Re-generated.
[bpt/emacs.git] / lisp / format.el
CommitLineData
a8581027 1;;; format.el --- read and save files in multiple formats
b578f267 2
c90f2757 3;; Copyright (C) 1994, 1995, 1997, 1999, 2001, 2002, 2003, 2004,
409cc4a3 4;; 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
029894b9 5
823139fb 6;; Author: Boris Goldowsky <boris@gnu.org>
029894b9
BG
7
8;; This file is part of GNU Emacs.
9
eb3fa2cf 10;; GNU Emacs is free software: you can redistribute it and/or modify
029894b9 11;; it under the terms of the GNU General Public License as published by
eb3fa2cf
GM
12;; the Free Software Foundation, either version 3 of the License, or
13;; (at your option) any later version.
b578f267 14
029894b9
BG
15;; GNU Emacs is distributed in the hope that it will be useful,
16;; but WITHOUT ANY WARRANTY; without even the implied warranty of
17;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18;; GNU General Public License for more details.
b578f267 19
029894b9 20;; You should have received a copy of the GNU General Public License
eb3fa2cf 21;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
029894b9
BG
22
23;;; Commentary:
b578f267
EN
24
25;; This file defines a unified mechanism for saving & loading files stored
26;; in different formats. `format-alist' contains information that directs
029894b9 27;; Emacs to call an encoding or decoding function when reading or writing
823139fb 28;; files that match certain conditions.
029894b9 29;;
b578f267
EN
30;; When a file is visited, its format is determined by matching the
31;; beginning of the file against regular expressions stored in
32;; `format-alist'. If this fails, you can manually translate the buffer
33;; using `format-decode-buffer'. In either case, the formats used are
34;; listed in the variable `buffer-file-format', and become the default
35;; format for saving the buffer. To save a buffer in a different format,
36;; change this variable, or use `format-write-file'.
029894b9
BG
37;;
38;; Auto-save files are normally created in the same format as the visited
4a2255ad 39;; file, but the variable `buffer-auto-save-file-format' can be set to a
b578f267
EN
40;; particularly fast or otherwise preferred format to be used for
41;; auto-saving (or nil to do no encoding on auto-save files, but then you
42;; risk losing any text-properties in the buffer).
029894b9 43;;
b578f267
EN
44;; You can manually translate a buffer into or out of a particular format
45;; with the functions `format-encode-buffer' and `format-decode-buffer'.
46;; To translate just the region use the functions `format-encode-region'
823139fb 47;; and `format-decode-region'.
029894b9 48;;
b578f267
EN
49;; You can define a new format by writing the encoding and decoding
50;; functions, and adding an entry to `format-alist'. See enriched.el for
51;; an example of how to implement a file format. There are various
52;; functions defined in this file that may be useful for writing the
53;; encoding and decoding functions:
54;; * `format-annotate-region' and `format-deannotate-region' allow a
55;; single alist of information to be used for encoding and decoding.
56;; The alist defines a correspondence between strings in the file
57;; ("annotations") and text-properties in the buffer.
029894b9
BG
58;; * `format-replace-strings' is similarly useful for doing simple
59;; string->string translations in a reversible manner.
60
b578f267
EN
61;;; Code:
62
029894b9 63(put 'buffer-file-format 'permanent-local t)
4a2255ad 64(put 'buffer-auto-save-file-format 'permanent-local t)
029894b9 65
823139fb 66(defvar format-alist
8d0dd9d2 67 '((text/enriched "Extended MIME text/enriched format."
029894b9
BG
68 "Content-[Tt]ype:[ \t]*text/enriched"
69 enriched-decode enriched-encode t enriched-mode)
c84cf2bf 70 (plain "ISO 8859-1 standard format, no text properties."
029894b9
BG
71 ;; Plain only exists so that there is an obvious neutral choice in
72 ;; the completion list.
c84cf2bf 73 nil nil nil nil nil)
823139fb
DL
74 (TeX "TeX (encoding)"
75 nil
c84cf2bf 76 iso-tex2iso iso-iso2tex t nil)
823139fb
DL
77 (gtex "German TeX (encoding)"
78 nil
c84cf2bf 79 iso-gtex2iso iso-iso2gtex t nil)
f1d6fe69 80 (html "HTML/SGML \"ISO 8879:1986//ENTITIES Added Latin 1//EN\" (encoding)"
823139fb 81 nil
f1d6fe69 82 iso-sgml2iso iso-iso2sgml t nil)
823139fb
DL
83 (rot13 "rot13"
84 nil
c84cf2bf 85 "tr a-mn-z n-za-m" "tr a-mn-z n-za-m" t nil)
823139fb
DL
86 (duden "Duden Ersatzdarstellung"
87 nil
88 "diac" iso-iso2duden t nil)
89 (de646 "German ASCII (ISO 646)"
90 nil
271af233 91 "recode -f iso646-ge:latin1" "recode -f latin1:iso646-ge" t nil)
823139fb
DL
92 (denet "net German"
93 nil
c84cf2bf 94 iso-german iso-cvt-read-only t nil)
823139fb
DL
95 (esnet "net Spanish"
96 nil
8d0dd9d2 97 iso-spanish iso-cvt-read-only t nil))
029894b9 98 "List of information about understood file formats.
ce44e932
GM
99Elements are of the form
100\(NAME DOC-STR REGEXP FROM-FN TO-FN MODIFY MODE-FN PRESERVE).
d1ae737b 101
029894b9 102NAME is a symbol, which is stored in `buffer-file-format'.
d1ae737b 103
029894b9
BG
104DOC-STR should be a single line providing more information about the
105 format. It is currently unused, but in the future will be shown to
106 the user if they ask for more information.
d1ae737b 107
029894b9 108REGEXP is a regular expression to match against the beginning of the file;
823139fb 109 it should match only files in that format. Use nil to avoid
12008438 110 matching at all for formats for which it isn't appropriate to
823139fb 111 require explicit encoding/decoding.
d1ae737b 112
12008438 113FROM-FN is called to decode files in that format; it takes two args, BEGIN
029894b9
BG
114 and END, and can make any modifications it likes, returning the new
115 end. It must make sure that the beginning of the file no longer
116 matches REGEXP, or else it will get called again.
d1ae737b
RS
117 Alternatively, FROM-FN can be a string, which specifies a shell command
118 (including options) to be used as a filter to perform the conversion.
119
12008438 120TO-FN is called to encode a region into that format; it takes three
d1abb42f
RS
121 arguments: BEGIN, END, and BUFFER. BUFFER is the original buffer that
122 the data being written came from, which the function could use, for
123 example, to find the values of local variables. TO-FN should either
124 return a list of annotations like `write-region-annotate-functions',
125 or modify the region and return the new end.
d1ae737b
RS
126 Alternatively, TO-FN can be a string, which specifies a shell command
127 (including options) to be used as a filter to perform the conversion.
128
d1abb42f
RS
129MODIFY, if non-nil, means the TO-FN wants to modify the region. If nil,
130 TO-FN will not make any changes but will instead return a list of
823139fb 131 annotations.
d1ae737b 132
3e224645
EZ
133MODE-FN, if specified, is called when visiting a file with that format.
134 It is called with a single positive argument, on the assumption
12008438 135 that this would turn on some minor mode.
34f30f79
RS
136
137PRESERVE, if non-nil, means that `format-write-file' should not remove
ec8dda2f 138 this format from `buffer-file-format'.")
029894b9
BG
139
140;;; Basic Functions (called from Lisp)
141
c84cf2bf 142(defun format-encode-run-method (method from to &optional buffer)
12008438
TTN
143 "Translate using METHOD the text from FROM to TO.
144If METHOD is a string, it is a shell command (including options);
c84cf2bf
RS
145otherwise, it should be a Lisp function.
146BUFFER should be the buffer that the output originally came from."
147 (if (stringp method)
823139fb 148 (let ((error-buff (get-buffer-create "*Format Errors*"))
7cc45a35 149 (coding-system-for-read 'no-conversion)
823139fb
DL
150 format-alist)
151 (with-current-buffer error-buff
152 (widen)
153 (erase-buffer))
10fa0073
GM
154 (if (and (zerop (save-window-excursion
155 (shell-command-on-region from to method t t
156 error-buff)))
823139fb
DL
157 ;; gzip gives zero exit status with bad args, for instance.
158 (zerop (with-current-buffer error-buff
159 (buffer-size))))
160 (bury-buffer error-buff)
161 (switch-to-buffer-other-window error-buff)
10fa0073 162 (error "Format encoding failed")))
c84cf2bf
RS
163 (funcall method from to buffer)))
164
165(defun format-decode-run-method (method from to &optional buffer)
12008438
TTN
166 "Decode using METHOD the text from FROM to TO.
167If METHOD is a string, it is a shell command (including options); otherwise,
168it should be a Lisp function. Decoding is done for the given BUFFER."
c84cf2bf 169 (if (stringp method)
823139fb 170 (let ((error-buff (get-buffer-create "*Format Errors*"))
7cc45a35 171 (coding-system-for-write 'no-conversion)
823139fb
DL
172 format-alist)
173 (with-current-buffer error-buff
174 (widen)
175 (erase-buffer))
176 ;; We should perhaps go via a temporary buffer and copy it
177 ;; back, in case of errors.
178 (if (and (zerop (save-window-excursion
179 (shell-command-on-region (point-min) (point-max)
180 method t t
181 error-buff)))
182 ;; gzip gives zero exit status with bad args, for instance.
183 (zerop (with-current-buffer error-buff
184 (buffer-size))))
185 (bury-buffer error-buff)
186 (switch-to-buffer-other-window error-buff)
187 (error "Format decoding failed"))
c84cf2bf
RS
188 (point))
189 (funcall method from to)))
190
10fa0073 191(defun format-annotate-function (format from to orig-buf format-count)
823139fb 192 "Return annotations for writing region as FORMAT.
12008438
TTN
193FORMAT is a symbol naming one of the formats defined in `format-alist'.
194It must be a single symbol, not a list like `buffer-file-format'.
d1abb42f
RS
195FROM and TO delimit the region to be operated on in the current buffer.
196ORIG-BUF is the original buffer that the data came from.
10fa0073
GM
197
198FORMAT-COUNT is an integer specifying how many times this function has
199been called in the process of decoding ORIG-BUF.
200
12008438 201This function works like a function in `write-region-annotate-functions':
029894b9 202it either returns a list of annotations, or returns with a different buffer
10fa0073
GM
203current, which contains the modified text to write. In the latter case,
204this function's value is nil.
029894b9
BG
205
206For most purposes, consider using `format-encode-region' instead."
10fa0073
GM
207 ;; This function is called by write-region (actually
208 ;; build_annotations) for each element of buffer-file-format.
029894b9
BG
209 (let* ((info (assq format format-alist))
210 (to-fn (nth 4 info))
211 (modify (nth 5 info)))
212 (if to-fn
213 (if modify
214 ;; To-function wants to modify region. Copy to safe place.
10fa0073 215 (let ((copy-buf (get-buffer-create (format " *Format Temp %d*"
8a2f1176
RS
216 format-count)))
217 (sel-disp selective-display)
5258c763
JB
218 (multibyte enable-multibyte-characters)
219 (coding-system buffer-file-coding-system))
8a2f1176
RS
220 (with-current-buffer copy-buf
221 (setq selective-display sel-disp)
5258c763
JB
222 (set-buffer-multibyte multibyte)
223 (setq buffer-file-coding-system coding-system))
029894b9
BG
224 (copy-to-buffer copy-buf from to)
225 (set-buffer copy-buf)
226 (format-insert-annotations write-region-annotations-so-far from)
c84cf2bf 227 (format-encode-run-method to-fn (point-min) (point-max) orig-buf)
029894b9
BG
228 nil)
229 ;; Otherwise just call function, it will return annotations.
d1abb42f 230 (funcall to-fn from to orig-buf)))))
029894b9
BG
231
232(defun format-decode (format length &optional visit-flag)
233 ;; This function is called by insert-file-contents whenever a file is read.
234 "Decode text from any known FORMAT.
823139fb 235FORMAT is a symbol appearing in `format-alist' or a list of such symbols,
029894b9
BG
236or nil, in which case this function tries to guess the format of the data by
237matching against the regular expressions in `format-alist'. After a match is
238found and the region decoded, the alist is searched again from the beginning
239for another match.
240
241Second arg LENGTH is the number of characters following point to operate on.
242If optional third arg VISIT-FLAG is true, set `buffer-file-format'
be227f22
SM
243to the reverted list of formats used, and call any mode functions defined
244for those formats.
029894b9 245
12008438 246Return the new length of the decoded region.
029894b9
BG
247
248For most purposes, consider using `format-decode-region' instead."
249 (let ((mod (buffer-modified-p))
10fa0073 250 (begin (point))
029894b9 251 (end (+ (point) length)))
10fa0073
GM
252 (unwind-protect
253 (progn
254 ;; Don't record undo information for the decoding.
71296446 255
10fa0073
GM
256 (if (null format)
257 ;; Figure out which format it is in, remember list in `format'.
258 (let ((try format-alist))
259 (while try
260 (let* ((f (car try))
261 (regexp (nth 2 f))
262 (p (point)))
263 (if (and regexp (looking-at regexp)
264 (< (match-end 0) (+ begin length)))
265 (progn
be227f22 266 (push (car f) format)
10fa0073
GM
267 ;; Decode it
268 (if (nth 3 f)
269 (setq end (format-decode-run-method (nth 3 f) begin end)))
270 ;; Call visit function if required
271 (if (and visit-flag (nth 6 f)) (funcall (nth 6 f) 1))
272 ;; Safeguard against either of the functions changing pt.
273 (goto-char p)
274 ;; Rewind list to look for another format
275 (setq try format-alist))
276 (setq try (cdr try))))))
277 ;; Deal with given format(s)
278 (or (listp format) (setq format (list format)))
279 (let ((do format) f)
280 (while do
281 (or (setq f (assq (car do) format-alist))
6b61353c 282 (error "Unknown format %s" (car do)))
10fa0073
GM
283 ;; Decode:
284 (if (nth 3 f)
285 (setq end (format-decode-run-method (nth 3 f) begin end)))
286 ;; Call visit function if required
287 (if (and visit-flag (nth 6 f)) (funcall (nth 6 f) 1))
be227f22
SM
288 (setq do (cdr do))))
289 ;; Encode in the opposite order.
290 (setq format (reverse format)))
10fa0073
GM
291 (if visit-flag
292 (setq buffer-file-format format)))
71296446 293
021cfbea
GM
294 (set-buffer-modified-p mod))
295
10fa0073 296 ;; Return new length of region
029894b9
BG
297 (- end begin)))
298
299;;;
300;;; Interactive functions & entry points
301;;;
302
303(defun format-decode-buffer (&optional format)
304 "Translate the buffer from some FORMAT.
12008438
TTN
305If the format is not specified, attempt a regexp-based guess.
306Set `buffer-file-format' to the format used, and call any
307format-specific mode functions."
029894b9 308 (interactive
5b76833f 309 (list (format-read "Translate buffer from format (default guess): ")))
029894b9
BG
310 (save-excursion
311 (goto-char (point-min))
312 (format-decode format (buffer-size) t)))
313
314(defun format-decode-region (from to &optional format)
315 "Decode the region from some format.
316Arg FORMAT is optional; if omitted the format will be determined by looking
317for identifying regular expressions at the beginning of the region."
318 (interactive
823139fb 319 (list (region-beginning) (region-end)
5b76833f 320 (format-read "Translate region from format (default guess): ")))
029894b9
BG
321 (save-excursion
322 (goto-char from)
323 (format-decode format (- to from) nil)))
324
325(defun format-encode-buffer (&optional format)
326 "Translate the buffer into FORMAT.
327FORMAT defaults to `buffer-file-format'. It is a symbol naming one of the
328formats defined in `format-alist', or a list of such symbols."
329 (interactive
330 (list (format-read (format "Translate buffer to format (default %s): "
331 buffer-file-format))))
332 (format-encode-region (point-min) (point-max) format))
333
334(defun format-encode-region (beg end &optional format)
823139fb 335 "Translate the region into some FORMAT.
12008438 336FORMAT defaults to `buffer-file-format'. It is a symbol naming
029894b9 337one of the formats defined in `format-alist', or a list of such symbols."
823139fb
DL
338 (interactive
339 (list (region-beginning) (region-end)
340 (format-read (format "Translate region to format (default %s): "
341 buffer-file-format))))
342 (if (null format) (setq format buffer-file-format))
343 (if (symbolp format) (setq format (list format)))
344 (save-excursion
345 (goto-char end)
346 (let ((cur-buf (current-buffer))
347 (end (point-marker)))
348 (while format
349 (let* ((info (assq (car format) format-alist))
350 (to-fn (nth 4 info))
351 (modify (nth 5 info))
352 result)
353 (if to-fn
354 (if modify
355 (setq end (format-encode-run-method to-fn beg end
356 (current-buffer)))
357 (format-insert-annotations
358 (funcall to-fn beg end (current-buffer)))))
359 (setq format (cdr format)))))))
029894b9 360
6b61353c 361(defun format-write-file (filename format &optional confirm)
823139fb 362 "Write current buffer into file FILENAME using some FORMAT.
6b61353c 363Make buffer visit that file and set the format as the default for future
08d55d13
GM
364saves (see below). If the buffer is already visiting a file, you can
365specify a directory name as FILENAME, to write a file of the same old name
366in that directory.
6b61353c 367
12008438
TTN
368If optional third arg CONFIRM is non-nil, ask for confirmation before
369overwriting an existing file. Interactively, confirmation is required
ce44e932
GM
370unless you supply a prefix argument.
371
08d55d13
GM
372This function sets the buffer-local value of `buffer-file-format'
373to FORMAT, then appends any elements from the previous value with
374a non-nil `preserve' flag (see the documentation of `format-alist'),
375if they are not already present in the new value."
029894b9
BG
376 (interactive
377 ;; Same interactive spec as write-file, plus format question.
378 (let* ((file (if buffer-file-name
379 (read-file-name "Write file: "
380 nil nil nil nil)
381 (read-file-name "Write file: "
382 (cdr (assq 'default-directory
383 (buffer-local-variables)))
384 nil nil (buffer-name))))
823139fb 385 (fmt (format-read (format "Write file `%s' in format: "
029894b9 386 (file-name-nondirectory file)))))
6b61353c 387 (list file fmt (not current-prefix-arg))))
34f30f79
RS
388 (let ((old-formats buffer-file-format)
389 preserve-formats)
390 (dolist (fmt old-formats)
391 (let ((aelt (assq fmt format-alist)))
392 (if (nth 7 aelt)
393 (push fmt preserve-formats))))
394 (setq buffer-file-format format)
395 (dolist (fmt preserve-formats)
396 (unless (memq fmt buffer-file-format)
397 (setq buffer-file-format (append buffer-file-format (list fmt))))))
6b61353c 398 (write-file filename confirm))
029894b9 399
ce6af52b 400(defun format-find-file (filename format)
823139fb 401 "Find the file FILENAME using data format FORMAT.
ce6af52b
KH
402If FORMAT is nil then do not do any format conversion."
403 (interactive
404 ;; Same interactive spec as write-file, plus format question.
405 (let* ((file (read-file-name "Find file: "))
823139fb 406 (fmt (format-read (format "Read file `%s' in format: "
ce6af52b
KH
407 (file-name-nondirectory file)))))
408 (list file fmt)))
409 (let ((format-alist nil))
410 (find-file filename))
411 (if format
412 (format-decode-buffer format)))
413
414(defun format-insert-file (filename format &optional beg end)
823139fb 415 "Insert the contents of file FILENAME using data format FORMAT.
ce6af52b
KH
416If FORMAT is nil then do not do any format conversion.
417The optional third and fourth arguments BEG and END specify
12008438 418the part (in bytes) of the file to read.
ce6af52b
KH
419
420The return value is like the value of `insert-file-contents':
6b61353c 421a list (ABSOLUTE-FILE-NAME SIZE)."
ce6af52b
KH
422 (interactive
423 ;; Same interactive spec as write-file, plus format question.
424 (let* ((file (read-file-name "Find file: "))
823139fb 425 (fmt (format-read (format "Read file `%s' in format: "
ce6af52b
KH
426 (file-name-nondirectory file)))))
427 (list file fmt)))
f70aa678
MR
428 (let (value size old-undo)
429 ;; Record only one undo entry for the insertion. Inhibit point-motion and
430 ;; modification hooks as with `insert-file-contents'.
431 (let ((inhibit-point-motion-hooks t)
432 (inhibit-modification-hooks t))
433 ;; Don't bind `buffer-undo-list' to t here to assert that
434 ;; `insert-file-contents' may record whether the buffer was unmodified
435 ;; before.
436 (let ((format-alist nil))
437 (setq value (insert-file-contents filename nil beg end))
438 (setq size (nth 1 value)))
439 (when (consp buffer-undo-list)
440 (let ((head (car buffer-undo-list)))
441 (when (and (consp head)
442 (equal (car head) (point))
443 (equal (cdr head) (+ (point) size)))
444 ;; Remove first entry from `buffer-undo-list', we shall insert
445 ;; another one below.
446 (setq old-undo (cdr buffer-undo-list)))))
447 (when format
448 (let ((buffer-undo-list t))
449 (setq size (format-decode format size)
450 value (list (car value) size)))
451 (unless (eq buffer-undo-list t)
452 (setq buffer-undo-list
453 (cons (cons (point) (+ (point) size)) old-undo)))))
454 (unless inhibit-modification-hooks
455 (run-hook-with-args 'after-change-functions (point) (+ (point) size) 0))
ce6af52b
KH
456 value))
457
029894b9
BG
458(defun format-read (&optional prompt)
459 "Read and return the name of a format.
460Return value is a list, like `buffer-file-format'; it may be nil.
461Formats are defined in `format-alist'. Optional arg is the PROMPT to use."
462 (let* ((table (mapcar (lambda (x) (list (symbol-name (car x))))
463 format-alist))
464 (ans (completing-read (or prompt "Format: ") table nil t)))
465 (if (not (equal "" ans)) (list (intern ans)))))
466
467
468;;;
469;;; Below are some functions that may be useful in writing encoding and
470;;; decoding functions for use in format-alist.
471;;;
472
473(defun format-replace-strings (alist &optional reverse beg end)
474 "Do multiple replacements on the buffer.
d0fc7e3b 475ALIST is a list of (FROM . TO) pairs, which should be proper arguments to
12008438
TTN
476`search-forward' and `replace-match', respectively.
477Optional second arg REVERSE, if non-nil, means the pairs are (TO . FROM),
478so that you can use the same list in both directions if it contains only
479literal strings.
823139fb 480Optional args BEG and END specify a region of the buffer on which to operate."
029894b9
BG
481 (save-excursion
482 (save-restriction
483 (or beg (setq beg (point-min)))
484 (if end (narrow-to-region (point-min) end))
485 (while alist
486 (let ((from (if reverse (cdr (car alist)) (car (car alist))))
2c3ac81d 487 (to (if reverse (car (car alist)) (cdr (car alist)))))
029894b9
BG
488 (goto-char beg)
489 (while (search-forward from nil t)
490 (goto-char (match-beginning 0))
491 (insert to)
492 (set-text-properties (- (point) (length to)) (point)
493 (text-properties-at (point)))
494 (delete-region (point) (+ (point) (- (match-end 0)
495 (match-beginning 0)))))
496 (setq alist (cdr alist)))))))
497
498;;; Some list-manipulation functions that we need.
499
500(defun format-delq-cons (cons list)
823139fb
DL
501 "Remove the given CONS from LIST by side effect and return the new LIST.
502Since CONS could be the first element of LIST, write
503`\(setq foo \(format-delq-cons element foo))' to be sure of changing
504the value of `foo'."
029894b9
BG
505 (if (eq cons list)
506 (cdr list)
507 (let ((p list))
508 (while (not (eq (cdr p) cons))
e8af40ee 509 (if (null p) (error "format-delq-cons: not an element"))
029894b9
BG
510 (setq p (cdr p)))
511 ;; Now (cdr p) is the cons to delete
512 (setcdr p (cdr cons))
513 list)))
71296446 514
029894b9
BG
515(defun format-make-relatively-unique (a b)
516 "Delete common elements of lists A and B, return as pair.
12008438 517Compare using `equal'."
029894b9
BG
518 (let* ((acopy (copy-sequence a))
519 (bcopy (copy-sequence b))
520 (tail acopy))
521 (while tail
522 (let ((dup (member (car tail) bcopy))
523 (next (cdr tail)))
524 (if dup (setq acopy (format-delq-cons tail acopy)
525 bcopy (format-delq-cons dup bcopy)))
526 (setq tail next)))
527 (cons acopy bcopy)))
528
529(defun format-common-tail (a b)
530 "Given two lists that have a common tail, return it.
12008438 531Compare with `equal', and return the part of A that is equal to the
029894b9 532equivalent part of B. If even the last items of the two are not equal,
12008438 533return nil."
029894b9
BG
534 (let ((la (length a))
535 (lb (length b)))
536 ;; Make sure they are the same length
823139fb 537 (if (> la lb)
029894b9
BG
538 (setq a (nthcdr (- la lb) a))
539 (setq b (nthcdr (- lb la) b))))
540 (while (not (equal a b))
541 (setq a (cdr a)
542 b (cdr b)))
543 a)
544
f0a6c717
GM
545(defun format-proper-list-p (list)
546 "Return t if LIST is a proper list.
547A proper list is a list ending with a nil cdr, not with an atom "
548 (when (listp list)
549 (while (consp list)
550 (setq list (cdr list)))
551 (null list)))
552
029894b9 553(defun format-reorder (items order)
12008438
TTN
554 "Arrange ITEMS to follow partial ORDER.
555Elements of ITEMS equal to elements of ORDER will be rearranged
556to follow the ORDER. Unmatched items will go last."
029894b9
BG
557 (if order
558 (let ((item (member (car order) items)))
559 (if item
823139fb 560 (cons (car item)
029894b9
BG
561 (format-reorder (format-delq-cons item items)
562 (cdr order)))
563 (format-reorder items (cdr order))))
564 items))
565
566(put 'face 'format-list-valued t) ; These text-properties take values
567(put 'unknown 'format-list-valued t) ; that are lists, the elements of which
568 ; should be considered separately.
569 ; See format-deannotate-region and
570 ; format-annotate-region.
571
f3bbef87
GM
572;; This text property has list values, but they are treated atomically.
573
574(put 'display 'format-list-atomic-p t)
575
029894b9
BG
576;;;
577;;; Decoding
578;;;
579
580(defun format-deannotate-region (from to translations next-fn)
581 "Translate annotations in the region into text properties.
823139fb 582This sets text properties between FROM to TO as directed by the
029894b9
BG
583TRANSLATIONS and NEXT-FN arguments.
584
585NEXT-FN is a function that searches forward from point for an annotation.
586It should return a list of 4 elements: \(BEGIN END NAME POSITIVE). BEGIN and
587END are buffer positions bounding the annotation, NAME is the name searched
588for in TRANSLATIONS, and POSITIVE should be non-nil if this annotation marks
589the beginning of a region with some property, or nil if it ends the region.
590NEXT-FN should return nil if there are no annotations after point.
591
592The basic format of the TRANSLATIONS argument is described in the
593documentation for the `format-annotate-region' function. There are some
594additional things to keep in mind for decoding, though:
595
596When an annotation is found, the TRANSLATIONS list is searched for a
597text-property name and value that corresponds to that annotation. If the
598text-property has several annotations associated with it, it will be used only
599if the other annotations are also in effect at that point. The first match
600found whose annotations are all present is used.
601
602The text property thus determined is set to the value over the region between
603the opening and closing annotations. However, if the text-property name has a
604non-nil `format-list-valued' property, then the value will be consed onto the
605surrounding value of the property, rather than replacing that value.
606
607There are some special symbols that can be used in the \"property\" slot of
608the TRANSLATIONS list: PARAMETER and FUNCTION \(spelled in uppercase).
609Annotations listed under the pseudo-property PARAMETER are considered to be
610arguments of the immediately surrounding annotation; the text between the
611opening and closing parameter annotations is deleted from the buffer but saved
92308d3a
RS
612as a string.
613
614The surrounding annotation should be listed under the pseudo-property
615FUNCTION. Instead of inserting a text-property for this annotation,
616the function listed in the VALUE slot is called to make whatever
617changes are appropriate. It can also return a list of the form
618\(START LOC PROP VALUE) which specifies a property to put on. The
619function's first two arguments are the START and END locations, and
620the rest of the arguments are any PARAMETERs found in that region.
029894b9
BG
621
622Any annotations that are found by NEXT-FN but not defined by TRANSLATIONS
623are saved as values of the `unknown' text-property \(which is list-valued).
624The TRANSLATIONS list should usually contain an entry of the form
625 \(unknown \(nil format-annotate-value))
626to write these unknown annotations back into the file."
627 (save-excursion
628 (save-restriction
629 (narrow-to-region (point-min) to)
630 (goto-char from)
631 (let (next open-ans todo loc unknown-ans)
632 (while (setq next (funcall next-fn))
633 (let* ((loc (nth 0 next))
634 (end (nth 1 next))
635 (name (nth 2 next))
636 (positive (nth 3 next))
637 (found nil))
638
639 ;; Delete the annotation
640 (delete-region loc end)
22828206
RS
641 (cond
642 ;; Positive annotations are stacked, remembering location
be227f22 643 (positive (push `(,name ((,loc . nil))) open-ans))
22828206
RS
644 ;; It is a negative annotation:
645 ;; Close the top annotation & add its text property.
646 ;; If the file's nesting is messed up, the close might not match
647 ;; the top thing on the open-annotations stack.
648 ;; If no matching annotation is open, just ignore the close.
649 ((not (assoc name open-ans))
650 (message "Extra closing annotation (%s) in file" name))
651 ;; If one is open, but not on the top of the stack, close
652 ;; the things in between as well. Set `found' when the real
653 ;; one is closed.
654 (t
655 (while (not found)
656 (let* ((top (car open-ans)) ; first on stack: should match.
657 (top-name (car top)) ; text property name
658 (top-extents (nth 1 top)) ; property regions
659 (params (cdr (cdr top))) ; parameters
660 (aalist translations)
661 (matched nil))
662 (if (equal name top-name)
663 (setq found t)
664 (message "Improper nesting in file."))
665 ;; Look through property names in TRANSLATIONS
666 (while aalist
667 (let ((prop (car (car aalist)))
668 (alist (cdr (car aalist))))
669 ;; And look through values for each property
670 (while alist
671 (let ((value (car (car alist)))
672 (ans (cdr (car alist))))
673 (if (member top-name ans)
674 ;; This annotation is listed, but still have to
675 ;; check if multiple annotations are satisfied
676 (if (member nil (mapcar (lambda (r)
677 (assoc r open-ans))
678 ans))
679 nil ; multiple ans not satisfied
680 ;; If there are multiple annotations going
681 ;; into one text property, split up the other
682 ;; annotations so they apply individually to
683 ;; the other regions.
684 (setcdr (car top-extents) loc)
685 (let ((to-split ans) this-one extents)
686 (while to-split
687 (setq this-one
688 (assoc (car to-split) open-ans)
689 extents (nth 1 this-one))
690 (if (not (eq this-one top))
691 (setcar (cdr this-one)
692 (format-subtract-regions
693 extents top-extents)))
694 (setq to-split (cdr to-split))))
695 ;; Set loop variables to nil so loop
696 ;; will exit.
697 (setq alist nil aalist nil matched t
698 ;; pop annotation off stack.
699 open-ans (cdr open-ans))
700 (let ((extents top-extents)
701 (start (car (car top-extents)))
702 (loc (cdr (car top-extents))))
703 (while extents
704 (cond
705 ;; Check for pseudo-properties
706 ((eq prop 'PARAMETER)
707 ;; A parameter of the top open ann:
708 ;; delete text and use as arg.
709 (if open-ans
710 ;; (If nothing open, discard).
711 (setq open-ans
712 (cons
713 (append (car open-ans)
714 (list
715 (buffer-substring
716 start loc)))
717 (cdr open-ans))))
718 (delete-region start loc))
719 ((eq prop 'FUNCTION)
720 ;; Not a property, but a function.
721 (let ((rtn
722 (apply value start loc params)))
be227f22 723 (if rtn (push rtn todo))))
22828206
RS
724 (t
725 ;; Normal property/value pair
726 (setq todo
727 (cons (list start loc prop value)
728 todo))))
729 (setq extents (cdr extents)
730 start (car (car extents))
731 loc (cdr (car extents))))))))
732 (setq alist (cdr alist))))
733 (setq aalist (cdr aalist)))
734 (if (not matched)
029894b9
BG
735 ;; Didn't find any match for the annotation:
736 ;; Store as value of text-property `unknown'.
22828206
RS
737 (let ((extents top-extents)
738 (start (car (car top-extents)))
0b140219 739 (loc (or (cdr (car top-extents)) loc)))
22828206
RS
740 (while extents
741 (setq open-ans (cdr open-ans)
742 todo (cons (list start loc 'unknown top-name)
743 todo)
744 unknown-ans (cons name unknown-ans)
745 extents (cdr extents)
746 start (car (car extents))
747 loc (cdr (car extents))))))))))))
029894b9
BG
748
749 ;; Once entire file has been scanned, add the properties.
750 (while todo
751 (let* ((item (car todo))
752 (from (nth 0 item))
753 (to (nth 1 item))
754 (prop (nth 2 item))
755 (val (nth 3 item)))
22828206
RS
756
757 (if (numberp val) ; add to ambient value if numeric
758 (format-property-increment-region from to prop val 0)
759 (put-text-property
029894b9 760 from to prop
22828206 761 (cond ((get prop 'format-list-valued) ; value gets consed onto
029894b9
BG
762 ; list-valued properties
763 (let ((prev (get-text-property from prop)))
764 (cons val (if (listp prev) prev (list prev)))))
22828206 765 (t val))))) ; normally, just set to val.
029894b9 766 (setq todo (cdr todo)))
22828206 767
029894b9
BG
768 (if unknown-ans
769 (message "Unknown annotations: %s" unknown-ans))))))
770
22828206 771(defun format-subtract-regions (minu subtra)
9c3054da 772 "Remove from the regions in MINUEND the regions in SUBTRAHEND.
d0fc7e3b 773A region is a dotted pair (FROM . TO). Both parameters are lists of
823139fb
DL
774regions. Each list must contain nonoverlapping, noncontiguous
775regions, in descending order. The result is also nonoverlapping,
776noncontiguous, and in descending order. The first element of MINUEND
777can have a cdr of nil, indicating that the end of that region is not
9c3054da
JB
778yet known.
779
780\(fn MINUEND SUBTRAHEND)"
22828206
RS
781 (let* ((minuend (copy-alist minu))
782 (subtrahend (copy-alist subtra))
783 (m (car minuend))
784 (s (car subtrahend))
785 results)
786 (while (and minuend subtrahend)
823139fb 787 (cond
22828206
RS
788 ;; The minuend starts after the subtrahend ends; keep it.
789 ((> (car m) (cdr s))
be227f22
SM
790 (push m results)
791 (setq minuend (cdr minuend)
22828206
RS
792 m (car minuend)))
793 ;; The minuend extends beyond the end of the subtrahend. Chop it off.
794 ((or (null (cdr m)) (> (cdr m) (cdr s)))
be227f22 795 (push (cons (1+ (cdr s)) (cdr m)) results)
22828206
RS
796 (setcdr m (cdr s)))
797 ;; The subtrahend starts after the minuend ends; throw it away.
798 ((< (cdr m) (car s))
799 (setq subtrahend (cdr subtrahend) s (car subtrahend)))
800 ;; The subtrahend extends beyond the end of the minuend. Chop it off.
801 (t ;(<= (cdr m) (cdr s)))
802 (if (>= (car m) (car s))
803 (setq minuend (cdr minuend) m (car minuend))
804 (setcdr m (1- (car s)))
805 (setq subtrahend (cdr subtrahend) s (car subtrahend))))))
806 (nconc (nreverse results) minuend)))
807
808;; This should probably go somewhere other than format.el. Then again,
809;; indent.el has alter-text-property. NOTE: We can also use
810;; next-single-property-change instead of text-property-not-all, but then
811;; we have to see if we passed TO.
812(defun format-property-increment-region (from to prop delta default)
12008438 813 "In the region from FROM to TO increment property PROP by amount DELTA.
823139fb 814DELTA may be negative. If property PROP is nil anywhere
22828206
RS
815in the region, it is treated as though it were DEFAULT."
816 (let ((cur from) val newval next)
817 (while cur
818 (setq val (get-text-property cur prop)
819 newval (+ (or val default) delta)
820 next (text-property-not-all cur to prop val))
821 (put-text-property cur (or next to) prop newval)
822 (setq cur next))))
823
029894b9
BG
824;;;
825;;; Encoding
826;;;
827
828(defun format-insert-annotations (list &optional offset)
829 "Apply list of annotations to buffer as `write-region' would.
12008438 830Insert each element of the given LIST of buffer annotations at its
029894b9
BG
831appropriate place. Use second arg OFFSET if the annotations' locations are
832not relative to the beginning of the buffer: annotations will be inserted
9c3054da
JB
833at their location-OFFSET+1 \(ie, the offset is treated as the position of
834the first character in the buffer)."
823139fb 835 (if (not offset)
029894b9
BG
836 (setq offset 0)
837 (setq offset (1- offset)))
838 (let ((l (reverse list)))
839 (while l
840 (goto-char (- (car (car l)) offset))
841 (insert (cdr (car l)))
842 (setq l (cdr l)))))
843
844(defun format-annotate-value (old new)
0b1ce6ba 845 "Return OLD and NEW as a \(CLOSE . OPEN) annotation pair.
029894b9
BG
846Useful as a default function for TRANSLATIONS alist when the value of the text
847property is the name of the annotation that you want to use, as it is for the
848`unknown' text property."
849 (cons (if old (list old))
850 (if new (list new))))
851
823139fb 852(defun format-annotate-region (from to translations format-fn ignore)
029894b9 853 "Generate annotations for text properties in the region.
12008438 854Search for changes between FROM and TO, and describe them with a list of
029894b9
BG
855annotations as defined by alist TRANSLATIONS and FORMAT-FN. IGNORE lists text
856properties not to consider; any text properties that are neither ignored nor
857listed in TRANSLATIONS are warned about.
858If you actually want to modify the region, give the return value of this
859function to `format-insert-annotations'.
860
861Format of the TRANSLATIONS argument:
862
863Each element is a list whose car is a PROPERTY, and the following
d0fc7e3b
RS
864elements have the form (VALUE ANNOTATIONS...).
865Whenever the property takes on the value VALUE, the annotations
029894b9
BG
866\(as formatted by FORMAT-FN) are inserted into the file.
867When the property stops having that value, the matching negated annotation
868will be inserted \(it may actually be closed earlier and reopened, if
823139fb 869necessary, to keep proper nesting).
029894b9 870
d0fc7e3b 871If VALUE is a list, then each element of the list is dealt with
029894b9
BG
872separately.
873
874If a VALUE is numeric, then it is assumed that there is a single annotation
875and each occurrence of it increments the value of the property by that number.
876Thus, given the entry \(left-margin \(4 \"indent\")), if the left margin
877changes from 4 to 12, two <indent> annotations will be generated.
878
879If the VALUE is nil, then instead of annotations, a function should be
880specified. This function is used as a default: it is called for all
881transitions not explicitly listed in the table. The function is called with
882two arguments, the OLD and NEW values of the property. It should return
d0fc7e3b 883a cons cell (CLOSE . OPEN) as `format-annotate-single-property-change' does.
029894b9 884
d0fc7e3b 885The same TRANSLATIONS structure can be used in reverse for reading files."
029894b9
BG
886 (let ((all-ans nil) ; All annotations - becomes return value
887 (open-ans nil) ; Annotations not yet closed
888 (loc nil) ; Current location
889 (not-found nil)) ; Properties that couldn't be saved
890 (while (or (null loc)
891 (and (setq loc (next-property-change loc nil to))
892 (< loc to)))
893 (or loc (setq loc from))
823139fb 894 (let* ((ans (format-annotate-location loc (= loc from) ignore translations))
029894b9
BG
895 (neg-ans (format-reorder (aref ans 0) open-ans))
896 (pos-ans (aref ans 1))
897 (ignored (aref ans 2)))
898 (setq not-found (append ignored not-found)
899 ignore (append ignored ignore))
900 ;; First do the negative (closing) annotations
901 (while neg-ans
902 ;; Check if it's missing. This can happen (eg, a numeric property
903 ;; going negative can generate closing annotations before there are
904 ;; any open). Warn user & ignore.
905 (if (not (member (car neg-ans) open-ans))
906 (message "Can't close %s: not open." (car neg-ans))
907 (while (not (equal (car neg-ans) (car open-ans)))
908 ;; To close anno. N, need to first close ans 1 to N-1,
909 ;; remembering to re-open them later.
be227f22 910 (push (car open-ans) pos-ans)
823139fb 911 (setq all-ans
029894b9
BG
912 (cons (cons loc (funcall format-fn (car open-ans) nil))
913 all-ans))
914 (setq open-ans (cdr open-ans)))
915 ;; Now remove the one we're really interested in from open list.
916 (setq open-ans (cdr open-ans))
917 ;; And put the closing annotation here.
be227f22
SM
918 (push (cons loc (funcall format-fn (car neg-ans) nil))
919 all-ans))
029894b9
BG
920 (setq neg-ans (cdr neg-ans)))
921 ;; Now deal with positive (opening) annotations
922 (let ((p pos-ans))
923 (while pos-ans
be227f22
SM
924 (push (car pos-ans) open-ans)
925 (push (cons loc (funcall format-fn (car pos-ans) t))
926 all-ans)
029894b9
BG
927 (setq pos-ans (cdr pos-ans))))))
928
929 ;; Close any annotations still open
930 (while open-ans
823139fb 931 (setq all-ans
029894b9
BG
932 (cons (cons to (funcall format-fn (car open-ans) nil))
933 all-ans))
934 (setq open-ans (cdr open-ans)))
935 (if not-found
936 (message "These text properties could not be saved:\n %s"
937 not-found))
938 (nreverse all-ans)))
939
940;;; Internal functions for format-annotate-region.
941
823139fb
DL
942(defun format-annotate-location (loc all ignore translations)
943 "Return annotation(s) needed at location LOC.
9c3054da 944This includes any properties that change between LOC - 1 and LOC.
029894b9
BG
945If ALL is true, don't look at previous location, but generate annotations for
946all non-nil properties.
947Third argument IGNORE is a list of text-properties not to consider.
d0fc7e3b 948Use the TRANSLATIONS alist (see `format-annotate-region' for doc).
029894b9
BG
949
950Return value is a vector of 3 elements:
d0fc7e3b
RS
9511. List of annotations to close
9522. List of annotations to open.
9533. List of properties that were ignored or couldn't be annotated.
954
955The annotations in lists 1 and 2 need not be strings.
956They can be whatever the FORMAT-FN in `format-annotate-region'
957can handle. If that is `enriched-make-annotation', they can be
958either strings, or lists of the form (PARAMETER VALUE)."
029894b9
BG
959 (let* ((prev-loc (1- loc))
960 (before-plist (if all nil (text-properties-at prev-loc)))
961 (after-plist (text-properties-at loc))
962 p negatives positives prop props not-found)
963 ;; make list of all property names involved
964 (setq p before-plist)
965 (while p
966 (if (not (memq (car p) props))
be227f22 967 (push (car p) props))
029894b9
BG
968 (setq p (cdr (cdr p))))
969 (setq p after-plist)
970 (while p
971 (if (not (memq (car p) props))
be227f22 972 (push (car p) props))
029894b9
BG
973 (setq p (cdr (cdr p))))
974
975 (while props
be227f22 976 (setq prop (pop props))
029894b9
BG
977 (if (memq prop ignore)
978 nil ; If it's been ignored before, ignore it now.
979 (let ((before (if all nil (car (cdr (memq prop before-plist)))))
980 (after (car (cdr (memq prop after-plist)))))
981 (if (equal before after)
982 nil ; no change; ignore
983 (let ((result (format-annotate-single-property-change
823139fb 984 prop before after translations)))
029894b9 985 (if (not result)
be227f22 986 (push prop not-found)
029894b9
BG
987 (setq negatives (nconc negatives (car result))
988 positives (nconc positives (cdr result)))))))))
989 (vector negatives positives not-found)))
990
d0fc7e3b 991(defun format-annotate-single-property-change (prop old new translations)
823139fb 992 "Return annotations for property PROP changing from OLD to NEW.
d0fc7e3b
RS
993These are searched for in the translations alist TRANSLATIONS
994 (see `format-annotate-region' for the format).
12008438
TTN
995If NEW does not appear in the list, but there is a default function,
996then call that function.
997Return a cons of the form (CLOSE . OPEN)
d0fc7e3b
RS
998where CLOSE is a list of annotations to close
999and OPEN is a list of annotations to open.
1000
1001The annotations in CLOSE and OPEN need not be strings.
1002They can be whatever the FORMAT-FN in `format-annotate-region'
1003can handle. If that is `enriched-make-annotation', they can be
1004either strings, or lists of the form (PARAMETER VALUE)."
1005
1006 (let ((prop-alist (cdr (assoc prop translations)))
029894b9
BG
1007 default)
1008 (if (not prop-alist)
1009 nil
029894b9 1010 ;; If either old or new is a list, have to treat both that way.
f0a6c717 1011 (if (and (or (listp old) (listp new))
f3bbef87 1012 (not (get prop 'format-list-atomic-p)))
f0a6c717
GM
1013 (if (or (not (format-proper-list-p old))
1014 (not (format-proper-list-p new)))
1015 (format-annotate-atomic-property-change prop-alist old new)
1016 (let* ((old (if (listp old) old (list old)))
1017 (new (if (listp new) new (list new)))
1018 (tail (format-common-tail old new))
1019 close open)
1020 (while old
1021 (setq close
1022 (append (car (format-annotate-atomic-property-change
1023 prop-alist (car old) nil))
1024 close)
1025 old (cdr old)))
1026 (while new
1027 (setq open
1028 (append (cdr (format-annotate-atomic-property-change
1029 prop-alist nil (car new)))
1030 open)
1031 new (cdr new)))
1032 (format-make-relatively-unique close open)))
029894b9
BG
1033 (format-annotate-atomic-property-change prop-alist old new)))))
1034
1035(defun format-annotate-atomic-property-change (prop-alist old new)
12008438 1036 "Internal function to annotate a single property change.
d0fc7e3b 1037PROP-ALIST is the relevant element of a TRANSLATIONS list.
029894b9 1038OLD and NEW are the values."
e5a60108
RS
1039 (let (num-ann)
1040 ;; If old and new values are numbers,
1041 ;; look for a number in PROP-ALIST.
5f525595
RS
1042 (if (and (or (null old) (numberp old))
1043 (or (null new) (numberp new)))
e5a60108
RS
1044 (progn
1045 (setq num-ann prop-alist)
1046 (while (and num-ann (not (numberp (car (car num-ann)))))
1047 (setq num-ann (cdr num-ann)))))
1048 (if num-ann
5f525595 1049 ;; Numerical annotation - use difference
34fbecde
RS
1050 (progn
1051 ;; If property is numeric, nil means 0
1052 (cond ((and (numberp old) (null new))
1053 (setq new 0))
1054 ((and (numberp new) (null old))
1055 (setq old 0)))
1056
1057 (let* ((entry (car num-ann))
1058 (increment (car entry))
1059 (n (ceiling (/ (float (- new old)) (float increment))))
1060 (anno (car (cdr entry))))
1061 (if (> n 0)
1062 (cons nil (make-list n anno))
1063 (cons (make-list (- n) anno) nil))))
e5a60108
RS
1064
1065 ;; Standard annotation
1066 (let ((close (and old (cdr (assoc old prop-alist))))
029894b9
BG
1067 (open (and new (cdr (assoc new prop-alist)))))
1068 (if (or close open)
1069 (format-make-relatively-unique close open)
1070 ;; Call "Default" function, if any
1071 (let ((default (assq nil prop-alist)))
1072 (if default
1073 (funcall (car (cdr default)) old new))))))))
1074
c84cf2bf 1075(provide 'format)
823139fb 1076
cbee283d 1077;; arch-tag: c387e9c7-a93d-47bf-89bc-8ca67e96755a
823139fb 1078;;; format.el ends here