152 @end iftex | |
21153 @end ifclear | |
21154 | |
21155 @noindent | |
21156 This expression returns 21. | |
21157 | |
21158 @need 1250 | |
21159 @noindent | |
21160 Similarly, we can write: | |
21161 | |
21162 @c lambda example diagram #2 | |
21163 @ifnottex | |
21164 @smallexample | |
21165 @group | |
21166 ((lambda (number) (* 7 number)) 3) | |
21167 \____________________________/ ^ | |
21168 | | | |
21169 anonymous function argument | |
21170 @end group | |
21171 @end smallexample | |
21172 @end ifnottex | |
21173 @ifset print-postscript-figures | |
21174 @sp 1 | |
21175 @tex | |
21176 @center @image{lambda-2} | |
21177 %%%% old method of including an image | |
21178 % \input /usr/local/lib/tex/inputs/psfig.tex | |
21179 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/lambda-2.eps}} | |
21180 % \catcode`\@=0 % | |
21181 @end tex | |
21182 @sp 1 | |
21183 @end ifset | |
21184 @ifclear print-postscript-figures | |
21185 @iftex | |
21186 @smallexample | |
21187 @group | |
21188 ((lambda (number) (* 7 number)) 3) | |
21189 \____________________________/ ^ | |
21190 | | | |
21191 anonymous function argument | |
21192 @end group | |
21193 @end smallexample | |
21194 @end iftex | |
21195 @end ifclear | |
21196 | |
21197 @need 1250 | |
21198 @noindent | |
21199 If we want to divide 100 by 50, we can write: | |
21200 | |
21201 @c lambda example diagram #3 | |
21202 @ifnottex | |
21203 @smallexample | |
21204 @group | |
21205 ((lambda (arg) (/ arg 50)) 100) | |
21206 \______________________/ \_/ | |
21207 | | | |
21208 anonymous function argument | |
21209 @end group | |
21210 @end smallexample | |
21211 @end ifnottex | |
21212 @ifset print-postscript-figures | |
21213 @sp 1 | |
21214 @tex | |
21215 @center @image{lambda-3} | |
21216 %%%% old method of including an image | |
21217 % \input /usr/local/lib/tex/inputs/psfig.tex | |
21218 % \centerline{\psfig{figure=/usr/local/lib/emacs/man/lambda-3.eps}} | |
21219 % \catcode`\@=0 % | |
21220 @end tex | |
21221 @sp 1 | |
21222 @end ifset | |
21223 @ifclear print-postscript-figures | |
21224 @iftex | |
21225 @smallexample | |
21226 @group | |
21227 ((lambda (arg) (/ arg 50)) 100) | |
21228 \______________________/ \_/ | |
21229 | | | |
21230 anonymous function argument | |
21231 @end group | |
21232 @end smallexample | |
21233 @end iftex | |
21234 @end ifclear | |
21235 | |
21236 @noindent | |
21237 This expression returns 2. The 100 is passed to the function, which | |
21238 divides that number by 50. | |
21239 | |
21240 @xref{Lambda Expressions, , Lambda Expressions, elisp, The GNU Emacs | |
21241 Lisp Reference Manual}, for more about @code{lambda}. Lisp and lambda | |
21242 expressions derive from the Lambda Calculus. | |
21243 | |
21244 @node mapcar, Another Bug, lambda, Print Whole Graph | |
21245 @appendixsubsec The @code{mapcar} Function | |
21246 @findex mapcar | |
21247 | |
21248 @code{mapcar} is a function that calls its first argument with each | |
21249 element of its second argument, in turn. The second argument must be | |
21250 a sequence. | |
21251 | |
21252 The @samp{map} part of the name comes from the mathematical phrase, | |
21253 `mapping over a domain', meaning to apply a function to each of the | |
21254 elements in a domain. The mathematical phrase is based on the | |
21255 metaphor of a surveyor walking, one step at a time, over an area he is | |
21256 mapping. And @samp{car}, of course, comes from the Lisp notion of the | |
21257 first of a list. | |
21258 | |
21259 @need 1250 | |
21260 @noindent | |
21261 For example, | |
21262 | |
21263 @smallexample | |
21264 @group | |
21265 (mapcar '1+ '(2 4 6)) | |
21266 @result{} (3 5 7) | |
21267 @end group | |
21268 @end smallexample | |
21269 | |
21270 @noindent | |
21271 The function @code{1+} which adds one to its argument, is executed on | |
21272 @emph{each} element of the list, and a new list is returned. | |
21273 | |
21274 Contrast this with @code{apply}, which applies its first argument to | |
21275 all the remaining. | |
21276 (@xref{Readying a Graph, , Readying a Graph}, for a explanation of | |
21277 @code{apply}.) | |
21278 | |
21279 @need 1250 | |
21280 In the definition of @code{one-fiftieth}, the first argument is the | |
21281 anonymous function: | |
21282 | |
21283 @smallexample | |
21284 (lambda (arg) (/ arg 50)) | |
21285 @end smallexample | |
21286 | |
21287 @noindent | |
21288 and the second argument is @code{full-range}, which will be bound to | |
21289 @code{list-for-graph}. | |
21290 | |
21291 @need 1250 | |
21292 The whole expression looks like this: | |
21293 | |
21294 @smallexample | |
21295 (mapcar '(lambda (arg) (/ arg 50)) full-range)) | |
21296 @end smallexample | |
21297 | |
21298 @xref{Mapping Functions, , Mapping Functions, elisp, The GNU Emacs | |
21299 Lisp Reference Manual}, for more about @code{mapcar}. | |
21300 | |
21301 Using the @code{one-fiftieth} function, we can generate a list in | |
21302 which each element is one-fiftieth the size of the corresponding | |
21303 element in @code{list-for-graph}. | |
21304 | |
21305 @smallexample | |
21306 @group | |
21307 (setq fiftieth-list-for-graph | |
21308 (one-fiftieth list-for-graph)) | |
21309 @end group | |
21310 @end smallexample | |
21311 | |
21312 @need 1250 | |
21313 The resulting list looks like this: | |
21314 | |
21315 @smallexample | |
21316 @group | |
21317 (10 20 19 15 11 9 6 5 4 3 3 2 2 | |
21318 1 1 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 4) | |
21319 @end group | |
21320 @end smallexample | |
21321 | |
21322 @noindent | |
21323 This, we are almost ready to print! (We also notice the loss of | |
21324 information: many of the higher ranges are 0, meaning that fewer than | |
21325 50 defuns had that many words or symbols---but not necessarily meaning | |
21326 that none had that many words or symbols.) | |
21327 | |
21328 @node Another Bug, Final printed graph, mapcar, Print Whole Graph | |
21329 @appendixsubsec Another Bug @dots{} Most Insidious | |
21330 @cindex Bug, most insidious type | |
21331 @cindex Insidious type of bug | |
21332 | |
21333 I said `almost ready to print'! Of course, there is a bug in the | |
21334 @code{print-graph} function @dots{} It has a @code{vertical-step} | |
21335 option, but not a @code{horizontal-step} option. The | |
21336 @code{top-of-range} scale goes from 10 to 300 by tens. But the | |
21337 @code{print-graph} function will print only by ones. | |
21338 | |
21339 This is a classic example of what some consider the most insidious | |
21340 type of bug, the bug of omission. This is not the kind of bug you can | |
21341 find by studying the code, for it is not in the code; it is an omitted | |
21342 feature. Your best actions are to try your program early and often; | |
21343 and try to arrange, as much as you can, to write code that is easy to | |
21344 understand and easy to change. Try to be aware, whenever you can, | |
21345 that whatever you have written, @emph{will} be rewritten, if not soon, | |
21346 eventually. A hard maxim to follow. | |
21347 | |
21348 It is the @code{print-X-axis-numbered-line} function that needs the | |
21349 work; and then the @code{print-X-axis} and the @code{print-graph} | |
21350 functions need to be adapted. Not much needs to be done; there is one | |
21351 nicety: the numbers ought to line up under the tic marks. This takes | |
21352 a little thought. | |
21353 | |
21354 @need 1250 | |
21355 Here is the corrected @code{print-X-axis-numbered-line}: | |
21356 | |
21357 @smallexample | |
21358 @group | |
21359 (defun print-X-axis-numbered-line | |
21360 (number-of-X-tics X-axis-leading-spaces | |
21361 &optional horizontal-step) | |
21362 "Print line of X-axis numbers" | |
21363 (let ((number X-axis-label-spacing) | |
21364 (horizontal-step (or horizontal-step 1))) | |
21365 @end group | |
21366 @group | |
21367 (insert X-axis-leading-spaces) | |
21368 ;; @r{Delete extra leading spaces.} | |
21369 (delete-char | |
21370 (- (1- | |
21371 (length (number-to-string horizontal-step))))) | |
21372 (insert (concat | |
21373 (make-string | |
21374 @end group | |
21375 @group | |
21376 ;; @r{Insert white space.} | |
21377 (- (* symbol-width | |
21378 X-axis-label-spacing) | |
21379 (1- | |
21380 (length | |
21381 (number-to-string horizontal-step))) | |
21382 2) | |
21383 ? ) | |
21384 (number-to-string | |
21385 (* number horizontal-step)))) | |
21386 @end group | |
21387 @group | |
21388 ;; @r{Insert remaining numbers.} | |
21389 (setq number (+ number X-axis-label-spacing)) | |
21390 (while (> number-of-X-tics 1) | |
21391 (insert (X-axis-element | |
21392 (* number horizontal-step))) | |
21393 (setq number (+ number X-axis-label-spacing)) | |
21394 (setq number-of-X-tics (1- number-of-X-tics))))) | |
21395 @end group | |
21396 @end smallexample | |
21397 | |
21398 @need 1500 | |
21399 If you are reading this in Info, you can see the new versions of | |
21400 @code{print-X-axis} @code{print-graph} and evaluate them. If you are | |
21401 reading this in a printed book, you can see the changed lines here | |
21402 (the full text is too much to print). | |
21403 | |
21404 @iftex | |
21405 @smallexample | |
21406 @group | |
21407 (defun print-X-axis (numbers-list horizontal-step) | |
21408 @dots{} | |
21409 (print-X-axis-numbered-line | |
21410 tic-number leading-spaces horizontal-step)) | |
21411 @end group | |
21412 @end smallexample | |
21413 | |
21414 @smallexample | |
21415 @group | |
21416 (defun print-graph | |
21417 (numbers-list | |
21418 &optional vertical-step horizontal-step) | |
21419 @dots{} | |
21420 (print-X-axis numbers-list horizontal-step)) | |
21421 @end group | |
21422 @end smallexample | |
21423 @end iftex | |
21424 | |
21425 @ifnottex | |
21426 @smallexample | |
21427 @group | |
21428 (defun print-X-axis (numbers-list horizontal-step) | |
21429 "Print X axis labels to length of NUMBERS-LIST. | |
21430 Optionally, HORIZONTAL-STEP, a positive integer, | |
21431 specifies how much an X axis label increments for | |
21432 each column." | |
21433 @end group | |
21434 @group | |
21435 ;; Value of symbol-width and full-Y-label-width | |
21436 ;; are passed by `print-graph'. | |
21437 (let* ((leading-spaces | |
21438 (make-string full-Y-label-width ? )) | |
21439 ;; symbol-width @r{is provided by} graph-body-print | |
21440 (tic-width (* symbol-width X-axis-label-spacing)) | |
21441 (X-length (length numbers-list)) | |
21442 @end group | |
21443 @group | |
21444 (X-tic | |
21445 (concat | |
21446 (make-string | |
21447 ;; @r{Make a string of blanks.} | |
21448 (- (* symbol-width X-axis-label-spacing) | |
21449 (length X-axis-tic-symbol)) | |
21450 ? ) | |
21451 @end group | |
21452 @group | |
21453 ;; @r{Concatenate blanks with tic symbol.} | |
21454 X-axis-tic-symbol)) | |
21455 (tic-number | |
21456 (if (zerop (% X-length tic-width)) | |
21457 (/ X-length tic-width) | |
21458 (1+ (/ X-length tic-width))))) | |
21459 @end group | |
21460 | |
21461 @group | |
21462 (print-X-axis-tic-line | |
21463 tic-number leading-spaces X-tic) | |
21464 (insert "\n") | |
21465 (print-X-axis-numbered-line | |
21466 tic-number leading-spaces horizontal-step))) | |
21467 @end group | |
21468 @end smallexample | |
21469 | |
21470 @smallexample | |
21471 @group | |
21472 (defun print-graph | |
21473 (numbers-list &optional vertical-step horizontal-step) | |
21474 "Print labelled bar graph of the NUMBERS-LIST. | |
21475 The numbers-list consists of the Y-axis values. | |
21476 @end group | |
21477 | |
21478 @group | |
21479 Optionally, VERTICAL-STEP, a positive integer, | |
21480 specifies how much a Y axis label increments for | |
21481 each line. For example, a step of 5 means that | |
21482 each row is five units. | |
21483 @end group | |
21484 | |
21485 @group | |
21486 Optionally, HORIZONTAL-STEP, a positive integer, | |
21487 specifies how much an X axis label increments for | |
21488 each column." | |
21489 (let* ((symbol-width (length graph-blank)) | |
21490 ;; @code{height} @r{is both the largest number} | |
21491 ;; @r{and the number with the most digits.} | |
21492 (height (apply 'max numbers-list)) | |
21493 @end group | |
21494 @group | |
21495 (height-of-top-line | |
21496 (if (zerop (% height Y-axis-label-spacing)) | |
21497 height | |
21498 ;; @r{else} | |
21499 (* (1+ (/ height Y-axis-label-spacing)) | |
21500 Y-axis-label-spacing))) | |
21501 @end group | |
21502 @group | |
21503 (vertical-step (or vertical-step 1)) | |
21504 (full-Y-label-width | |
21505 (length | |
21506 (concat | |
21507 (number-to-string | |
21508 (* height-of-top-line vertical-step)) | |
21509 Y-axis-tic)))) | |
21510 @end group | |
21511 @group | |
21512 (print-Y-axis | |
21513 height-of-top-line full-Y-label-width vertical-step) | |
21514 (graph-body-print | |
21515 numbers-list height-of-top-line symbol-width) | |
21516 (print-X-axis numbers-list horizontal-step))) | |
21517 @end group | |
21518 @end smallexample | |
21519 @end ifnottex | |
21520 | |
21521 @c qqq | |
21522 @ignore | |
21523 Graphing Definitions Re-listed | |
21524 | |
21525 @need 1250 | |
21526 Here are all the graphing definitions in their final form: | |
21527 | |
21528 @smallexample | |
21529 @group | |
21530 (defvar top-of-ranges | |
21531 '(10 20 30 40 50 | |
21532 60 70 80 90 100 | |
21533 110 120 130 140 150 | |
21534 160 170 180 190 200 | |
21535 210 220 230 240 250) | |
21536 "List specifying ranges for `defuns-per-range'.") | |
21537 @end group | |
21538 | |
21539 @group | |
21540 (defvar graph-symbol "*" | |
21541 "String used as symbol in graph, usually an asterisk.") | |
21542 @end group | |
21543 | |
21544 @group | |
21545 (defvar graph-blank " " | |
21546 "String used as blank in graph, usually a blank space. | |
21547 graph-blank must be the same number of columns wide | |
21548 as graph-symbol.") | |
21549 @end group | |
21550 | |
21551 @group | |
21552 (defvar Y-axis-tic " - " | |
21553 "String that follows number in a Y axis label.") | |
21554 @end group | |
21555 | |
21556 @group | |
21557 (defvar Y-axis-label-spacing 5 | |
21558 "Number of lines from one Y axis label to next.") | |
21559 @end group | |
21560 | |
21561 @group | |
21562 (defvar X-axis-tic-symbol "|" | |
21563 "String to insert to point to a column in X axis.") | |
21564 @end group | |
21565 | |
21566 @group | |
21567 (defvar X-axis-label-spacing | |
21568 (if (boundp 'graph-blank) | |
21569 (* 5 (length graph-blank)) 5) | |
21570 "Number of units from one X axis label to next.") | |
21571 @end group | |
21572 @end smallexample | |
21573 | |
21574 @smallexample | |
21575 @group | |
21576 (defun count-words-in-defun () | |
21577 "Return the number of words and symbols in a defun." | |
21578 (beginning-of-defun) | |
21579 (let ((count 0) | |
21580 (end (save-excursion (end-of-defun) (point)))) | |
21581 @end group | |
21582 | |
21583 @group | |
21584 (while | |
21585 (and (< (point) end) | |
21586 (re-search-forward | |
21587 "\\(\\w\\|\\s_\\)+[^ \t\n]*[ \t\n]*" | |
21588 end t)) | |
21589 (setq count (1+ count))) | |
21590 count)) | |
21591 @end group | |
21592 @end smallexample | |
21593 | |
21594 @smallexample | |
21595 @group | |
21596 (defun lengths-list-file (filename) | |
21597 "Return list of definitions' lengths within FILE. | |
21598 The returned list is a list of numbers. | |
21599 Each number is the number of words or | |
21600 symbols in one function definition." | |
21601 @end group | |
21602 | |
21603 @group | |
21604 (message "Working on `%s' ... " filename) | |
21605 (save-excursion | |
21606 (let ((buffer (find-file-noselect filename)) | |
21607 (lengths-list)) | |
21608 (set-buffer buffer) | |
21609 (setq buffer-read-only t) | |
21610 (widen) | |
21611 (goto-char (point-min)) | |
21612 @end group | |
21613 | |
21614 @group | |
21615 (while (re-search-forward "^(defun" nil t) | |
21616 (setq lengths-list | |
21617 (cons (count-words-in-defun) lengths-list))) | |
21618 (kill-buffer buffer) | |
21619 lengths-list))) | |
21620 @end group | |
21621 @end smallexample | |
21622 | |
21623 @smallexample | |
21624 @group | |
21625 (defun lengths-list-many-files (list-of-files) | |
21626 "Return list of lengths of defuns in LIST-OF-FILES." | |
21627 (let (lengths-list) | |
21628 ;;; @r{true-or-false-test} | |
21629 (while list-of-files | |
21630 (setq lengths-list | |
21631 (append | |
21632 lengths-list | |
21633 @end group | |
21634 @group | |
21635 ;;; @r{Generate a lengths' list.} | |
21636 (lengths-list-file | |
21637 (expand-file-name (car list-of-files))))) | |
21638 ;;; @r{Make files' list shorter.} | |
21639 (setq list-of-files (cdr list-of-files))) | |
21640 ;;; @r{Return final value of lengths' list.} | |
21641 lengths-list)) | |
21642 @end group | |
21643 @end smallexample | |
21644 | |
21645 @smallexample | |
21646 @group | |
21647 (defun defuns-per-range (sorted-lengths top-of-ranges) | |
21648 "SORTED-LENGTHS defuns in each TOP-OF-RANGES range." | |
21649 (let ((top-of-range (car top-of-ranges)) | |
21650 (number-within-range 0) | |
21651 defuns-per-range-list) | |
21652 @end group | |
21653 | |
21654 @group | |
21655 ;; @r{Outer loop.} | |
21656 (while top-of-ranges | |
21657 | |
21658 ;; @r{Inner loop.} | |
21659 (while (and | |
21660 ;; @r{Need number for numeric test.} | |
21661 (car sorted-lengths) | |
21662 (< (car sorted-lengths) top-of-range)) | |
21663 | |
21664 ;; @r{Count number of definitions within current range.} | |
21665 (setq number-within-range (1+ number-within-range)) | |
21666 (setq sorted-lengths (cdr sorted-lengths))) | |
21667 @end group | |
21668 | |
21669 @group | |
21670 ;; @r{Exit inner loop but remain within outer loop.} | |
21671 | |
21672 (setq defuns-per-range-list | |
21673 (cons number-within-range defuns-per-range-list)) | |
21674 (setq number-within-range 0) ; @r{Reset count to zero.} | |
21675 | |
21676 ;; @r{Move to next range.} | |
21677 (setq top-of-ranges (cdr top-of-ranges)) | |
21678 ;; @r{Specify next top of range value.} | |
21679 (setq top-of-range (car top-of-ranges))) | |
21680 @end group | |
21681 | |
21682 @group | |
21683 ;; @r{Exit outer loop and count the number of defuns larger than} | |
21684 ;; @r{ the largest top-of-range value.} | |
21685 (setq defuns-per-range-list | |
21686 (cons | |
21687 (length sorted-lengths) | |
21688 defuns-per-range-list)) | |
21689 | |
21690 ;; @r{Return a list of the number of definitions within each range,} | |
21691 ;; @r{ smallest to largest.} | |
21692 (nreverse defuns-per-range-list))) | |
21693 @end group | |
21694 @end smallexample | |
21695 | |
21696 @smallexample | |
21697 @group | |
21698 (defun column-of-graph (max-graph-height actual-height) | |
21699 "Return list of MAX-GRAPH-HEIGHT strings; | |
21700 ACTUAL-HEIGHT are graph-symbols. | |
21701 The graph-symbols are contiguous entries at the end | |
21702 of the list. | |
21703 The list will be inserted as one column of a graph. | |
21704 The strings are either graph-blank or graph-symbol." | |
21705 @end group | |
21706 | |
21707 @group | |
21708 (let ((insert-list nil) | |
21709 (number-of-top-blanks | |
21710 (- max-graph-height actual-height))) | |
21711 | |
21712 ;; @r{Fill in @code{graph-symbols}.} | |
21713 (while (> actual-height 0) | |
21714 (setq insert-list (cons graph-symbol insert-list)) | |
21715 (setq actual-height (1- actual-height))) | |
21716 @end group | |
21717 | |
21718 @group | |
21719 ;; @r{Fill in @code{graph-blanks}.} | |
21720 (while (> number-of-top-blanks 0) | |
21721 (setq insert-list (cons graph-blank insert-list)) | |
21722 (setq number-of-top-blanks | |
21723 (1- number-of-top-blanks))) | |
21724 | |
21725 ;; @r{Return whole list.} | |
21726 insert-list)) | |
21727 @end group | |
21728 @end smallexample | |
21729 | |
21730 @smallexample | |
21731 @group | |
21732 (defun Y-axis-element (number full-Y-label-width) | |
21733 "Construct a NUMBERed label element. | |
21734 A numbered element looks like this ` 5 - ', | |
21735 and is padded as needed so all line up with | |
21736 the element for the largest number." | |
21737 @end group | |
21738 @group | |
21739 (let* ((leading-spaces | |
21740 (- full-Y-label-width | |
21741 (length | |
21742 (concat (number-to-string number) | |
21743 Y-axis-tic))))) | |
21744 @end group | |
21745 @group | |
21746 (concat | |
21747 (make-string leading-spaces ? ) | |
21748 (number-to-string number) | |
21749 Y-axis-tic))) | |
21750 @end group | |
21751 @end smallexample | |
21752 | |
21753 @smallexample | |
21754 @group | |
21755 (defun print-Y-axis | |
21756 (height full-Y-label-width &optional vertical-step) | |
21757 "Insert Y axis by HEIGHT and FULL-Y-LABEL-WIDTH. | |
21758 Height must be the maximum height of the graph. | |
21759 Full width is the width of the highest label element. | |
21760 Optionally, print according to VERTICAL-STEP." | |
21761 @end group | |
21762 @group | |
21763 ;; Value of height and full-Y-label-width | |
21764 ;; are passed by `print-graph'. | |
21765 (let ((start (point))) | |
21766 (insert-rectangle | |
21767 (Y-axis-column height full-Y-label-width vertical-step)) | |
21768 @end group | |
21769 @group | |
21770 ;; @r{Place point ready for inserting graph.} | |
21771 (goto-char start) | |
21772 ;; @r{Move point forward by value of} full-Y-label-width | |
21773 (forward-char full-Y-label-width))) | |
21774 @end group | |
21775 @end smallexample | |
21776 | |
21777 @smallexample | |
21778 @group | |
21779 (defun print-X-axis-tic-line | |
21780 (number-of-X-tics X-axis-leading-spaces X-axis-tic-element) | |
21781 "Print ticks for X axis." | |
21782 (insert X-axis-leading-spaces) | |
21783 (insert X-axis-tic-symbol) ; @r{Under first column.} | |
21784 @end group | |
21785 @group | |
21786 ;; @r{Insert second tic in the right spot.} | |
21787 (insert (concat | |
21788 (make-string | |
21789 (- (* symbol-width X-axis-label-spacing) | |
21790 ;; @r{Insert white space up to second tic symbol.} | |
21791 (* 2 (length X-axis-tic-symbol))) | |
21792 ? ) | |
21793 X-axis-tic-symbol)) | |
21794 @end group | |
21795 @group | |
21796 ;; @r{Insert remaining ticks.} | |
21797 (while (> number-of-X-tics 1) | |
21798 (insert X-axis-tic-element) | |
21799 (setq number-of-X-tics (1- number-of-X-tics)))) | |
21800 @end group | |
21801 @end smallexample | |
21802 | |
21803 @smallexample | |
21804 @group | |
21805 (defun X-axis-element (number) | |
21806 "Construct a numbered X axis element." | |
21807 (let ((leading-spaces | |
21808 (- (* symbol-width X-axis-label-spacing) | |
21809 (length (number-to-string number))))) | |
21810 (concat (make-string leading-spaces ? ) | |
21811 (number-to-string number)))) | |
21812 @end group | |
21813 @end smallexample | |
21814 | |
21815 @smallexample | |
21816 @group | |
21817 (defun graph-body-print (numbers-list height symbol-width) | |
21818 "Print a bar graph of the NUMBERS-LIST. | |
21819 The numbers-list consists of the Y-axis values. | |
21820 HEIGHT is maximum height of graph. | |
21821 SYMBOL-WIDTH is number of each column." | |
21822 @end group | |
21823 @group | |
21824 (let (from-position) | |
21825 (while numbers-list | |
21826 (setq from-position (point)) | |
21827 (insert-rectangle | |
21828 (column-of-graph height (car numbers-list))) | |
21829 (goto-char from-position) | |
21830 (forward-char symbol-width) | |
21831 @end group | |
21832 @group | |
21833 ;; @r{Draw graph column by column.} | |
21834 (sit-for 0) | |
21835 (setq numbers-list (cdr numbers-list))) | |
21836 ;; @r{Place point for X axis labels.} | |
21837 (forward-line height) | |
21838 (insert "\n"))) | |
21839 @end group | |
21840 @end smallexample | |
21841 | |
21842 @smallexample | |
21843 @group | |
21844 (defun Y-axis-column | |
21845 (height width-of-label &optional vertical-step) | |
21846 "Construct list of labels for Y axis. | |
21847 HEIGHT is maximum height of graph. | |
21848 WIDTH-OF-LABEL is maximum width of label. | |
21849 @end group | |
21850 @group | |
21851 VERTICAL-STEP, an option, is a positive integer | |
21852 that specifies how much a Y axis label increments | |
21853 for each line. For example, a step of 5 means | |
21854 that each line is five units of the graph." | |
21855 (let (Y-axis | |
21856 (number-per-line (or vertical-step 1))) | |
21857 @end group | |
21858 @group | |
21859 (while (> height 1) | |
21860 (if (zerop (% height Y-axis-label-spacing)) | |
21861 ;; @r{Insert label.} | |
21862 (setq Y-axis | |
21863 (cons | |
21864 (Y-axis-element | |
21865 (* height number-per-line) | |
21866 width-of-label) | |
21867 Y-axis)) | |
21868 @end group | |
21869 @group | |
21870 ;; @r{Else, insert blanks.} | |
21871 (setq Y-axis | |
21872 (cons | |
21873 (make-string width-of-label ? ) | |
21874 Y-axis))) | |
21875 (setq height (1- height))) | |
21876 @end group | |
21877 @group | |
21878 ;; @r{Insert base line.} | |
21879 (setq Y-axis (cons (Y-axis-element | |
21880 (or vertical-step 1) | |
21881 width-of-label) | |
21882 Y-axis)) | |
21883 (nreverse Y-axis))) | |
21884 @end group | |
21885 @end smallexample | |
21886 | |
21887 @smallexample | |
21888 @group | |
21889 (defun print-X-axis-numbered-line | |
21890 (number-of-X-tics X-axis-leading-spaces | |
21891 &optional horizontal-step) | |
21892 "Print line of X-axis numbers" | |
21893 (let ((number X-axis-label-spacing) | |
21894 (horizontal-step (or horizontal-step 1))) | |
21895 @end group | |
21896 @group | |
21897 (insert X-axis-leading-spaces) | |
21898 ;; line up number | |
21899 (delete-char (- (1- (length (number-to-string horizontal-step))))) | |
21900 (insert (concat | |
21901 (make-string | |
21902 ;; @r{Insert white space up to next number.} | |
21903 (- (* symbol-width X-axis-label-spacing) | |
21904 (1- (length (number-to-string horizontal-step))) | |
21905 2) | |
21906 ? ) | |
21907 (number-to-string (* number horizontal-step)))) | |
21908 @end group | |
21909 @group | |
21910 ;; @r{Insert remaining numbers.} | |
21911 (setq number (+ number X-axis-label-spacing)) | |
21912 (while (> number-of-X-tics 1) | |
21913 (insert (X-axis-element (* number horizontal-step))) | |
21914 (setq number (+ number X-axis-label-spacing)) | |
21915 (setq number-of-X-tics (1- number-of-X-tics))))) | |
21916 @end group | |
21917 @end smallexample | |
21918 | |
21919 @smallexample | |
21920 @group | |
21921 (defun print-X-axis (numbers-list horizontal-step) | |
21922 "Print X axis labels to length of NUMBERS-LIST. | |
21923 Optionally, HORIZONTAL-STEP, a positive integer, | |
21924 specifies how much an X axis label increments for | |
21925 each column." | |
21926 @end group | |
21927 @group | |
21928 ;; Value of symbol-width and full-Y-label-width | |
21929 ;; are passed by `print-graph'. | |
21930 (let* ((leading-spaces | |
21931 (make-string full-Y-label-width ? )) | |
21932 ;; symbol-width @r{is provided by} graph-body-print | |
21933 (tic-width (* symbol-width X-axis-label-spacing)) | |
21934 (X-length (length numbers-list)) | |
21935 @end group | |
21936 @group | |
21937 (X-tic | |
21938 (concat | |
21939 (make-string | |
21940 ;; @r{Make a string of blanks.} | |
21941 (- (* symbol-width X-axis-label-spacing) | |
21942 (length X-axis-tic-symbol)) | |
21943 ? ) | |
21944 @end group | |
21945 @group | |
21946 ;; @r{Concatenate blanks with tic symbol.} | |
21947 X-axis-tic-symbol)) | |
21948 (tic-number | |
21949 (if (zerop (% X-length tic-width)) | |
21950 (/ X-length tic-width) | |
21951 (1+ (/ X-length tic-width))))) | |
21952 @end group | |
21953 | |
21954 @group | |
21955 (print-X-axis-tic-line | |
21956 tic-number leading-spaces X-tic) | |
21957 (insert "\n") | |
21958 (print-X-axis-numbered-line | |
21959 tic-number leading-spaces horizontal-step))) | |
21960 @end group | |
21961 @end smallexample | |
21962 | |
21963 @smallexample | |
21964 @group | |
21965 (defun one-fiftieth (full-range) | |
21966 "Return list, each number of which is 1/50th previous." | |
21967 (mapcar '(lambda (arg) (/ arg 50)) full-range)) | |
21968 @end group | |
21969 @end smallexample | |
21970 | |
21971 @smallexample | |
21972 @group | |
21973 (defun print-graph | |
21974 (numbers-list &optional vertical-step horizontal-step) | |
21975 "Print labelled bar graph of the NUMBERS-LIST. | |
21976 The numbers-list consists of the Y-axis values. | |
21977 @end group | |
21978 | |
21979 @group | |
21980 Optionally, VERTICAL-STEP, a positive integer, | |
21981 specifies how much a Y axis label increments for | |
21982 each line. For example, a step of 5 means that | |
21983 each row is five units. | |
21984 @end group | |
21985 | |
21986 @group | |
21987 Optionally, HORIZONTAL-STEP, a positive integer, | |
21988 specifies how much an X axis label increments for | |
21989 each column." | |
21990 (let* ((symbol-width (length graph-blank)) | |
21991 ;; @code{height} @r{is both the largest number} | |
21992 ;; @r{and the number with the most digits.} | |
21993 (height (apply 'max numbers-list)) | |
21994 @end group | |
21995 @group | |
21996 (height-of-top-line | |
21997 (if (zerop (% height Y-axis-label-spacing)) | |
21998 height | |
21999 ;; @r{else} | |
22000 (* (1+ (/ height Y-axis-label-spacing)) | |
22001 Y-axis-label-spacing))) | |
22002 @end group | |
22003 @group | |
22004 (vertical-step (or vertical-step 1)) | |
22005 (full-Y-label-width | |
22006 (length | |
22007 (concat | |
22008 (number-to-string | |
22009 (* height-of-top-line vertical-step)) | |
22010 Y-axis-tic)))) | |
22011 @end group | |
22012 @group | |
22013 | |
22014 (print-Y-axis | |
22015 height-of-top-line full-Y-label-width vertical-step) | |
22016 (graph-body-print | |
22017 numbers-list height-of-top-line symbol-width) | |
22018 (print-X-axis numbers-list horizontal-step))) | |
22019 @end group | |
22020 @end smallexample | |
22021 @c qqq | |
22022 @end ignore | |
22023 | |
22024 @page | |
22025 @node Final printed graph, , Another Bug, Print Whole Graph | |
22026 @appendixsubsec The Printed Graph | |
22027 | |
22028 When made and installed, you can call the @code{print-graph} command | |
22029 like this: | |
22030 @sp 1 | |
22031 | |
22032 @smallexample | |
22033 @group | |
22034 (print-graph fiftieth-list-for-graph 50 10) | |
22035 @end group | |
22036 @end smallexample | |
22037 @sp 1 | |
22038 | |
22039 @noindent | |
22040 Here is the graph: | |
22041 @sp 2 | |
22042 | |
22043 @smallexample | |
22044 @group | |
22045 1000 - * | |
22046 ** | |
22047 ** | |
22048 ** | |
22049 ** | |
22050 750 - *** | |
22051 *** | |
22052 *** | |
22053 *** | |
22054 **** | |
22055 500 - ***** | |
22056 ****** | |
22057 ****** | |
22058 ****** | |
22059 ******* | |
22060 250 - ******** | |
22061 ********* * | |
22062 *********** * | |
22063 ************* * | |
22064 50 - ***************** * * | |
22065 | | | | | | | | | |
22066 10 50 100 150 200 250 300 350 | |
22067 @end group | |
22068 @end smallexample | |
22069 | |
22070 @sp 2 | |
22071 | |
22072 @noindent | |
22073 The largest group of functions contain 10 -- 19 words and symbols each. | |
22074 | |
22075 @node Free Software and Free Manuals, GNU Free Documentation License, Full Graph, Top | |
22076 @appendix Free Software and Free Manuals | |
22077 | |
22078 @strong{by Richard M. Stallman} | |
22079 @sp 1 | |
22080 | |
22081 The biggest deficiency in free operating systems is not in the | |
22082 software---it is the lack of good free manuals that we can include in | |
22083 these systems. Many of our most important programs do not come with | |
22084 full manuals. Documentation is an essential part of any software | |
22085 package; when an important free software package does not come with a | |
22086 free manual, that is a major gap. We have many such gaps today. | |
22087 | |
22088 Once upon a time, many years ago, I thought I would learn Perl. I got | |
22089 a copy of a free manual, but I found it hard to read. When I asked | |
22090 Perl users about alternatives, they told me that there were better | |
22091 introductory manuals---but those were not free. | |
22092 | |
22093 Why was this? The authors of the good manuals had written them for | |
22094 O'Reilly Associates, which published them with restrictive terms---no | |
22095 copying, no modification, source files not available---which exclude | |
22096 them from the free software community. | |
22097 | |
22098 That wasn't the first time this sort of thing has happened, and (to | |
22099 our community's great loss) it was far from the last. Proprietary | |
22100 manual publishers have enticed a great many authors to restrict their | |
22101 manuals since then. Many times I have heard a GNU user eagerly tell me | |
22102 about a manual that he is writing, with which he expects to help the | |
22103 GNU project---and then had my hopes dashed, as he proceeded to explain | |
22104 that he had signed a contract with a publisher that would restrict it | |
22105 so that we cannot use it. | |
22106 | |
22107 Given that writing good English is a rare skill among programmers, we | |
22108 can ill afford to lose manuals this way. | |
22109 | |
22110 @c (texinfo)uref | |
22111 (The Free Software Foundation | |
22112 @uref{http://www.gnu.org/doc/doc.html#DescriptionsOfGNUDocumentation, , | |
22113 sells printed copies} of free @uref{http://www.gnu.org/doc/doc.html, | |
22114 GNU manuals}, too.) | |
22115 | |
22116 Free documentation, like free software, is a matter of freedom, not | |
22117 price. The problem with these manuals was not that O'Reilly Associates | |
22118 charged a price for printed copies---that in itself is fine. (The Free | |
22119 Software Foundation sells printed copies of free GNU manuals, too.) | |
22120 But GNU manuals are available in source code form, while these manuals | |
22121 are available only on paper. GNU manuals come with permission to copy | |
22122 and modify; the Perl manuals do not. These restrictions are the | |
22123 problems. | |
22124 | |
22125 The criterion for a free manual is pretty much the same as for free | |
22126 software: it is a matter of giving all users certain | |
22127 freedoms. Redistribution (including commercial redistribution) must be | |
22128 permitted, so that the manual can accompany every copy of the program, | |
22129 on-line or on paper. Permission for modification is crucial too. | |
22130 | |
22131 As a general rule, I don't believe that it is essential for people to | |
22132 have permission to modify all sorts of articles and books. The issues | |
22133 for writings are not necessarily the same as those for software. For | |
22134 example, I don't think you or I are obliged to give permission to | |
22135 modify articles like this one, which describe our actions and our | |
22136 views. | |
22137 | |
22138 But there is a particular reason why the freedom to modify is crucial | |
22139 for documentation for free software. When people exercise their right | |
22140 to modify the software, and add or change its features, if they are | |
22141 conscientious they will change the manual too---so they can provide | |
22142 accurate and usable documentation with the modified program. A manual | |
22143 which forbids programmers to be conscientious and finish the job, or | |
22144 more precisely requires them to write a new manual from scratch if | |
22145 they change the program, does not fill our community's needs. | |
22146 | |
22147 While a blanket prohibition on modification is unacceptable, some | |
22148 kinds of limits on the method of modification pose no problem. For | |
22149 example, requirements to preserve the original author's copyright | |
22150 notice, the distribution terms, or the list of authors, are ok. It is | |
22151 also no problem to require modified versions to include notice that | |
22152 they were modified, even to have entire sections that may not be | |
22153 deleted or changed, as long as these sections deal with nontechnical | |
22154 topics. (Some GNU manuals have them.) | |
22155 | |
22156 These kinds of restrictions are not a problem because, as a practical | |
22157 matter, they don't stop the conscientious programmer from adapting the | |
22158 manual to fit the modified program. In other words, they don't block | |
22159 the free software community from making full use of the manual. | |
22160 | |
22161 However, it must be possible to modify all the technical content of | |
22162 the manual, and then distribute the result in all the usual media, | |
22163 through all the usual channels; otherwise, the restrictions do block | |
22164 the community, the manual is not free, and so we need another manual. | |
22165 | |
22166 Unfortunately, it is often hard to find someone to write another | |
22167 manual when a proprietary manual exists. The obstacle is that many | |
22168 users think that a proprietary manual is good enough---so they don't | |
22169 see the need to write a free manual. They do not see that the free | |
22170 operating system has a gap that needs filling. | |
22171 | |
22172 Why do users think that proprietary manuals are good enough? Some have | |
22173 not considered the issue. I hope this article will do something to | |
22174 change that. | |
22175 | |
22176 Other users consider proprietary manuals acceptable for the same | |
22177 reason so many people consider proprietary software acceptable: they | |
22178 judge in purely practical terms, not using freedom as a | |
22179 criterion. These people are entitled to their opinions, but since | |
22180 those opinions spring from values which do not include freedom, they | |
22181 are no guide for those of us who do value freedom. | |
22182 | |
22183 Please spread the word about this issue. We continue to lose manuals | |
22184 to proprietary publishing. If we spread the word that proprietary | |
22185 manuals are not sufficient, perhaps the next person who wants to help | |
22186 GNU by writing documentation will realize, before it is too late, that | |
22187 he must above all make it free. | |
22188 | |
22189 We can also encourage commercial publishers to sell free, copylefted | |
22190 manuals instead of proprietary ones. One way you can help this is to | |
22191 check the distribution terms of a manual before you buy it, and prefer | |
22192 copylefted manuals to non-copylefted ones. | |
22193 | |
22194 @sp 2 | |
22195 @noindent | |
22196 Note: The Free Software Foundation maintains a page on its Web site | |
22197 that lists free books available from other publishers:@* | |
22198 @uref{http://www.gnu.org/doc/other-free-books.html} | |
22199 | |
22200 @node GNU Free Documentation License, Index, Free Software and Free Manuals, Top | |
22201 @appendix GNU Free Documentation License | |
22202 | |
22203 @cindex FDL, GNU Free Documentation License | |
22204 @center Version 1.2, November 2002 | |
22205 | |
22206 @display | |
22207 Copyright @copyright{} 2000,2001,2002 Free Software Foundation, Inc. | |
22208 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA | |
22209 | |
22210 Everyone is permitted to copy and distribute verbatim copies | |
22211 of this license document, but changing it is not allowed. | |
22212 @end display | |
22213 | |
22214 @enumerate 0 | |
22215 @item | |
22216 PREAMBLE | |
22217 | |
22218 The purpose of this License is to make a manual, textbook, or other | |
22219 functional and useful document @dfn{free} in the sense of freedom: to | |
22220 assure everyone the effective freedom to copy and redistribute it, | |
22221 with or without modifying it, either commercially or noncommercially. | |
22222 Secondarily, this License preserves for the author and publisher a way | |
22223 to get credit for their work, while not being considered responsible | |
22224 for modifications made by others. | |
22225 | |
22226 This License is a kind of ``copyleft'', which means that derivative | |
22227 works of the document must themselves be free in the same sense. It | |
22228 complements the GNU General Public License, which is a copyleft | |
22229 license designed for free software. | |
22230 | |
22231 We have designed this License in order to use it for manuals for free | |
22232 software, because free software needs free documentation: a free | |
22233 program should come with manuals providing the same freedoms that the | |
22234 software does. But this License is not limited to software manuals; | |
22235 it can be used for any textual work, regardless of subject matter or | |
22236 whether it is published as a printed book. We recommend this License | |
22237 principally for works whose purpose is instruction or reference. | |
22238 | |
22239 @item | |
22240 APPLICABILITY AND DEFINITIONS | |
22241 | |
22242 This License applies to any manual or other work, in any medium, that | |
22243 contains a notice placed by the copyright holder saying it can be | |
22244 distributed under the terms of this License. Such a notice grants a | |
22245 world-wide, royalty-free license, unlimited in duration, to use that | |
22246 work under the conditions stated herein. The ``Document'', below, | |
22247 refers to any such manual or work. Any member of the public is a | |
22248 licensee, and is addressed as ``you''. You accept the license if you | |
22249 copy, modify or distribute the work in a way requiring permission | |
22250 under copyright law. | |
22251 | |
22252 A ``Modified Version'' of the Document means any work containing the | |
22253 Document or a portion of it, either copied verbatim, or with | |
22254 modifications and/or translated into another language. | |
22255 | |
22256 A ``Secondary Section'' is a named appendix or a front-matter section | |
22257 of the Document that deals exclusively with the relationship of the | |
22258 publishers or authors of the Document to the Document's overall | |
22259 subject (or to related matters) and contains nothing that could fall | |
22260 directly within that overall subject. (Thus, if the Document is in | |
22261 part a textbook of mathematics, a Secondary Section may not explain | |
22262 any mathematics.) The relationship could be a matter of historical | |
22263 connection with the subject or with related matters, or of legal, | |
22264 commercial, philosophical, ethical or political position regarding | |
22265 them. | |
22266 | |
22267 The ``Invariant Sections'' are certain Secondary Sections whose titles | |
22268 are designated, as being those of Invariant Sections, in the notice | |
22269 that says that the Document is released under this License. If a | |
22270 section does not fit the above definition of Secondary then it is not | |
22271 allowed to be designated as Invariant. The Document may contain zero | |
22272 Invariant Sections. If the Document does not identify any Invariant | |
22273 Sections then there are none. | |
22274 | |
22275 The ``Cover Texts'' are certain short passages of text that are listed, | |
22276 as Front-Cover Texts or Back-Cover Texts, in the notice that says that | |
22277 the Document is released under this License. A Front-Cover Text may | |
22278 be at most 5 words, and a Back-Cover Text may be at most 25 words. | |
22279 | |
22280 A ``Transparent'' copy of the Document means a machine-readable copy, | |
22281 represented in a format whose specification is available to the | |
22282 general public, that is suitable for revising the document | |
22283 straightforwardly with generic text editors or (for images composed of | |
22284 pixels) generic paint programs or (for drawings) some widely available | |
22285 drawing editor, and that is suitable for input to text formatters or | |
22286 for automatic translation to a variety of formats suitable for input | |
22287 to text formatters. A copy made in an otherwise Transparent file | |
22288 format whose markup, or absence of markup, has been arranged to thwart | |
22289 or discourage subsequent modification by readers is not Transparent. | |
22290 An image format is not Transparent if used for any substantial amount | |
22291 of text. A copy that is not ``Transparent'' is called ``Opaque''. | |
22292 | |
22293 Examples of suitable formats for Transparent copies include plain | |
22294 @sc{ascii} without markup, Texinfo input format, La@TeX{} input | |
22295 format, @acronym{SGML} or @acronym{XML} using a publicly available | |
22296 @acronym{DTD}, and standard-conforming simple @acronym{HTML}, | |
22297 PostScript or @acronym{PDF} designed for human modification. Examples | |
22298 of transparent image formats include @acronym{PNG}, @acronym{XCF} and | |
22299 @acronym{JPG}. Opaque formats include proprietary formats that can be | |
22300 read and edited only by proprietary word processors, @acronym{SGML} or | |
22301 @acronym{XML} for which the @acronym{DTD} and/or processing tools are | |
22302 not generally available, and the machine-generated @acronym{HTML}, | |
22303 PostScript or @acronym{PDF} produced by some word processors for | |
22304 output purposes only. | |
22305 | |
22306 The ``Title Page'' means, for a printed book, the title page itself, | |
22307 plus such following pages as are needed to hold, legibly, the material | |
22308 this License requires to appear in the title page. For works in | |
22309 formats which do not have any title page as such, ``Title Page'' means | |
22310 the text near the most prominent appearance of the work's title, | |
22311 preceding the beginning of the body of the text. | |
22312 | |
22313 A section ``Entitled XYZ'' means a named subunit of the Document whose | |
22314 title either is precisely XYZ or contains XYZ in parentheses following | |
22315 text that translates XYZ in another language. (Here XYZ stands for a | |
22316 specific section name mentioned below, such as ``Acknowledgements'', | |
22317 ``Dedications'', ``Endorsements'', or ``History''.) To ``Preserve the Title'' | |
22318 of such a section when you modify the Document means that it remains a | |
22319 section ``Entitled XYZ'' according to this definition. | |
22320 | |
22321 The Document may include Warranty Disclaimers next to the notice which | |
22322 states that this License applies to the Document. These Warranty | |
22323 Disclaimers are considered to be included by reference in this | |
22324 License, but only as regards disclaiming warranties: any other | |
22325 implication that these Warranty Disclaimers may have is void and has | |
22326 no effect on the meaning of this License. | |
22327 | |
22328 @item | |
22329 VERBATIM COPYING | |
22330 | |
22331 You may copy and distribute the Document in any medium, either | |
22332 commercially or noncommercially, provided that this License, the | |
22333 copyright notices, and the license notice saying this License applies | |
22334 to the Document are reproduced in all copies, and that you add no other | |
22335 conditions whatsoever to those of this License. You may not use | |
22336 technical measures to obstruct or control the reading or further | |
22337 copying of the copies you make or distribute. However, you may accept | |
22338 compensation in exchange for copies. If you distribute a large enough | |
22339 number of copies you must also follow the conditions in section 3. | |
22340 | |
22341 You may also lend copies, under the same conditions stated above, and | |
22342 you may publicly display copies. | |
22343 | |
22344 @item | |
22345 COPYING IN QUANTITY | |
22346 | |
22347 If you publish printed copies (or copies in media that commonly have | |
22348 printed covers) of the Document, numbering more than 100, and the | |
22349 Document's license notice requires Cover Texts, you must enclose the | |
22350 copies in covers that carry, clearly and legibly, all these Cover | |
22351 Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on | |
22352 the back cover. Both covers must also clearly and legibly identify | |
22353 you as the publisher of these copies. The front cover must present | |
22354 the full title with all words of the title equally prominent and | |
22355 visible. You may add other material on the covers in addition. | |
22356 Copying with changes limited to the covers, as long as they preserve | |
22357 the title of the Document and satisfy these conditions, can be treated | |
22358 as verbatim copying in other respects. | |
22359 | |
22360 If the required texts for either cover are too voluminous to fit | |
22361 legibly, you should put the first ones listed (as many as fit | |
22362 reasonably) on the actual cover, and continue the rest onto adjacent | |
22363 pages. | |
22364 | |
22365 If you publish or distribute Opaque copies of the Document numbering | |
22366 more than 100, you must either include a machine-readable Transparent | |
22367 copy along with each Opaque copy, or state in or with each Opaque copy | |
22368 a computer-network location from which the general network-using | |
22369 public has access to download using public-standard network protocols | |
22370 a complete Transparent copy of the Document, free of added material. | |
22371 If you use the latter option, you must take reasonably prudent steps, | |
22372 when you begin distribution of Opaque copies in quantity, to ensure | |
22373 that this Transparent copy will remain thus accessible at the stated | |
22374 location until at least one year after the last time you distribute an | |
22375 Opaque copy (directly or through your agents or retailers) of that | |
22376 edition to the public. | |
22377 | |
22378 It is requested, but not required, that you contact the authors of the | |
22379 Document well before redistributing any large number of copies, to give | |
22380 them a chance to provide you with an updated version of the Document. | |
22381 | |
22382 @item | |
22383 MODIFICATIONS | |
22384 | |
22385 You may copy and distribute a Modified Version of the Document under | |
22386 the conditions of sections 2 and 3 above, provided that you release | |
22387 the Modified Version under precisely this License, with the Modified | |
22388 Version filling the role of the Document, thus licensing distribution | |
22389 and modification of the Modified Version to whoever possesses a copy | |
22390 of it. In addition, you must do these things in the Modified Version: | |
22391 | |
22392 @enumerate A | |
22393 @item | |
22394 Use in the Title Page (and on the covers, if any) a title distinct | |
22395 from that of the Document, and from those of previous versions | |
22396 (which should, if there were any, be listed in the History section | |
22397 of the Document). You may use the same title as a previous version | |
22398 if the original publisher of that version gives permission. | |
22399 | |
22400 @item | |
22401 List on the Title Page, as authors, one or more persons or entities | |
22402 responsible for authorship of the modifications in the Modified | |
22403 Version, together with at least five of the principal authors of the | |
22404 Document (all of its principal authors, if it has fewer than five), | |
22405 unless they release you from this requirement. | |
22406 | |
22407 @item | |
22408 State on the Title page the name of the publisher of the | |
22409 Modified Version, as the publisher. | |
22410 | |
22411 @item | |
22412 Preserve all the copyright notices of the Document. | |
22413 | |
22414 @item | |
22415 Add an appropriate copyright notice for your modifications | |
22416 adjacent to the other copyright notices. | |
22417 | |
22418 @item | |
22419 Include, immediately after the copyright notices, a license notice | |
22420 giving the public permission to use the Modified Version under the | |
22421 terms of this License, in the form shown in the Addendum below. | |
22422 | |
22423 @item | |
22424 Preserve in that license notice the full lists of Invariant Sections | |
22425 and required Cover Texts given in the Document's license notice. | |
22426 | |
22427 @item | |
22428 Include an unaltered copy of this License. | |
22429 | |
22430 @item | |
22431 Preserve the section Entitled ``History'', Preserve its Title, and add | |
22432 to it an item stating at least the title, year, new authors, and | |
22433 publisher of the Modified Version as given on the Title Page. If | |
22434 there is no section Entitled ``History'' in the Document, create one | |
22435 stating the title, year, authors, and publisher of the Document as | |
22436 given on its Title Page, then add an item describing the Modified | |
22437 Version as stated in the previous sentence. | |
22438 | |
22439 @item | |
22440 Preserve the network location, if any, given in the Document for | |
22441 public access to a Transparent copy of the Document, and likewise | |
22442 the network locations given in the Document for previous versions | |
22443 it was based on. These may be placed in the ``History'' section. | |
22444 You may omit a network location for a work that was published at | |
22445 least four years before the Document itself, or if the original | |
22446 publisher of the version it refers to gives permission. | |
22447 | |
22448 @item | |
22449 For any section Entitled ``Acknowledgements'' or ``Dedications'', Preserve | |
22450 the Title of the section, and preserve in the section all the | |
22451 substance and tone of each of the contributor acknowledgements and/or | |
22452 dedications given therein. | |
22453 | |
22454 @item | |
22455 Preserve all the Invariant Sections of the Document, | |
22456 unaltered in their text and in their titles. Section numbers | |
22457 or the equivalent are not considered part of the section titles. | |
22458 | |
22459 @item | |
22460 Delete any section Entitled ``Endorsements''. Such a section | |
22461 may not be included in the Modified Version. | |
22462 | |
22463 @item | |
22464 Do not retitle any existing section to be Entitled ``Endorsements'' or | |
22465 to conflict in title with any Invariant Section. | |
22466 | |
22467 @item | |
22468 Preserve any Warranty Disclaimers. | |
22469 @end enumerate | |
22470 | |
22471 If the Modified Version includes new front-matter sections or | |
22472 appendices that qualify as Secondary Sections and contain no material | |
22473 copied from the Document, you may at your option designate some or all | |
22474 of these sections as invariant. To do this, add their titles to the | |
22475 list of Invariant Sections in the Modified Version's license notice. | |
22476 These titles must be distinct from any other section titles. | |
22477 | |
22478 You may add a section Entitled ``Endorsements'', provided it contains | |
22479 nothing but endorsements of your Modified Version by various | |
22480 parties---for example, statements of peer review or that the text has | |
22481 been approved by an organization as the authoritative definition of a | |
22482 standard. | |
22483 | |
22484 You may add a passage of up to five words as a Front-Cover Text, and a | |
22485 passage of up to 25 words as a Back-Cover Text, to the end of the list | |
22486 of Cover Texts in the Modified Version. Only one passage of | |
22487 Front-Cover Text and one of Back-Cover Text may be added by (or | |
22488 through arrangements made by) any one entity. If the Document already | |
22489 includes a cover text for the same cover, previously added by you or | |
22490 by arrangement made by the same entity you are acting on behalf of, | |
22491 you may not add another; but you may replace the old one, on explicit | |
22492 permission from the previous publisher that added the old one. | |
22493 | |
22494 The author(s) and publisher(s) of the Document do not by this License | |
22495 give permission to use their names for publicity for or to assert or | |
22496 imply endorsement of any Modified Version. | |
22497 | |
22498 @item | |
22499 COMBINING DOCUMENTS | |
22500 | |
22501 You may combine the Document with other documents released under this | |
22502 License, under the terms defined in section 4 above for modified | |
22503 versions, provided that you include in the combination all of the | |
22504 Invariant Sections of all of the original documents, unmodified, and | |
22505 list them all as Invariant Sections of your combined work in its | |
22506 license notice, and that you preserve all their Warranty Disclaimers. | |
22507 | |
22508 The combined work need only contain one copy of this License, and | |
22509 multiple identical Invariant Sections may be replaced with a single | |
22510 copy. If there are multiple Invariant Sections with the same name but | |
22511 different contents, make the title of each such section unique by | |
22512 adding at the end of it, in parentheses, the name of the original | |
22513 author or publisher of that section if known, or else a unique number. | |
22514 Make the same adjustment to the section titles in the list of | |
22515 Invariant Sections in the license notice of the combined work. | |
22516 | |
22517 In the combination, you must combine any sections Entitled ``History'' | |
22518 in the various original documents, forming one section Entitled | |
22519 ``History''; likewise combine any sections Entitled ``Acknowledgements'', | |
22520 and any sections Entitled ``Dedications''. You must delete all | |
22521 sections Entitled ``Endorsements.'' | |
22522 | |
22523 @item | |
22524 COLLECTIONS OF DOCUMENTS | |
22525 | |
22526 You may make a collection consisting of the Document and other documents | |
22527 released under this License, and replace the individual copies of this | |
22528 License in the various documents with a single copy that is included in | |
22529 the collection, provided that you follow the rules of this License for | |
22530 verbatim copying of each of the documents in all other respects. | |
22531 | |
22532 You may extract a single document from such a collection, and distribute | |
22533 it individually under this License, provided you insert a copy of this | |
22534 License into the extracted document, and follow this License in all | |
22535 other respects regarding verbatim copying of that document. | |
22536 | |
22537 @item | |
22538 AGGREGATION WITH INDEPENDENT WORKS | |
22539 | |
22540 A compilation of the Document or its derivatives with other separate | |
22541 and independent documents or works, in or on a volume of a storage or | |
22542 distribution medium, is called an ``aggregate'' if the copyright | |
22543 resulting from the compilation is not used to limit the legal rights | |
22544 of the compilation's users beyond what the individual works permit. | |
22545 When the Document is included in an aggregate, this License does not | |
22546 apply to the other works in the aggregate which are not themselves | |
22547 derivative works of the Document. | |
22548 | |
22549 If the Cover Text requirement of section 3 is applicable to these | |
22550 copies of the Document, then if the Document is less than one half of | |
22551 the entire aggregate, the Document's Cover Texts may be placed on | |
22552 covers that bracket the Document within the aggregate, or the | |
22553 electronic equivalent of covers if the Document is in electronic form. | |
22554 Otherwise they must appear on printed covers that bracket the whole | |
22555 aggregate. | |
22556 | |
22557 @item | |
22558 TRANSLATION | |
22559 | |
22560 Translation is considered a kind of modification, so you may | |
22561 distribute translations of the Document under the terms of section 4. | |
22562 Replacing Invariant Sections with translations requires special | |
22563 permission from their copyright holders, but you may include | |
22564 translations of some or all Invariant Sections in addition to the | |
22565 original versions of these Invariant Sections. You may include a | |
22566 translation of this License, and all the license notices in the | |
22567 Document, and any Warranty Disclaimers, provided that you also include | |
22568 the original English version of this License and the original versions | |
22569 of those notices and disclaimers. In case of a disagreement between | |
22570 the translation and the original version of this License or a notice | |
22571 or disclaimer, the original version will prevail. | |
22572 | |
22573 If a section in the Document is Entitled ``Acknowledgements'', | |
22574 ``Dedications'', or ``History'', the requirement (section 4) to Preserve | |
22575 its Title (section 1) will typically require changing the actual | |
22576 title. | |
22577 | |
22578 @item | |
22579 TERMINATION | |
22580 | |
22581 You may not copy, modify, sublicense, or distribute the Document except | |
22582 as expressly provided for under this License. Any other attempt to | |
22583 copy, modify, sublicense or distribute the Document is void, and will | |
22584 automatically terminate your rights under this License. However, | |
22585 parties who have received copies, or rights, from you under this | |
22586 License will not have their licenses terminated so long as such | |
22587 parties remain in full compliance. | |
22588 | |
22589 @item | |
22590 FUTURE REVISIONS OF THIS LICENSE | |
22591 | |
22592 The Free Software Foundation may publish new, revised versions | |
22593 of the GNU Free Documentation License from time to time. Such new | |
22594 versions will be similar in spirit to the present version, but may | |
22595 differ in detail to address new problems or concerns. See | |
22596 @uref{http://www.gnu.org/copyleft/}. | |
22597 | |
22598 Each version of the License is given a distinguishing version number. | |
22599 If the Document specifies that a particular numbered version of this | |
22600 License ``or any later version'' applies to it, you have the option of | |
22601 following the terms and conditions either of that specified version or | |
22602 of any later version that has been published (not as a draft) by the | |
22603 Free Software Foundation. If the Document does not specify a version | |
22604 number of this License, you may choose any version ever published (not | |
22605 as a draft) by the Free Software Foundation. | |
22606 @end enumerate | |
22607 | |
22608 @page | |
22609 @appendixsubsec ADDENDUM: How to use this License for your documents | |
22610 | |
22611 To use this License in a document you have written, include a copy of | |
22612 the License in the document and put the following copyright and | |
22613 license notices just after the title page: | |
22614 | |
22615 @smallexample | |
22616 @group | |
22617 Copyright (C) @var{year} @var{your name}. | |
22618 Permission is granted to copy, distribute and/or modify this document | |
22619 under the terms of the GNU Free Documentation License, Version 1.2 | |
22620 or any later version published by the Free Software Foundation; | |
22621 with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. | |
22622 A copy of the license is included in the section entitled ``GNU | |
22623 Free Documentation License''. | |
22624 @end group | |
22625 @end smallexample | |
22626 | |
22627 If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, | |
22628 replace the ``with...Texts.'' line with this: | |
22629 | |
22630 @smallexample | |
22631 @group | |
22632 with the Invariant Sections being @var{list their titles}, with | |
22633 the Front-Cover Texts being @var{list}, and with the Back-Cover Texts | |
22634 being @var{list}. | |
22635 @end group | |
22636 @end smallexample | |
22637 | |
22638 If you have Invariant Sections without Cover Texts, or some other | |
22639 combination of the three, merge those two alternatives to suit the | |
22640 situation. | |
22641 | |
22642 If your document contains nontrivial examples of program code, we | |
22643 recommend releasing these examples in parallel under your choice of | |
22644 free software license, such as the GNU General Public License, | |
22645 to permit their use in free software. | |
22646 | |
22647 @node Index, About the Author, GNU Free Documentation License, Top | |
22648 @comment node-name, next, previous, up | |
22649 @unnumbered Index | |
22650 | |
22651 @ignore | |
22652 MENU ENTRY: NODE NAME. | |
22653 @end ignore | |
22654 | |
22655 @printindex cp | |
22656 | |
22657 @iftex | |
22658 @c Place biographical information on right-hand (verso) page | |
22659 | |
22660 @tex | |
22661 \ifodd\pageno | |
22662 \par\vfill\supereject | |
22663 \global\evenheadline={\hfil} \global\evenfootline={\hfil} | |
22664 \global\oddheadline={\hfil} \global\oddfootline={\hfil} | |
22665 \page\hbox{}\page | |
22666 \else | |
22667 \par\vfill\supereject | |
22668 \par\vfill\supereject | |
22669 \global\evenheadline={\hfil} \global\evenfootline={\hfil} | |
22670 \global\oddheadline={\hfil} \global\oddfootline={\hfil} | |
22671 \page\hbox{}\page | |
22672 \page\hbox{}\page | |
22673 \fi | |
22674 @end tex | |
22675 | |
22676 @page | |
22677 @w{ } | |
22678 | |
22679 @c ================ Biographical information ================ | |
22680 | |
22681 @w{ } | |
22682 @sp 8 | |
22683 @center About the Author | |
22684 @sp 1 | |
22685 @end iftex | |
22686 | |
22687 @ifnottex | |
22688 @node About the Author, , Index, Top | |
22689 @unnumbered About the Author | |
22690 @end ifnottex | |
22691 | |
22692 @quotation | |
22693 Robert J. Chassell has worked with GNU Emacs since 1985. He writes | |
22694 and edits, teaches Emacs and Emacs Lisp, and speaks throughout the | |
22695 world on software freedom. Chassell was a founding Director and | |
22696 Treasurer of the Free Software Foundation, Inc. He is co-author of | |
22697 the @cite{Texinfo} manual, and has edited more than a dozen other | |
22698 books. He graduated from Cambridge University, in England. He has an | |
22699 abiding interest in social and economic history and flies his own | |
22700 airplane. | |
22701 @end quotation | |
22702 | |
22703 @page | |
22704 @w{ } | |
22705 | |
22706 @c Prevent page number on blank verso, so eject it first. | |
22707 @tex | |
22708 \par\vfill\supereject | |
22709 @end tex | |
22710 | |
22711 @iftex | |
22712 @headings off | |
22713 @evenheading @thispage @| @| @thistitle | |
22714 @oddheading @| @| @thispage | |
22715 @end iftex | |
22716 | |
22717 @bye | |
22718 | |
22719 @ignore | |
22720 arch-tag: da1a2154-531f-43a8-8e33-fc7faad10acf | |
22721 @end ignore |