Mercurial > emacs
annotate lispref/control.texi @ 19630:95743e18a01c
(defgroup, defcustom): Add Emacs 19 compatibility definitions.
author | Richard M. Stallman <rms@gnu.org> |
---|---|
date | Sat, 30 Aug 1997 19:48:14 +0000 |
parents | ee4b53003fd5 |
children | 66d807bdc5b4 |
rev | line source |
---|---|
6453 | 1 @c -*-texinfo-*- |
2 @c This is part of the GNU Emacs Lisp Reference Manual. | |
3 @c Copyright (C) 1990, 1991, 1992, 1993, 1994 Free Software Foundation, Inc. | |
4 @c See the file elisp.texi for copying conditions. | |
5 @setfilename ../info/control | |
6 @node Control Structures, Variables, Evaluation, Top | |
7 @chapter Control Structures | |
8 @cindex special forms for control structures | |
9 @cindex control structures | |
10 | |
11 A Lisp program consists of expressions or @dfn{forms} (@pxref{Forms}). | |
12 We control the order of execution of the forms by enclosing them in | |
13 @dfn{control structures}. Control structures are special forms which | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
14 control when, whether, or how many times to execute the forms they |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
15 contain. |
6453 | 16 |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
17 The simplest order of execution is sequential execution: first form |
6453 | 18 @var{a}, then form @var{b}, and so on. This is what happens when you |
19 write several forms in succession in the body of a function, or at top | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
20 level in a file of Lisp code---the forms are executed in the order |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
21 written. We call this @dfn{textual order}. For example, if a function |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
22 body consists of two forms @var{a} and @var{b}, evaluation of the |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
23 function evaluates first @var{a} and then @var{b}, and the function's |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
24 value is the value of @var{b}. |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
25 |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
26 Explicit control structures make possible an order of execution other |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
27 than sequential. |
6453 | 28 |
29 Emacs Lisp provides several kinds of control structure, including | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
30 other varieties of sequencing, conditionals, iteration, and (controlled) |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
31 jumps---all discussed below. The built-in control structures are |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
32 special forms since their subforms are not necessarily evaluated or not |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
33 evaluated sequentially. You can use macros to define your own control |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
34 structure constructs (@pxref{Macros}). |
6453 | 35 |
36 @menu | |
37 * Sequencing:: Evaluation in textual order. | |
38 * Conditionals:: @code{if}, @code{cond}. | |
39 * Combining Conditions:: @code{and}, @code{or}, @code{not}. | |
40 * Iteration:: @code{while} loops. | |
41 * Nonlocal Exits:: Jumping out of a sequence. | |
42 @end menu | |
43 | |
44 @node Sequencing | |
45 @section Sequencing | |
46 | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
47 Evaluating forms in the order they appear is the most common way |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
48 control passes from one form to another. In some contexts, such as in a |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
49 function body, this happens automatically. Elsewhere you must use a |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
50 control structure construct to do this: @code{progn}, the simplest |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
51 control construct of Lisp. |
6453 | 52 |
53 A @code{progn} special form looks like this: | |
54 | |
55 @example | |
56 @group | |
57 (progn @var{a} @var{b} @var{c} @dots{}) | |
58 @end group | |
59 @end example | |
60 | |
61 @noindent | |
62 and it says to execute the forms @var{a}, @var{b}, @var{c} and so on, in | |
63 that order. These forms are called the body of the @code{progn} form. | |
64 The value of the last form in the body becomes the value of the entire | |
65 @code{progn}. | |
66 | |
67 @cindex implicit @code{progn} | |
68 In the early days of Lisp, @code{progn} was the only way to execute | |
69 two or more forms in succession and use the value of the last of them. | |
70 But programmers found they often needed to use a @code{progn} in the | |
71 body of a function, where (at that time) only one form was allowed. So | |
72 the body of a function was made into an ``implicit @code{progn}'': | |
73 several forms are allowed just as in the body of an actual @code{progn}. | |
74 Many other control structures likewise contain an implicit @code{progn}. | |
75 As a result, @code{progn} is not used as often as it used to be. It is | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
76 needed now most often inside an @code{unwind-protect}, @code{and}, |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
77 @code{or}, or in the @var{then}-part of an @code{if}. |
6453 | 78 |
79 @defspec progn forms@dots{} | |
80 This special form evaluates all of the @var{forms}, in textual | |
81 order, returning the result of the final form. | |
82 | |
83 @example | |
84 @group | |
85 (progn (print "The first form") | |
86 (print "The second form") | |
87 (print "The third form")) | |
88 @print{} "The first form" | |
89 @print{} "The second form" | |
90 @print{} "The third form" | |
91 @result{} "The third form" | |
92 @end group | |
93 @end example | |
94 @end defspec | |
95 | |
96 Two other control constructs likewise evaluate a series of forms but return | |
97 a different value: | |
98 | |
99 @defspec prog1 form1 forms@dots{} | |
100 This special form evaluates @var{form1} and all of the @var{forms}, in | |
101 textual order, returning the result of @var{form1}. | |
102 | |
103 @example | |
104 @group | |
105 (prog1 (print "The first form") | |
106 (print "The second form") | |
107 (print "The third form")) | |
108 @print{} "The first form" | |
109 @print{} "The second form" | |
110 @print{} "The third form" | |
111 @result{} "The first form" | |
112 @end group | |
113 @end example | |
114 | |
115 Here is a way to remove the first element from a list in the variable | |
116 @code{x}, then return the value of that former element: | |
117 | |
118 @example | |
119 (prog1 (car x) (setq x (cdr x))) | |
120 @end example | |
121 @end defspec | |
122 | |
123 @defspec prog2 form1 form2 forms@dots{} | |
124 This special form evaluates @var{form1}, @var{form2}, and all of the | |
125 following @var{forms}, in textual order, returning the result of | |
126 @var{form2}. | |
127 | |
128 @example | |
129 @group | |
130 (prog2 (print "The first form") | |
131 (print "The second form") | |
132 (print "The third form")) | |
133 @print{} "The first form" | |
134 @print{} "The second form" | |
135 @print{} "The third form" | |
136 @result{} "The second form" | |
137 @end group | |
138 @end example | |
139 @end defspec | |
140 | |
141 @node Conditionals | |
142 @section Conditionals | |
143 @cindex conditional evaluation | |
144 | |
145 Conditional control structures choose among alternatives. Emacs Lisp | |
16850 | 146 has four conditional forms: @code{if}, which is much the same as in |
147 other languages; @code{when} and @code{unless}, which are variants of | |
148 @code{if}; and @code{cond}, which is a generalized case statement. | |
6453 | 149 |
150 @defspec if condition then-form else-forms@dots{} | |
151 @code{if} chooses between the @var{then-form} and the @var{else-forms} | |
152 based on the value of @var{condition}. If the evaluated @var{condition} is | |
153 non-@code{nil}, @var{then-form} is evaluated and the result returned. | |
154 Otherwise, the @var{else-forms} are evaluated in textual order, and the | |
155 value of the last one is returned. (The @var{else} part of @code{if} is | |
156 an example of an implicit @code{progn}. @xref{Sequencing}.) | |
157 | |
158 If @var{condition} has the value @code{nil}, and no @var{else-forms} are | |
159 given, @code{if} returns @code{nil}. | |
160 | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
161 @code{if} is a special form because the branch that is not selected is |
6453 | 162 never evaluated---it is ignored. Thus, in the example below, |
163 @code{true} is not printed because @code{print} is never called. | |
164 | |
165 @example | |
166 @group | |
167 (if nil | |
168 (print 'true) | |
169 'very-false) | |
170 @result{} very-false | |
171 @end group | |
172 @end example | |
173 @end defspec | |
174 | |
16850 | 175 @defspec when condition then-forms@dots{} |
176 This is a variant of @code{if} where there are no @var{else-forms}, | |
177 and possibly several @var{then-forms}. In particular, | |
178 | |
179 @example | |
180 (when @var{condition} @var{a} @var{b} @var{c}) | |
181 @end example | |
182 | |
183 @noindent | |
184 is entirely equivalent to | |
185 | |
186 @example | |
187 (if @var{condition} (progn @var{a} @var{b} @var{c}) nil) | |
188 @end example | |
189 @end defspec | |
190 | |
191 @defspec unless condition forms@dots{} | |
192 This is a variant of @code{if} where there is no @var{then-form}: | |
193 | |
194 @example | |
195 (unless @var{condition} @var{a} @var{b} @var{c}) | |
196 @end example | |
197 | |
198 @noindent | |
199 is entirely equivalent to | |
200 | |
201 @example | |
202 (if @var{condition} nil | |
203 @var{a} @var{b} @var{c}) | |
204 @end example | |
205 @end defspec | |
206 | |
6453 | 207 @defspec cond clause@dots{} |
208 @code{cond} chooses among an arbitrary number of alternatives. Each | |
209 @var{clause} in the @code{cond} must be a list. The @sc{car} of this | |
210 list is the @var{condition}; the remaining elements, if any, the | |
211 @var{body-forms}. Thus, a clause looks like this: | |
212 | |
213 @example | |
214 (@var{condition} @var{body-forms}@dots{}) | |
215 @end example | |
216 | |
217 @code{cond} tries the clauses in textual order, by evaluating the | |
218 @var{condition} of each clause. If the value of @var{condition} is | |
219 non-@code{nil}, the clause ``succeeds''; then @code{cond} evaluates its | |
220 @var{body-forms}, and the value of the last of @var{body-forms} becomes | |
221 the value of the @code{cond}. The remaining clauses are ignored. | |
222 | |
223 If the value of @var{condition} is @code{nil}, the clause ``fails'', so | |
224 the @code{cond} moves on to the following clause, trying its | |
225 @var{condition}. | |
226 | |
227 If every @var{condition} evaluates to @code{nil}, so that every clause | |
228 fails, @code{cond} returns @code{nil}. | |
229 | |
230 A clause may also look like this: | |
231 | |
232 @example | |
233 (@var{condition}) | |
234 @end example | |
235 | |
236 @noindent | |
237 Then, if @var{condition} is non-@code{nil} when tested, the value of | |
238 @var{condition} becomes the value of the @code{cond} form. | |
239 | |
240 The following example has four clauses, which test for the cases where | |
241 the value of @code{x} is a number, string, buffer and symbol, | |
242 respectively: | |
243 | |
244 @example | |
245 @group | |
246 (cond ((numberp x) x) | |
247 ((stringp x) x) | |
248 ((bufferp x) | |
249 (setq temporary-hack x) ; @r{multiple body-forms} | |
250 (buffer-name x)) ; @r{in one clause} | |
251 ((symbolp x) (symbol-value x))) | |
252 @end group | |
253 @end example | |
254 | |
255 Often we want to execute the last clause whenever none of the previous | |
256 clauses was successful. To do this, we use @code{t} as the | |
257 @var{condition} of the last clause, like this: @code{(t | |
258 @var{body-forms})}. The form @code{t} evaluates to @code{t}, which is | |
259 never @code{nil}, so this clause never fails, provided the @code{cond} | |
260 gets to it at all. | |
261 | |
262 For example, | |
263 | |
264 @example | |
265 @group | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
266 (cond ((eq a 'hack) 'foo) |
6453 | 267 (t "default")) |
268 @result{} "default" | |
269 @end group | |
270 @end example | |
271 | |
272 @noindent | |
273 This expression is a @code{cond} which returns @code{foo} if the value | |
274 of @code{a} is 1, and returns the string @code{"default"} otherwise. | |
275 @end defspec | |
276 | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
277 Any conditional construct can be expressed with @code{cond} or with |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
278 @code{if}. Therefore, the choice between them is a matter of style. |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
279 For example: |
6453 | 280 |
281 @example | |
282 @group | |
283 (if @var{a} @var{b} @var{c}) | |
284 @equiv{} | |
285 (cond (@var{a} @var{b}) (t @var{c})) | |
286 @end group | |
287 @end example | |
288 | |
289 @node Combining Conditions | |
290 @section Constructs for Combining Conditions | |
291 | |
292 This section describes three constructs that are often used together | |
293 with @code{if} and @code{cond} to express complicated conditions. The | |
294 constructs @code{and} and @code{or} can also be used individually as | |
295 kinds of multiple conditional constructs. | |
296 | |
297 @defun not condition | |
298 This function tests for the falsehood of @var{condition}. It returns | |
299 @code{t} if @var{condition} is @code{nil}, and @code{nil} otherwise. | |
300 The function @code{not} is identical to @code{null}, and we recommend | |
301 using the name @code{null} if you are testing for an empty list. | |
302 @end defun | |
303 | |
304 @defspec and conditions@dots{} | |
305 The @code{and} special form tests whether all the @var{conditions} are | |
306 true. It works by evaluating the @var{conditions} one by one in the | |
307 order written. | |
308 | |
309 If any of the @var{conditions} evaluates to @code{nil}, then the result | |
310 of the @code{and} must be @code{nil} regardless of the remaining | |
311 @var{conditions}; so @code{and} returns right away, ignoring the | |
312 remaining @var{conditions}. | |
313 | |
314 If all the @var{conditions} turn out non-@code{nil}, then the value of | |
315 the last of them becomes the value of the @code{and} form. | |
316 | |
317 Here is an example. The first condition returns the integer 1, which is | |
318 not @code{nil}. Similarly, the second condition returns the integer 2, | |
319 which is not @code{nil}. The third condition is @code{nil}, so the | |
320 remaining condition is never evaluated. | |
321 | |
322 @example | |
323 @group | |
324 (and (print 1) (print 2) nil (print 3)) | |
325 @print{} 1 | |
326 @print{} 2 | |
327 @result{} nil | |
328 @end group | |
329 @end example | |
330 | |
331 Here is a more realistic example of using @code{and}: | |
332 | |
333 @example | |
334 @group | |
335 (if (and (consp foo) (eq (car foo) 'x)) | |
336 (message "foo is a list starting with x")) | |
337 @end group | |
338 @end example | |
339 | |
340 @noindent | |
341 Note that @code{(car foo)} is not executed if @code{(consp foo)} returns | |
342 @code{nil}, thus avoiding an error. | |
343 | |
344 @code{and} can be expressed in terms of either @code{if} or @code{cond}. | |
345 For example: | |
346 | |
347 @example | |
348 @group | |
349 (and @var{arg1} @var{arg2} @var{arg3}) | |
350 @equiv{} | |
351 (if @var{arg1} (if @var{arg2} @var{arg3})) | |
352 @equiv{} | |
353 (cond (@var{arg1} (cond (@var{arg2} @var{arg3})))) | |
354 @end group | |
355 @end example | |
356 @end defspec | |
357 | |
358 @defspec or conditions@dots{} | |
359 The @code{or} special form tests whether at least one of the | |
360 @var{conditions} is true. It works by evaluating all the | |
361 @var{conditions} one by one in the order written. | |
362 | |
363 If any of the @var{conditions} evaluates to a non-@code{nil} value, then | |
364 the result of the @code{or} must be non-@code{nil}; so @code{or} returns | |
365 right away, ignoring the remaining @var{conditions}. The value it | |
366 returns is the non-@code{nil} value of the condition just evaluated. | |
367 | |
368 If all the @var{conditions} turn out @code{nil}, then the @code{or} | |
369 expression returns @code{nil}. | |
370 | |
371 For example, this expression tests whether @code{x} is either 0 or | |
372 @code{nil}: | |
373 | |
374 @example | |
375 (or (eq x nil) (eq x 0)) | |
376 @end example | |
377 | |
378 Like the @code{and} construct, @code{or} can be written in terms of | |
379 @code{cond}. For example: | |
380 | |
381 @example | |
382 @group | |
383 (or @var{arg1} @var{arg2} @var{arg3}) | |
384 @equiv{} | |
385 (cond (@var{arg1}) | |
386 (@var{arg2}) | |
387 (@var{arg3})) | |
388 @end group | |
389 @end example | |
390 | |
391 You could almost write @code{or} in terms of @code{if}, but not quite: | |
392 | |
393 @example | |
394 @group | |
395 (if @var{arg1} @var{arg1} | |
396 (if @var{arg2} @var{arg2} | |
397 @var{arg3})) | |
398 @end group | |
399 @end example | |
400 | |
401 @noindent | |
402 This is not completely equivalent because it can evaluate @var{arg1} or | |
403 @var{arg2} twice. By contrast, @code{(or @var{arg1} @var{arg2} | |
404 @var{arg3})} never evaluates any argument more than once. | |
405 @end defspec | |
406 | |
407 @node Iteration | |
408 @section Iteration | |
409 @cindex iteration | |
410 @cindex recursion | |
411 | |
412 Iteration means executing part of a program repetitively. For | |
413 example, you might want to repeat some computation once for each element | |
414 of a list, or once for each integer from 0 to @var{n}. You can do this | |
415 in Emacs Lisp with the special form @code{while}: | |
416 | |
417 @defspec while condition forms@dots{} | |
418 @code{while} first evaluates @var{condition}. If the result is | |
419 non-@code{nil}, it evaluates @var{forms} in textual order. Then it | |
420 reevaluates @var{condition}, and if the result is non-@code{nil}, it | |
421 evaluates @var{forms} again. This process repeats until @var{condition} | |
422 evaluates to @code{nil}. | |
423 | |
424 There is no limit on the number of iterations that may occur. The loop | |
425 will continue until either @var{condition} evaluates to @code{nil} or | |
426 until an error or @code{throw} jumps out of it (@pxref{Nonlocal Exits}). | |
427 | |
428 The value of a @code{while} form is always @code{nil}. | |
429 | |
430 @example | |
431 @group | |
432 (setq num 0) | |
433 @result{} 0 | |
434 @end group | |
435 @group | |
436 (while (< num 4) | |
437 (princ (format "Iteration %d." num)) | |
438 (setq num (1+ num))) | |
439 @print{} Iteration 0. | |
440 @print{} Iteration 1. | |
441 @print{} Iteration 2. | |
442 @print{} Iteration 3. | |
443 @result{} nil | |
444 @end group | |
445 @end example | |
446 | |
447 If you would like to execute something on each iteration before the | |
448 end-test, put it together with the end-test in a @code{progn} as the | |
449 first argument of @code{while}, as shown here: | |
450 | |
451 @example | |
452 @group | |
453 (while (progn | |
454 (forward-line 1) | |
455 (not (looking-at "^$")))) | |
456 @end group | |
457 @end example | |
458 | |
459 @noindent | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
460 This moves forward one line and continues moving by lines until it |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
461 reaches an empty. It is unusual in that the @code{while} has no body, |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
462 just the end test (which also does the real work of moving point). |
6453 | 463 @end defspec |
464 | |
465 @node Nonlocal Exits | |
466 @section Nonlocal Exits | |
467 @cindex nonlocal exits | |
468 | |
469 A @dfn{nonlocal exit} is a transfer of control from one point in a | |
470 program to another remote point. Nonlocal exits can occur in Emacs Lisp | |
471 as a result of errors; you can also use them under explicit control. | |
472 Nonlocal exits unbind all variable bindings made by the constructs being | |
473 exited. | |
474 | |
475 @menu | |
476 * Catch and Throw:: Nonlocal exits for the program's own purposes. | |
477 * Examples of Catch:: Showing how such nonlocal exits can be written. | |
478 * Errors:: How errors are signaled and handled. | |
479 * Cleanups:: Arranging to run a cleanup form if an error happens. | |
480 @end menu | |
481 | |
482 @node Catch and Throw | |
483 @subsection Explicit Nonlocal Exits: @code{catch} and @code{throw} | |
484 | |
485 Most control constructs affect only the flow of control within the | |
486 construct itself. The function @code{throw} is the exception to this | |
487 rule of normal program execution: it performs a nonlocal exit on | |
488 request. (There are other exceptions, but they are for error handling | |
489 only.) @code{throw} is used inside a @code{catch}, and jumps back to | |
490 that @code{catch}. For example: | |
491 | |
492 @example | |
493 @group | |
494 (catch 'foo | |
495 (progn | |
496 @dots{} | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
497 (throw 'foo t) |
6453 | 498 @dots{})) |
499 @end group | |
500 @end example | |
501 | |
502 @noindent | |
503 The @code{throw} transfers control straight back to the corresponding | |
504 @code{catch}, which returns immediately. The code following the | |
505 @code{throw} is not executed. The second argument of @code{throw} is used | |
506 as the return value of the @code{catch}. | |
507 | |
508 The @code{throw} and the @code{catch} are matched through the first | |
509 argument: @code{throw} searches for a @code{catch} whose first argument | |
510 is @code{eq} to the one specified. Thus, in the above example, the | |
511 @code{throw} specifies @code{foo}, and the @code{catch} specifies the | |
512 same symbol, so that @code{catch} is applicable. If there is more than | |
513 one applicable @code{catch}, the innermost one takes precedence. | |
514 | |
515 Executing @code{throw} exits all Lisp constructs up to the matching | |
516 @code{catch}, including function calls. When binding constructs such as | |
517 @code{let} or function calls are exited in this way, the bindings are | |
518 unbound, just as they are when these constructs exit normally | |
519 (@pxref{Local Variables}). Likewise, @code{throw} restores the buffer | |
520 and position saved by @code{save-excursion} (@pxref{Excursions}), and | |
521 the narrowing status saved by @code{save-restriction} and the window | |
522 selection saved by @code{save-window-excursion} (@pxref{Window | |
523 Configurations}). It also runs any cleanups established with the | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
524 @code{unwind-protect} special form when it exits that form |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
525 (@pxref{Cleanups}). |
6453 | 526 |
527 The @code{throw} need not appear lexically within the @code{catch} | |
528 that it jumps to. It can equally well be called from another function | |
529 called within the @code{catch}. As long as the @code{throw} takes place | |
530 chronologically after entry to the @code{catch}, and chronologically | |
531 before exit from it, it has access to that @code{catch}. This is why | |
532 @code{throw} can be used in commands such as @code{exit-recursive-edit} | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
533 that throw back to the editor command loop (@pxref{Recursive Editing}). |
6453 | 534 |
535 @cindex CL note---only @code{throw} in Emacs | |
536 @quotation | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
537 @b{Common Lisp note:} Most other versions of Lisp, including Common Lisp, |
6453 | 538 have several ways of transferring control nonsequentially: @code{return}, |
539 @code{return-from}, and @code{go}, for example. Emacs Lisp has only | |
540 @code{throw}. | |
541 @end quotation | |
542 | |
543 @defspec catch tag body@dots{} | |
544 @cindex tag on run time stack | |
545 @code{catch} establishes a return point for the @code{throw} function. The | |
546 return point is distinguished from other such return points by @var{tag}, | |
547 which may be any Lisp object. The argument @var{tag} is evaluated normally | |
548 before the return point is established. | |
549 | |
550 With the return point in effect, @code{catch} evaluates the forms of the | |
551 @var{body} in textual order. If the forms execute normally, without | |
552 error or nonlocal exit, the value of the last body form is returned from | |
553 the @code{catch}. | |
554 | |
555 If a @code{throw} is done within @var{body} specifying the same value | |
556 @var{tag}, the @code{catch} exits immediately; the value it returns is | |
557 whatever was specified as the second argument of @code{throw}. | |
558 @end defspec | |
559 | |
560 @defun throw tag value | |
561 The purpose of @code{throw} is to return from a return point previously | |
562 established with @code{catch}. The argument @var{tag} is used to choose | |
563 among the various existing return points; it must be @code{eq} to the value | |
564 specified in the @code{catch}. If multiple return points match @var{tag}, | |
565 the innermost one is used. | |
566 | |
567 The argument @var{value} is used as the value to return from that | |
568 @code{catch}. | |
569 | |
570 @kindex no-catch | |
571 If no return point is in effect with tag @var{tag}, then a @code{no-catch} | |
572 error is signaled with data @code{(@var{tag} @var{value})}. | |
573 @end defun | |
574 | |
575 @node Examples of Catch | |
576 @subsection Examples of @code{catch} and @code{throw} | |
577 | |
578 One way to use @code{catch} and @code{throw} is to exit from a doubly | |
579 nested loop. (In most languages, this would be done with a ``go to''.) | |
580 Here we compute @code{(foo @var{i} @var{j})} for @var{i} and @var{j} | |
581 varying from 0 to 9: | |
582 | |
583 @example | |
584 @group | |
585 (defun search-foo () | |
586 (catch 'loop | |
587 (let ((i 0)) | |
588 (while (< i 10) | |
589 (let ((j 0)) | |
590 (while (< j 10) | |
591 (if (foo i j) | |
592 (throw 'loop (list i j))) | |
593 (setq j (1+ j)))) | |
594 (setq i (1+ i)))))) | |
595 @end group | |
596 @end example | |
597 | |
598 @noindent | |
599 If @code{foo} ever returns non-@code{nil}, we stop immediately and return a | |
600 list of @var{i} and @var{j}. If @code{foo} always returns @code{nil}, the | |
601 @code{catch} returns normally, and the value is @code{nil}, since that | |
602 is the result of the @code{while}. | |
603 | |
604 Here are two tricky examples, slightly different, showing two | |
605 return points at once. First, two return points with the same tag, | |
606 @code{hack}: | |
607 | |
608 @example | |
609 @group | |
610 (defun catch2 (tag) | |
611 (catch tag | |
612 (throw 'hack 'yes))) | |
613 @result{} catch2 | |
614 @end group | |
615 | |
616 @group | |
617 (catch 'hack | |
618 (print (catch2 'hack)) | |
619 'no) | |
620 @print{} yes | |
621 @result{} no | |
622 @end group | |
623 @end example | |
624 | |
625 @noindent | |
626 Since both return points have tags that match the @code{throw}, it goes to | |
627 the inner one, the one established in @code{catch2}. Therefore, | |
628 @code{catch2} returns normally with value @code{yes}, and this value is | |
629 printed. Finally the second body form in the outer @code{catch}, which is | |
630 @code{'no}, is evaluated and returned from the outer @code{catch}. | |
631 | |
632 Now let's change the argument given to @code{catch2}: | |
633 | |
634 @example | |
635 @group | |
636 (defun catch2 (tag) | |
637 (catch tag | |
638 (throw 'hack 'yes))) | |
639 @result{} catch2 | |
640 @end group | |
641 | |
642 @group | |
643 (catch 'hack | |
644 (print (catch2 'quux)) | |
645 'no) | |
646 @result{} yes | |
647 @end group | |
648 @end example | |
649 | |
650 @noindent | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
651 We still have two return points, but this time only the outer one has |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
652 the tag @code{hack}; the inner one has the tag @code{quux} instead. |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
653 Therefore, @code{throw} makes the outer @code{catch} return the value |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
654 @code{yes}. The function @code{print} is never called, and the |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
655 body-form @code{'no} is never evaluated. |
6453 | 656 |
657 @node Errors | |
658 @subsection Errors | |
659 @cindex errors | |
660 | |
661 When Emacs Lisp attempts to evaluate a form that, for some reason, | |
662 cannot be evaluated, it @dfn{signals} an @dfn{error}. | |
663 | |
664 When an error is signaled, Emacs's default reaction is to print an | |
665 error message and terminate execution of the current command. This is | |
666 the right thing to do in most cases, such as if you type @kbd{C-f} at | |
667 the end of the buffer. | |
668 | |
669 In complicated programs, simple termination may not be what you want. | |
670 For example, the program may have made temporary changes in data | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
671 structures, or created temporary buffers that should be deleted before |
6453 | 672 the program is finished. In such cases, you would use |
673 @code{unwind-protect} to establish @dfn{cleanup expressions} to be | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
674 evaluated in case of error. (@xref{Cleanups}.) Occasionally, you may |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
675 wish the program to continue execution despite an error in a subroutine. |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
676 In these cases, you would use @code{condition-case} to establish |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
677 @dfn{error handlers} to recover control in case of error. |
6453 | 678 |
679 Resist the temptation to use error handling to transfer control from | |
680 one part of the program to another; use @code{catch} and @code{throw} | |
681 instead. @xref{Catch and Throw}. | |
682 | |
683 @menu | |
684 * Signaling Errors:: How to report an error. | |
685 * Processing of Errors:: What Emacs does when you report an error. | |
686 * Handling Errors:: How you can trap errors and continue execution. | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
687 * Error Symbols:: How errors are classified for trapping them. |
6453 | 688 @end menu |
689 | |
690 @node Signaling Errors | |
691 @subsubsection How to Signal an Error | |
692 @cindex signaling errors | |
693 | |
694 Most errors are signaled ``automatically'' within Lisp primitives | |
695 which you call for other purposes, such as if you try to take the | |
696 @sc{car} of an integer or move forward a character at the end of the | |
697 buffer; you can also signal errors explicitly with the functions | |
698 @code{error} and @code{signal}. | |
699 | |
700 Quitting, which happens when the user types @kbd{C-g}, is not | |
701 considered an error, but it is handled almost like an error. | |
702 @xref{Quitting}. | |
703 | |
704 @defun error format-string &rest args | |
705 This function signals an error with an error message constructed by | |
706 applying @code{format} (@pxref{String Conversion}) to | |
707 @var{format-string} and @var{args}. | |
708 | |
709 These examples show typical uses of @code{error}: | |
710 | |
711 @example | |
712 @group | |
713 (error "You have committed an error. | |
714 Try something else.") | |
715 @error{} You have committed an error. | |
716 Try something else. | |
717 @end group | |
718 | |
719 @group | |
720 (error "You have committed %d errors." 10) | |
721 @error{} You have committed 10 errors. | |
722 @end group | |
723 @end example | |
724 | |
725 @code{error} works by calling @code{signal} with two arguments: the | |
726 error symbol @code{error}, and a list containing the string returned by | |
727 @code{format}. | |
728 | |
729 If you want to use your own string as an error message verbatim, don't | |
730 just write @code{(error @var{string})}. If @var{string} contains | |
731 @samp{%}, it will be interpreted as a format specifier, with undesirable | |
732 results. Instead, use @code{(error "%s" @var{string})}. | |
733 @end defun | |
734 | |
735 @defun signal error-symbol data | |
736 This function signals an error named by @var{error-symbol}. The | |
737 argument @var{data} is a list of additional Lisp objects relevant to the | |
738 circumstances of the error. | |
739 | |
740 The argument @var{error-symbol} must be an @dfn{error symbol}---a symbol | |
741 bearing a property @code{error-conditions} whose value is a list of | |
742 condition names. This is how Emacs Lisp classifies different sorts of | |
743 errors. | |
744 | |
745 The number and significance of the objects in @var{data} depends on | |
746 @var{error-symbol}. For example, with a @code{wrong-type-arg} error, | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
747 there are two objects in the list: a predicate that describes the type |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
748 that was expected, and the object that failed to fit that type. |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
749 @xref{Error Symbols}, for a description of error symbols. |
6453 | 750 |
751 Both @var{error-symbol} and @var{data} are available to any error | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
752 handlers that handle the error: @code{condition-case} binds a local |
6453 | 753 variable to a list of the form @code{(@var{error-symbol} .@: |
754 @var{data})} (@pxref{Handling Errors}). If the error is not handled, | |
755 these two values are used in printing the error message. | |
756 | |
757 The function @code{signal} never returns (though in older Emacs versions | |
758 it could sometimes return). | |
759 | |
760 @smallexample | |
761 @group | |
762 (signal 'wrong-number-of-arguments '(x y)) | |
763 @error{} Wrong number of arguments: x, y | |
764 @end group | |
765 | |
766 @group | |
767 (signal 'no-such-error '("My unknown error condition.")) | |
768 @error{} peculiar error: "My unknown error condition." | |
769 @end group | |
770 @end smallexample | |
771 @end defun | |
772 | |
773 @cindex CL note---no continuable errors | |
774 @quotation | |
775 @b{Common Lisp note:} Emacs Lisp has nothing like the Common Lisp | |
776 concept of continuable errors. | |
777 @end quotation | |
778 | |
779 @node Processing of Errors | |
780 @subsubsection How Emacs Processes Errors | |
781 | |
782 When an error is signaled, @code{signal} searches for an active | |
783 @dfn{handler} for the error. A handler is a sequence of Lisp | |
784 expressions designated to be executed if an error happens in part of the | |
785 Lisp program. If the error has an applicable handler, the handler is | |
786 executed, and control resumes following the handler. The handler | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
787 executes in the environment of the @code{condition-case} that |
6453 | 788 established it; all functions called within that @code{condition-case} |
789 have already been exited, and the handler cannot return to them. | |
790 | |
791 If there is no applicable handler for the error, the current command is | |
792 terminated and control returns to the editor command loop, because the | |
793 command loop has an implicit handler for all kinds of errors. The | |
794 command loop's handler uses the error symbol and associated data to | |
795 print an error message. | |
796 | |
797 @cindex @code{debug-on-error} use | |
798 An error that has no explicit handler may call the Lisp debugger. The | |
799 debugger is enabled if the variable @code{debug-on-error} (@pxref{Error | |
800 Debugging}) is non-@code{nil}. Unlike error handlers, the debugger runs | |
801 in the environment of the error, so that you can examine values of | |
802 variables precisely as they were at the time of the error. | |
803 | |
804 @node Handling Errors | |
805 @subsubsection Writing Code to Handle Errors | |
806 @cindex error handler | |
807 @cindex handling errors | |
808 | |
809 The usual effect of signaling an error is to terminate the command | |
810 that is running and return immediately to the Emacs editor command loop. | |
811 You can arrange to trap errors occurring in a part of your program by | |
812 establishing an error handler, with the special form | |
813 @code{condition-case}. A simple example looks like this: | |
814 | |
815 @example | |
816 @group | |
817 (condition-case nil | |
818 (delete-file filename) | |
819 (error nil)) | |
820 @end group | |
821 @end example | |
822 | |
823 @noindent | |
824 This deletes the file named @var{filename}, catching any error and | |
825 returning @code{nil} if an error occurs. | |
826 | |
827 The second argument of @code{condition-case} is called the | |
828 @dfn{protected form}. (In the example above, the protected form is a | |
829 call to @code{delete-file}.) The error handlers go into effect when | |
830 this form begins execution and are deactivated when this form returns. | |
831 They remain in effect for all the intervening time. In particular, they | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
832 are in effect during the execution of functions called by this form, in |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
833 their subroutines, and so on. This is a good thing, since, strictly |
6453 | 834 speaking, errors can be signaled only by Lisp primitives (including |
835 @code{signal} and @code{error}) called by the protected form, not by the | |
836 protected form itself. | |
837 | |
838 The arguments after the protected form are handlers. Each handler | |
839 lists one or more @dfn{condition names} (which are symbols) to specify | |
840 which errors it will handle. The error symbol specified when an error | |
841 is signaled also defines a list of condition names. A handler applies | |
842 to an error if they have any condition names in common. In the example | |
843 above, there is one handler, and it specifies one condition name, | |
844 @code{error}, which covers all errors. | |
845 | |
846 The search for an applicable handler checks all the established handlers | |
847 starting with the most recently established one. Thus, if two nested | |
848 @code{condition-case} forms offer to handle the same error, the inner of | |
849 the two will actually handle it. | |
850 | |
851 When an error is handled, control returns to the handler. Before this | |
852 happens, Emacs unbinds all variable bindings made by binding constructs | |
853 that are being exited and executes the cleanups of all | |
854 @code{unwind-protect} forms that are exited. Once control arrives at | |
855 the handler, the body of the handler is executed. | |
856 | |
16736
981e116b4ac6
Minor cleanups for overfull hboxes.
Richard M. Stallman <rms@gnu.org>
parents:
15725
diff
changeset
|
857 After execution of the handler body, execution returns from the |
981e116b4ac6
Minor cleanups for overfull hboxes.
Richard M. Stallman <rms@gnu.org>
parents:
15725
diff
changeset
|
858 @code{condition-case} form. Because the protected form is exited |
981e116b4ac6
Minor cleanups for overfull hboxes.
Richard M. Stallman <rms@gnu.org>
parents:
15725
diff
changeset
|
859 completely before execution of the handler, the handler cannot resume |
981e116b4ac6
Minor cleanups for overfull hboxes.
Richard M. Stallman <rms@gnu.org>
parents:
15725
diff
changeset
|
860 execution at the point of the error, nor can it examine variable |
6453 | 861 bindings that were made within the protected form. All it can do is |
862 clean up and proceed. | |
863 | |
864 @code{condition-case} is often used to trap errors that are | |
865 predictable, such as failure to open a file in a call to | |
866 @code{insert-file-contents}. It is also used to trap errors that are | |
867 totally unpredictable, such as when the program evaluates an expression | |
868 read from the user. | |
869 | |
870 Error signaling and handling have some resemblance to @code{throw} and | |
871 @code{catch}, but they are entirely separate facilities. An error | |
872 cannot be caught by a @code{catch}, and a @code{throw} cannot be handled | |
873 by an error handler (though using @code{throw} when there is no suitable | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
874 @code{catch} signals an error that can be handled). |
6453 | 875 |
876 @defspec condition-case var protected-form handlers@dots{} | |
877 This special form establishes the error handlers @var{handlers} around | |
878 the execution of @var{protected-form}. If @var{protected-form} executes | |
879 without error, the value it returns becomes the value of the | |
880 @code{condition-case} form; in this case, the @code{condition-case} has | |
881 no effect. The @code{condition-case} form makes a difference when an | |
882 error occurs during @var{protected-form}. | |
883 | |
884 Each of the @var{handlers} is a list of the form @code{(@var{conditions} | |
885 @var{body}@dots{})}. Here @var{conditions} is an error condition name | |
886 to be handled, or a list of condition names; @var{body} is one or more | |
887 Lisp expressions to be executed when this handler handles an error. | |
888 Here are examples of handlers: | |
889 | |
890 @smallexample | |
891 @group | |
892 (error nil) | |
893 | |
894 (arith-error (message "Division by zero")) | |
895 | |
896 ((arith-error file-error) | |
897 (message | |
898 "Either division by zero or failure to open a file")) | |
899 @end group | |
900 @end smallexample | |
901 | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
902 Each error that occurs has an @dfn{error symbol} that describes what |
6453 | 903 kind of error it is. The @code{error-conditions} property of this |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
904 symbol is a list of condition names (@pxref{Error Symbols}). Emacs |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
905 searches all the active @code{condition-case} forms for a handler that |
6453 | 906 specifies one or more of these condition names; the innermost matching |
907 @code{condition-case} handles the error. Within this | |
908 @code{condition-case}, the first applicable handler handles the error. | |
909 | |
910 After executing the body of the handler, the @code{condition-case} | |
911 returns normally, using the value of the last form in the handler body | |
912 as the overall value. | |
913 | |
15725 | 914 @cindex error description |
6453 | 915 The argument @var{var} is a variable. @code{condition-case} does not |
916 bind this variable when executing the @var{protected-form}, only when it | |
15725 | 917 handles an error. At that time, it binds @var{var} locally to an |
918 @dfn{error description}, which is a list giving the particulars of the | |
919 error. The error description has the form @code{(@var{error-symbol} | |
920 . @var{data})}. The handler can refer to this list to decide what to | |
921 do. For example, if the error is for failure opening a file, the file | |
922 name is the second element of @var{data}---the third element of the | |
923 error description. | |
6453 | 924 |
925 If @var{var} is @code{nil}, that means no variable is bound. Then the | |
926 error symbol and associated data are not available to the handler. | |
927 @end defspec | |
928 | |
15725 | 929 @defun error-message-string error-description |
930 This function returns the error message string for a given error | |
931 descriptor. It is useful if you want to handle an error by printing the | |
932 usual error message for that error. | |
933 @end defun | |
934 | |
6453 | 935 @cindex @code{arith-error} example |
936 Here is an example of using @code{condition-case} to handle the error | |
15725 | 937 that results from dividing by zero. The handler displays the error |
938 message (but without a beep), then returns a very large number. | |
6453 | 939 |
940 @smallexample | |
941 @group | |
942 (defun safe-divide (dividend divisor) | |
943 (condition-case err | |
944 ;; @r{Protected form.} | |
945 (/ dividend divisor) | |
946 ;; @r{The handler.} | |
947 (arith-error ; @r{Condition.} | |
15725 | 948 ;; @r{Display the usual message for this error.} |
949 (message "%s" (error-message-string err)) | |
6453 | 950 1000000))) |
951 @result{} safe-divide | |
952 @end group | |
953 | |
954 @group | |
955 (safe-divide 5 0) | |
956 @print{} Arithmetic error: (arith-error) | |
957 @result{} 1000000 | |
958 @end group | |
959 @end smallexample | |
960 | |
961 @noindent | |
962 The handler specifies condition name @code{arith-error} so that it will handle only division-by-zero errors. Other kinds of errors will not be handled, at least not by this @code{condition-case}. Thus, | |
963 | |
964 @smallexample | |
965 @group | |
966 (safe-divide nil 3) | |
967 @error{} Wrong type argument: integer-or-marker-p, nil | |
968 @end group | |
969 @end smallexample | |
970 | |
971 Here is a @code{condition-case} that catches all kinds of errors, | |
972 including those signaled with @code{error}: | |
973 | |
974 @smallexample | |
975 @group | |
976 (setq baz 34) | |
977 @result{} 34 | |
978 @end group | |
979 | |
980 @group | |
981 (condition-case err | |
982 (if (eq baz 35) | |
983 t | |
984 ;; @r{This is a call to the function @code{error}.} | |
12098 | 985 (error "Rats! The variable %s was %s, not 35" 'baz baz)) |
6453 | 986 ;; @r{This is the handler; it is not a form.} |
987 (error (princ (format "The error was: %s" err)) | |
988 2)) | |
12098 | 989 @print{} The error was: (error "Rats! The variable baz was 34, not 35") |
6453 | 990 @result{} 2 |
991 @end group | |
992 @end smallexample | |
993 | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
994 @node Error Symbols |
6453 | 995 @subsubsection Error Symbols and Condition Names |
996 @cindex error symbol | |
997 @cindex error name | |
998 @cindex condition name | |
999 @cindex user-defined error | |
1000 @kindex error-conditions | |
1001 | |
1002 When you signal an error, you specify an @dfn{error symbol} to specify | |
1003 the kind of error you have in mind. Each error has one and only one | |
1004 error symbol to categorize it. This is the finest classification of | |
1005 errors defined by the Emacs Lisp language. | |
1006 | |
1007 These narrow classifications are grouped into a hierarchy of wider | |
1008 classes called @dfn{error conditions}, identified by @dfn{condition | |
1009 names}. The narrowest such classes belong to the error symbols | |
1010 themselves: each error symbol is also a condition name. There are also | |
1011 condition names for more extensive classes, up to the condition name | |
1012 @code{error} which takes in all kinds of errors. Thus, each error has | |
1013 one or more condition names: @code{error}, the error symbol if that | |
1014 is distinct from @code{error}, and perhaps some intermediate | |
1015 classifications. | |
1016 | |
1017 In order for a symbol to be an error symbol, it must have an | |
1018 @code{error-conditions} property which gives a list of condition names. | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1019 This list defines the conditions that this kind of error belongs to. |
6453 | 1020 (The error symbol itself, and the symbol @code{error}, should always be |
1021 members of this list.) Thus, the hierarchy of condition names is | |
1022 defined by the @code{error-conditions} properties of the error symbols. | |
1023 | |
1024 In addition to the @code{error-conditions} list, the error symbol | |
1025 should have an @code{error-message} property whose value is a string to | |
1026 be printed when that error is signaled but not handled. If the | |
1027 @code{error-message} property exists, but is not a string, the error | |
1028 message @samp{peculiar error} is used. | |
1029 @cindex peculiar error | |
1030 | |
1031 Here is how we define a new error symbol, @code{new-error}: | |
1032 | |
1033 @example | |
1034 @group | |
1035 (put 'new-error | |
1036 'error-conditions | |
1037 '(error my-own-errors new-error)) | |
1038 @result{} (error my-own-errors new-error) | |
1039 @end group | |
1040 @group | |
1041 (put 'new-error 'error-message "A new error") | |
1042 @result{} "A new error" | |
1043 @end group | |
1044 @end example | |
1045 | |
1046 @noindent | |
1047 This error has three condition names: @code{new-error}, the narrowest | |
1048 classification; @code{my-own-errors}, which we imagine is a wider | |
1049 classification; and @code{error}, which is the widest of all. | |
12098 | 1050 |
1051 The error string should start with a capital letter but it should | |
1052 not end with a period. This is for consistency with the rest of Emacs. | |
6453 | 1053 |
1054 Naturally, Emacs will never signal @code{new-error} on its own; only | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1055 an explicit call to @code{signal} (@pxref{Signaling Errors}) in your |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1056 code can do this: |
6453 | 1057 |
1058 @example | |
1059 @group | |
1060 (signal 'new-error '(x y)) | |
1061 @error{} A new error: x, y | |
1062 @end group | |
1063 @end example | |
1064 | |
1065 This error can be handled through any of the three condition names. | |
1066 This example handles @code{new-error} and any other errors in the class | |
1067 @code{my-own-errors}: | |
1068 | |
1069 @example | |
1070 @group | |
1071 (condition-case foo | |
1072 (bar nil t) | |
1073 (my-own-errors nil)) | |
1074 @end group | |
1075 @end example | |
1076 | |
1077 The significant way that errors are classified is by their condition | |
1078 names---the names used to match errors with handlers. An error symbol | |
1079 serves only as a convenient way to specify the intended error message | |
1080 and list of condition names. It would be cumbersome to give | |
1081 @code{signal} a list of condition names rather than one error symbol. | |
1082 | |
1083 By contrast, using only error symbols without condition names would | |
1084 seriously decrease the power of @code{condition-case}. Condition names | |
1085 make it possible to categorize errors at various levels of generality | |
1086 when you write an error handler. Using error symbols alone would | |
1087 eliminate all but the narrowest level of classification. | |
1088 | |
1089 @xref{Standard Errors}, for a list of all the standard error symbols | |
1090 and their conditions. | |
1091 | |
1092 @node Cleanups | |
1093 @subsection Cleaning Up from Nonlocal Exits | |
1094 | |
1095 The @code{unwind-protect} construct is essential whenever you | |
1096 temporarily put a data structure in an inconsistent state; it permits | |
1097 you to ensure the data are consistent in the event of an error or throw. | |
1098 | |
1099 @defspec unwind-protect body cleanup-forms@dots{} | |
1100 @cindex cleanup forms | |
1101 @cindex protected forms | |
1102 @cindex error cleanup | |
1103 @cindex unwinding | |
1104 @code{unwind-protect} executes the @var{body} with a guarantee that the | |
1105 @var{cleanup-forms} will be evaluated if control leaves @var{body}, no | |
1106 matter how that happens. The @var{body} may complete normally, or | |
1107 execute a @code{throw} out of the @code{unwind-protect}, or cause an | |
1108 error; in all cases, the @var{cleanup-forms} will be evaluated. | |
1109 | |
1110 If the @var{body} forms finish normally, @code{unwind-protect} returns | |
1111 the value of the last @var{body} form, after it evaluates the | |
1112 @var{cleanup-forms}. If the @var{body} forms do not finish, | |
1113 @code{unwind-protect} does not return any value in the normal sense. | |
1114 | |
1115 Only the @var{body} is actually protected by the @code{unwind-protect}. | |
1116 If any of the @var{cleanup-forms} themselves exits nonlocally (e.g., via | |
1117 a @code{throw} or an error), @code{unwind-protect} is @emph{not} | |
1118 guaranteed to evaluate the rest of them. If the failure of one of the | |
1119 @var{cleanup-forms} has the potential to cause trouble, then protect it | |
1120 with another @code{unwind-protect} around that form. | |
1121 | |
1122 The number of currently active @code{unwind-protect} forms counts, | |
1123 together with the number of local variable bindings, against the limit | |
1124 @code{max-specpdl-size} (@pxref{Local Variables}). | |
1125 @end defspec | |
1126 | |
1127 For example, here we make an invisible buffer for temporary use, and | |
1128 make sure to kill it before finishing: | |
1129 | |
1130 @smallexample | |
1131 @group | |
1132 (save-excursion | |
1133 (let ((buffer (get-buffer-create " *temp*"))) | |
1134 (set-buffer buffer) | |
1135 (unwind-protect | |
1136 @var{body} | |
1137 (kill-buffer buffer)))) | |
1138 @end group | |
1139 @end smallexample | |
1140 | |
1141 @noindent | |
1142 You might think that we could just as well write @code{(kill-buffer | |
1143 (current-buffer))} and dispense with the variable @code{buffer}. | |
1144 However, the way shown above is safer, if @var{body} happens to get an | |
1145 error after switching to a different buffer! (Alternatively, you could | |
1146 write another @code{save-excursion} around the body, to ensure that the | |
1147 temporary buffer becomes current in time to kill it.) | |
1148 | |
1149 @findex ftp-login | |
7337
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1150 Here is an actual example taken from the file @file{ftp.el}. It |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1151 creates a process (@pxref{Processes}) to try to establish a connection |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1152 to a remote machine. As the function @code{ftp-login} is highly |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1153 susceptible to numerous problems that the writer of the function cannot |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1154 anticipate, it is protected with a form that guarantees deletion of the |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1155 process in the event of failure. Otherwise, Emacs might fill up with |
cd57cd335fff
*** empty log message ***
Richard M. Stallman <rms@gnu.org>
parents:
6453
diff
changeset
|
1156 useless subprocesses. |
6453 | 1157 |
1158 @smallexample | |
1159 @group | |
1160 (let ((win nil)) | |
1161 (unwind-protect | |
1162 (progn | |
1163 (setq process (ftp-setup-buffer host file)) | |
1164 (if (setq win (ftp-login process host user password)) | |
1165 (message "Logged in") | |
1166 (error "Ftp login failed"))) | |
1167 (or win (and process (delete-process process))))) | |
1168 @end group | |
1169 @end smallexample | |
1170 | |
1171 This example actually has a small bug: if the user types @kbd{C-g} to | |
1172 quit, and the quit happens immediately after the function | |
1173 @code{ftp-setup-buffer} returns but before the variable @code{process} is | |
1174 set, the process will not be killed. There is no easy way to fix this bug, | |
1175 but at least it is very unlikely. | |
1176 | |
1177 Here is another example which uses @code{unwind-protect} to make sure | |
1178 to kill a temporary buffer. In this example, the value returned by | |
1179 @code{unwind-protect} is used. | |
1180 | |
7734 | 1181 @smallexample |
6453 | 1182 (defun shell-command-string (cmd) |
1183 "Return the output of the shell command CMD, as a string." | |
1184 (save-excursion | |
1185 (set-buffer (generate-new-buffer " OS*cmd")) | |
1186 (shell-command cmd t) | |
1187 (unwind-protect | |
1188 (buffer-string) | |
1189 (kill-buffer (current-buffer))))) | |
7734 | 1190 @end smallexample |