view lispref/two-volume-cross-refs.txt @ 61126:c0aa521e0ca7

Revision: miles@gnu.org--gnu-2005/emacs--cvs-trunk--0--patch-220 Merge from gnus--rel--5.10 Patches applied: * gnus--rel--5.10 (patch 45-52) - Update from CVS - Update from CVS: texi Makefile.in CVS keyw cruft - Update from CVS: ChangeLog tweaks 2005-03-29 Reiner Steib <Reiner.Steib@gmx.de> * etc/gnus-refcard.tex, etc/gnus-logo.eps: New files. 2005-03-25 Katsumi Yamaoka <yamaoka@jpl.org> * lisp/gnus/message.el (message-resend): Bind rfc2047-encode-encoded-words. * lisp/gnus/mm-util.el (mm-replace-in-string): New function. (mm-xemacs-find-mime-charset-1): Ignore errors while loading latin-unity, which cannot be used with XEmacs 21.1. * lisp/gnus/rfc2047.el (rfc2047-encode-function-alist): Rename from rfc2047-encoding-function-alist in order to avoid conflicting with the old version. (rfc2047-encode-message-header): Remove useless goto-char. (rfc2047-encodable-p): Don't move point. (rfc2047-syntax-table): Treat `(' and `)' as is. (rfc2047-encode-region): Concatenate words containing non-ASCII characters in structured fields; don't encode space-delimited ASCII words even in unstructured fields; don't break words at char-category boundaries; encode encoded words in structured fields; treat text within parentheses as special; show the original text when error has occurred; move point to the end of the region after encoding, suggested by IRIE Tetsuya <irie@t.email.ne.jp>; treat backslash-quoted characters as non-special; check carefully whether to encode special characters; fix some kind of misconfigured headers; signal a real error if debug-on-quit or debug-on-error is non-nil; don't infloop, suggested by Hiroshi Fujishima <pooh@nature.tsukuba.ac.jp>; assume the close parenthesis may be included in the encoded word; encode bogus delimiters. (rfc2047-encode-string): Use mm-with-multibyte-buffer. (rfc2047-encode-max-chars): New variable. (rfc2047-encode-1): New function. (rfc2047-encode): Use it; encode text so that it occupies the maximum width within 76-column; work correctly on Q encoding for iso-2022-* charsets; fold the line before encoding; don't append a space if the encoded word includes close parenthesis. (rfc2047-fold-region): Use existing whitespace for LWSP; make it sure not to break a line just after the header name. (rfc2047-b-encode-region): Remove. (rfc2047-b-encode-string): New function. (rfc2047-q-encode-region): Remove. (rfc2047-q-encode-string): New function. (rfc2047-encode-parameter): New function. (rfc2047-encoded-word-regexp): Don't use shy group. (rfc2047-decode-region): Follow rfc2047-encoded-word-regexp change. (rfc2047-parse-and-decode): Ditto. (rfc2047-decode): Treat the ascii coding-system as raw-text by default. 2005-03-25 Lars Magne Ingebrigtsen <larsi@gnus.org> * lisp/gnus/rfc2047.el (rfc2047-encode-encoded-words): New variable. (rfc2047-field-value): Strip props. (rfc2047-encode-message-header): Disabled header folding -- not all headers can be folded, and this should be done by the message composition mode. Probably. I think. (rfc2047-encodable-p): Say that =? needs encoding. (rfc2047-encode-region): Encode =? strings. 2005-03-25 Jesper Harder <harder@ifa.au.dk> * lisp/gnus/rfc2047.el (rfc2047-encoded-word-regexp): Support RFC 2231 language tags; remove unnecessary '+'. Reported by Stefan Wiens <s.wi@gmx.net>. (rfc2047-decode-string): Don't cons a string unnecessarily. (rfc2047-parse-and-decode, rfc2047-decode): Use a character for the encoding to avoid consing a string. (rfc2047-decode): Use mm-subst-char-in-string instead of mm-replace-chars-in-string. 2005-03-25 TSUCHIYA Masatoshi <tsuchiya@namazu.org> * lisp/gnus/rfc2047.el (rfc2047-encode): Use uppercase letters to specify encodings of MIME-encoded words, in order to improve interoperability with several broken MUAs. 2005-03-21 Reiner Steib <Reiner.Steib@gmx.de> * lisp/gnus/gnus-srvr.el (gnus-browse-select-group): Add NUMBER argument and pass it to `gnus-browse-read-group'. (gnus-browse-read-group): Add NUMBER argument and pass it to `gnus-group-read-ephemeral-group'. * lisp/gnus/gnus-group.el (gnus-group-read-ephemeral-group): Add NUMBER argument and pass it to `gnus-group-read-group'. 2005-03-19 Aidan Kehoe <kehoea@parhasard.net> * lisp/gnus/mm-util.el (mm-xemacs-find-mime-charset): Only call mm-xemacs-find-mime-charset-1 if we have the mule feature available at runtime. 2005-03-25 Katsumi Yamaoka <yamaoka@jpl.org> * man/emacs-mime.texi (Display Customization): Markup fixes. (rfc2047): Update. 2005-03-23 Reiner Steib <Reiner.Steib@gmx.de> * man/gnus-faq.texi: Replaced with auto-generated version.
author Miles Bader <miles@gnu.org>
date Wed, 30 Mar 2005 08:14:32 +0000
parents 23a1cea22d13
children 9f4849fee703
line wrap: on
line source

Two Volume Cross References
===========================

18 March 1992

This enables you to create manuals in *two* volumes, with tables of
contents, cross references, and indices in each volume referring to
*both* volumes.

The procedure is tedious.  However, the resulting two volumes are
conveniently organized.  Each has an index of the whole two volumes.
Each volume starts with page 1.  (I don't like multi-volume works
where each volume starts with a higher page number since I find it
harder to go to the right place in the volume.)

References to the same volume are just the page number; references to
the other volume are a volumne number (in Roman numerals) preceding
the page number.

For example, in Volume I:

    list length ......... 90
    list motion ......II:117

and in Volume II:

    list length ....... I:90
    list motion .........117

All other references and the table of contents work the same way.  I
find this *very* helpful.


In brief: you run tex on a .texi file with

  a. redefined @contents and @summarycontents inputting elisp-toc-2vol.toc file
  b. redone .aux file
  c. redone .fns file


Here are the steps in detail:

% tex vol1.texi
% texindex vol1.??
% tex vol1.texi

% tex vol2.texi
% texindex vol2.??
% tex vol2.texi

### Create .aux files with volume numbers for other volume.

% cp vol1.aux elisp1-aux
% cp vol2.aux elisp2-aux

% cp vol1.aux elisp1-aux-vol-number-added
% cp vol2.aux elisp2-aux-vol-number-added

on elisp1-aux-vol-number-added
(volume-aux-markup 1)           see defun for volum-aux-markup below.
to create             elisp1-aux-vol-number-added

on elisp2-aux-vol-number-added
(volume-aux-markup 2)
to create             elisp2-aux-vol-number-added

insert elisp2-aux-vol-number-added into vol1.aux  (append)
insert elisp1-aux-vol-number-added into vol2.aux  (prepend)

(so you dont have to do it again)
% cp vol1.aux elisp1-aux-2vol-ready
% cp vol2.aux elisp2-aux-2vol-ready


### Create .fn files with volume numbers for other volume.

% cp vol1.fn elisp1-fn
% cp vol2.fn elisp2-fn

% cp vol1.fn elisp1-fn-vol-number-added
% cp vol2.fn elisp2-fn-vol-number-added

on elisp1-fn-vol-number-added
(volume-index-markup "I")
to create             elisp1-fn-vol-number-added

on elisp2-fn-vol-number-added
(volume-index-markup "II")
to create             elisp2-fn-vol-number-added

insert elisp2-fn-vol-number-added into vol1.fn: do following `cat'
insert elisp1-fn-vol-number-added into vol2.fn: do following `cat'

% cat elisp2-fn-vol-number-added >> vol1.fn
% cat elisp1-fn-vol-number-added >> vol2.fn

Be sure to handle special case entries by hand.
Be sure that .fn file has no blank lines.

% texindex vol1.fn
% texindex vol2.fn

(so you dont have to do it again)
% cp vol1.fns elisp1-fns-2vol-ready
% cp vol2.fns elisp2-fns-2vol-ready

### Create merged .toc file with volume number headings.

append vol2.toc to vol1.toc  with following `cat'

% cat vol1.toc vol2.toc > elisp-toc-2vol.toc

and edit in Volume titles

\unnumbchapentry {Volume 1}{}
\unnumbchapentry {}{}

\unnumbchapentry {Index}{295}
\unnumbchapentry {}{}
\unnumbchapentry {Volume 2}{}
\unnumbchapentry {}{}

If you want to put in volume numbers for TOC, then do this:
Create volume specific .toc files with volume numbers in them.

% cp elisp-toc-2vol.toc elisp1-toc.toc
% cp elisp-toc-2vol.toc elisp2-toc.toc

Use keyboard macro to put I: in first half of elisp1-toc.toc and
II: in first half of elisp2-toc.toc

Copy the tocs to something you can remember more easily

% cp elisp2-toc.toc elisp1-toc-ready.toc
% cp elisp1-toc.toc elisp2-toc-ready.toc

Then, edit vol1.texi to input elisp1-toc-ready.toc
and vol2.texi to input elisp2-toc-ready.toc


### Now format the two volumes:

% cp elisp1-aux-2vol-ready vol1.aux
% cp elisp2-aux-2vol-ready vol2.aux

% tex vol1.texi
% tex vol2.texi



For every additional run:

### recopy aux files so the correct ones are read:
% cp elisp1-aux-2vol-ready vol1.aux
% cp elisp2-aux-2vol-ready vol2.aux

Do not run texindex.  Then proper sorted index will stay.
 else do: % cp elisp2-fns-2vol-ready vol2.fns

Do not change the .texi files; they will call the elisp-toc-2vol.toc file.

% tex vol1.texi
% tex vol2.texi

================================================================

@c ================================================================
@tex
% Special @contents  command
% This inputs fixed up table of contents file rather than create new one.
\global\def\contents{%
   \startcontents{Table of Contents}%
      \input elisp-toc-2vol.toc
   \endgroup
   \vfill \eject
}

% Special @summarycontents  command
% This inputs fixed up table of contents file rather than create new one.
\outer\def\summarycontents{%
   \startcontents{Short Contents}%
      %
      \let\chapentry = \shortchapentry
      \let\unnumbchapentry = \shortunnumberedentry
      % We want a true roman here for the page numbers.
      \secfonts
      \let\rm=\shortcontrm \let\bf=\shortcontbf \let\sl=\shortcontsl
      \rm
      \advance\baselineskip by 1pt % Open it up a little.
      \def\secentry ##1##2##3##4{}
      \def\unnumbsecentry ##1##2{}
      \def\subsecentry ##1##2##3##4##5{}
      \def\unnumbsubsecentry ##1##2{}
      \def\subsubsecentry ##1##2##3##4##5##6{}
      \def\unnumbsubsubsecentry ##1##2{}
      \input elisp-toc-2vol.toc
   \endgroup
   \vfill \eject
}
@end tex
@c ================================================================


================================================================


(defun volume-aux-markup (arg)
  "Append `vol. NUMBER' to page number.
Apply to aux file that you save.
Then insert marked file into other volume's .aux file."
  (interactive "sType volume number, 1 or 2: " )
  (goto-char (point-min))
  (while (search-forward "-pg" nil t)
    (end-of-line 1)
    (delete-backward-char 1 nil)
    (insert ", vol.'tie" arg "}")))

(defun volume-index-markup (arg)
  "Prepend  `NUMBER:' to page number.  Use Roman Numeral.
Apply only to unsorted index file,
Then insert marked file into other volume's unsorted index file.
Then run texindex on that file and save."
  (interactive
   "sType volume number,  roman number I or II: " )
  (goto-char (point-min))
  (while (search-forward "\\entry" nil t)
    (search-forward "}{" (save-excursion (end-of-line) (point)) nil)
    (insert arg ":")))


================================================================


The steps:

1. Run TeX, texindex and TeX on file1.
2. Run TeX, texindex and TeX on file2.

3. Copy both .aux files into specially named files

4. In the case of the elisp ref manual,

   copy the *unsorted* function index files into specially named files
   (no other index used in elisp ref manual)


5. For aux files:

   Run a function on the specially named .aux files to label each
   entry according to volume.  Save these files.

   i.e., convert
   'xrdef {Special-pg}{7}  to 'xrdef {Special-pg}{7, vol.'tie1}

5a.Insert each specially named .aux file into the regular .aux file of
   the other volume.

6. For index files:

   Run a function on the specially named unsorted index files to label
   each entry according to volume.  Save these files.

6b.Insert each specially named marked unsorted index file into the
   regular unsorted file of the other volume.  Run texindex on this

7. Insert the other volumes .toc file into the .toc, edit, and rename to
   elisp-toc-2vol.toc

7a. insert special @contents and @summarycontents defs into .texi files.

8. Run TeX on each .texi file.

================



Here is the discursive commentary:

I've been running some small test files, called test1.texi and
test2.texi.  As far as I can see, if we run tex on the two test files,
tex creates a .aux for each that includes the names of all the nodes
in that file.  The node names are used for cross references.

If you insert the .aux file for the second test file, test2.aux, into
the .aux file for the first test file, test1.aux, then when you next
run TeX on the first test file, test1.texi, the second volume cross
references are inserted.

You can edit the text of the cross reference in test2.aux to include
the volume number.

For example, you can take the following two lines from test1.texi and
insert them into test2.texi:

    'xrdef {Special-pg}{7}
    'xrdef {Special-snt}{Section'tie1.6}

You can re-edit this to show that the page is in volume 1:

    'xrdef {Special-pg}{7, vol.'tie1}
    'xrdef {Special-snt}{Section'tie1.6}

(The  'tie  is a TeX special command to keep the number tied on one
line to the previous word.  I don't know if it works after a period in
the "vol." but figure it is worth trying.  {The  ' is the @  of  .aux files.}
Apparently 'tie is like the tilde in plain tex; in texinfo.tex, the
definition for 'tie is the following:

    \def\tie{\penalty 10000\ }     % Save plain tex definition of ~.

)

After running tex on the test2.texi file with the augmented test2.aux
file, you can see the following in the resulting DVI file:

    See Section 1.6 [Special], page 7, vol. 1

Note that TeX rewrites the .aux file each time TeX is run, so after
running Tex using an .aux file augmented with the .aux file from the
other volume, the new .aux file will *lack* the other volumes cross
references.  Save your augmented .aux file in some other name for
another run!