Add arch taglines
[bpt/emacs.git] / lisp / gnus / mm-util.el
CommitLineData
95fa1ff7 1;;; mm-util.el --- Utility functions for Mule and low level things
d62d49df 2;; Copyright (C) 1998, 1999, 2000, 2001, 2002 Free Software Foundation, Inc.
c113de23
GM
3
4;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
5;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
6;; This file is part of GNU Emacs.
7
8;; GNU Emacs is free software; you can redistribute it and/or modify
9;; it under the terms of the GNU General Public License as published by
10;; the Free Software Foundation; either version 2, or (at your option)
11;; any later version.
12
13;; GNU Emacs is distributed in the hope that it will be useful,
14;; but WITHOUT ANY WARRANTY; without even the implied warranty of
15;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16;; GNU General Public License for more details.
17
18;; You should have received a copy of the GNU General Public License
19;; along with GNU Emacs; see the file COPYING. If not, write to the
20;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
21;; Boston, MA 02111-1307, USA.
22
23;;; Commentary:
24
25;;; Code:
26
56e09c09
DL
27(eval-when-compile
28 (require 'cl)
29 (defvar mm-mime-mule-charset-alist))
c113de23
GM
30(require 'mail-prsvr)
31
f53b2875
DL
32(eval-and-compile
33 (mapcar
34 (lambda (elem)
35 (let ((nfunc (intern (format "mm-%s" (car elem)))))
36 (if (fboundp (car elem))
37 (defalias nfunc (car elem))
38 (defalias nfunc (cdr elem)))))
39 '((decode-coding-string . (lambda (s a) s))
40 (encode-coding-string . (lambda (s a) s))
41 (encode-coding-region . ignore)
42 (coding-system-list . ignore)
43 (decode-coding-region . ignore)
44 (char-int . identity)
45 (device-type . ignore)
46 (coding-system-equal . equal)
47 (annotationp . ignore)
48 (set-buffer-file-coding-system . ignore)
49 (make-char
50 . (lambda (charset int)
51 (int-to-char int)))
f53b2875
DL
52 (read-charset
53 . (lambda (prompt)
54 "Return a charset."
55 (intern
56 (completing-read
57 prompt
58 (mapcar (lambda (e) (list (symbol-name (car e))))
59 mm-mime-mule-charset-alist)
60 nil t))))
95fa1ff7
SZ
61 (subst-char-in-string
62 . (lambda (from to string) ;; stolen (and renamed) from nnheader.el
63 "Replace characters in STRING from FROM to TO."
64 (let ((string (substring string 0)) ;Copy string.
65 (len (length string))
66 (idx 0))
67 ;; Replace all occurrences of FROM with TO.
68 (while (< idx len)
69 (when (= (aref string idx) from)
70 (aset string idx to))
71 (setq idx (1+ idx)))
72 string)))
f53b2875 73 (string-as-unibyte . identity)
95fa1ff7 74 (string-as-multibyte . identity)
56e09c09
DL
75 (multibyte-string-p . ignore)
76 (point-at-bol . line-beginning-position)
77 (point-at-eol . line-end-position)
78 (insert-byte . insert-char)
79 (multibyte-char-to-unibyte . identity))))
f53b2875 80
c113de23
GM
81(eval-and-compile
82 (defalias 'mm-char-or-char-int-p
95fa1ff7 83 (cond
c113de23 84 ((fboundp 'char-or-char-int-p) 'char-or-char-int-p)
95fa1ff7 85 ((fboundp 'char-valid-p) 'char-valid-p)
c113de23
GM
86 (t 'identity))))
87
95fa1ff7
SZ
88(eval-and-compile
89 (defalias 'mm-read-coding-system
90 (cond
91 ((fboundp 'read-coding-system)
92 (if (and (featurep 'xemacs)
93 (<= (string-to-number emacs-version) 21.1))
94 (lambda (prompt &optional default-coding-system)
95 (read-coding-system prompt))
96 'read-coding-system))
97 (t (lambda (prompt &optional default-coding-system)
98 "Prompt the user for a coding system."
99 (completing-read
100 prompt (mapcar (lambda (s) (list (symbol-name (car s))))
101 mm-mime-mule-charset-alist)))))))
102
c113de23
GM
103(defvar mm-coding-system-list nil)
104(defun mm-get-coding-system-list ()
105 "Get the coding system list."
106 (or mm-coding-system-list
107 (setq mm-coding-system-list (mm-coding-system-list))))
108
95fa1ff7
SZ
109(defun mm-coding-system-p (sym)
110 "Return non-nil if SYM is a coding system."
111 (or (and (fboundp 'coding-system-p) (coding-system-p sym))
112 (memq sym (mm-get-coding-system-list))))
113
c113de23 114(defvar mm-charset-synonym-alist
95fa1ff7
SZ
115 `(
116 ;; Perfectly fine? A valid MIME name, anyhow.
72eb5fc7
SZ
117 ,@(unless (mm-coding-system-p 'big5)
118 '((big5 . cn-big5)))
95fa1ff7 119 ;; Not in XEmacs, but it's not a proper MIME charset anyhow.
72eb5fc7
SZ
120 ,@(unless (mm-coding-system-p 'x-ctext)
121 '((x-ctext . ctext)))
95fa1ff7 122 ;; Apparently not defined in Emacs 20, but is a valid MIME name.
72eb5fc7
SZ
123 ,@(unless (mm-coding-system-p 'gb2312)
124 '((gb2312 . cn-gb-2312)))
125 ;; ISO-8859-15 is very similar to ISO-8859-1.
56e09c09 126 ;; But this is just wrong. --fx
72eb5fc7
SZ
127 ,@(unless (mm-coding-system-p 'iso-8859-15) ; Emacs 21 defines it.
128 '((iso-8859-15 . iso-8859-1)))
d1a7bc93
DL
129 ;; Windows-1252 is actually a superset of Latin-1. See also
130 ;; `gnus-article-dumbquotes-map'.
a1506d29 131 ,@(unless (mm-coding-system-p 'windows-1252)
72eb5fc7
SZ
132 (if (mm-coding-system-p 'cp1252)
133 '((windows-1252 . cp1252))
134 '((windows-1252 . iso-8859-1))))
20c381cf
GM
135 ;; Windows-1250 is a variant of Latin-2 heavily used by Microsoft
136 ;; Outlook users in Czech republic. Use this to allow reading of their
137 ;; e-mails. cp1250 should be defined by M-x codepage-setup.
56e09c09
DL
138
139 ;; This is not TRT, the MIME name, windows-1250, should be an
140 ;; alias, and cp1250 should have a mime-charset property, per
141 ;; code-page.el. -- fx
72eb5fc7
SZ
142 ,@(if (and (not (mm-coding-system-p 'windows-1250))
143 (mm-coding-system-p 'cp1250))
144 '((windows-1250 . cp1250)))
95fa1ff7 145 )
c113de23
GM
146 "A mapping from invalid charset names to the real charset names.")
147
c113de23 148(defvar mm-binary-coding-system
95fa1ff7 149 (cond
c113de23
GM
150 ((mm-coding-system-p 'binary) 'binary)
151 ((mm-coding-system-p 'no-conversion) 'no-conversion)
152 (t nil))
153 "100% binary coding system.")
154
155(defvar mm-text-coding-system
156 (or (if (memq system-type '(windows-nt ms-dos ms-windows))
157 (and (mm-coding-system-p 'raw-text-dos) 'raw-text-dos)
158 (and (mm-coding-system-p 'raw-text) 'raw-text))
159 mm-binary-coding-system)
160 "Text-safe coding system (For removing ^M).")
161
162(defvar mm-text-coding-system-for-write nil
163 "Text coding system for write.")
164
165(defvar mm-auto-save-coding-system
95fa1ff7 166 (cond
56e09c09
DL
167 ((mm-coding-system-p 'utf-8-emacs)
168 (if (memq system-type '(windows-nt ms-dos ms-windows))
169 (if (mm-coding-system-p 'utf-8-emacs-dos)
170 'utf-8-emacs-dos mm-binary-coding-system)
171 'utf-8-emacs))
c113de23
GM
172 ((mm-coding-system-p 'emacs-mule)
173 (if (memq system-type '(windows-nt ms-dos ms-windows))
95fa1ff7 174 (if (mm-coding-system-p 'emacs-mule-dos)
c113de23
GM
175 'emacs-mule-dos mm-binary-coding-system)
176 'emacs-mule))
177 ((mm-coding-system-p 'escape-quoted) 'escape-quoted)
178 (t mm-binary-coding-system))
179 "Coding system of auto save file.")
180
95fa1ff7 181(defvar mm-universal-coding-system mm-auto-save-coding-system
47b63dfa 182 "The universal coding system.")
95fa1ff7
SZ
183
184;; Fixme: some of the cars here aren't valid MIME charsets. That
185;; should only matter with XEmacs, though.
186(defvar mm-mime-mule-charset-alist
187 `((us-ascii ascii)
188 (iso-8859-1 latin-iso8859-1)
189 (iso-8859-2 latin-iso8859-2)
190 (iso-8859-3 latin-iso8859-3)
191 (iso-8859-4 latin-iso8859-4)
192 (iso-8859-5 cyrillic-iso8859-5)
193 ;; Non-mule (X)Emacs uses the last mule-charset for 8bit characters.
194 ;; The fake mule-charset, gnus-koi8-r, tells Gnus that the default
195 ;; charset is koi8-r, not iso-8859-5.
196 (koi8-r cyrillic-iso8859-5 gnus-koi8-r)
197 (iso-8859-6 arabic-iso8859-6)
198 (iso-8859-7 greek-iso8859-7)
199 (iso-8859-8 hebrew-iso8859-8)
200 (iso-8859-9 latin-iso8859-9)
201 (iso-8859-14 latin-iso8859-14)
202 (iso-8859-15 latin-iso8859-15)
203 (viscii vietnamese-viscii-lower)
204 (iso-2022-jp latin-jisx0201 japanese-jisx0208 japanese-jisx0208-1978)
205 (euc-kr korean-ksc5601)
206 (gb2312 chinese-gb2312)
207 (big5 chinese-big5-1 chinese-big5-2)
208 (tibetan tibetan)
209 (thai-tis620 thai-tis620)
210 (iso-2022-7bit ethiopic arabic-1-column arabic-2-column)
211 (iso-2022-jp-2 latin-iso8859-1 greek-iso8859-7
212 latin-jisx0201 japanese-jisx0208-1978
213 chinese-gb2312 japanese-jisx0208
214 korean-ksc5601 japanese-jisx0212
215 katakana-jisx0201)
216 (iso-2022-int-1 latin-iso8859-1 greek-iso8859-7
217 latin-jisx0201 japanese-jisx0208-1978
218 chinese-gb2312 japanese-jisx0208
219 korean-ksc5601 japanese-jisx0212
220 chinese-cns11643-1 chinese-cns11643-2)
221 (iso-2022-int-1 latin-iso8859-1 latin-iso8859-2
222 cyrillic-iso8859-5 greek-iso8859-7
223 latin-jisx0201 japanese-jisx0208-1978
224 chinese-gb2312 japanese-jisx0208
225 korean-ksc5601 japanese-jisx0212
226 chinese-cns11643-1 chinese-cns11643-2
227 chinese-cns11643-3 chinese-cns11643-4
228 chinese-cns11643-5 chinese-cns11643-6
229 chinese-cns11643-7)
230 ,(if (or (not (fboundp 'charsetp)) ;; non-Mule case
231 (charsetp 'unicode-a)
232 (not (mm-coding-system-p 'mule-utf-8)))
233 '(utf-8 unicode-a unicode-b unicode-c unicode-d unicode-e)
234 ;; If we have utf-8 we're in Mule 5+.
235 (append '(utf-8)
236 (delete 'ascii
237 (coding-system-get 'mule-utf-8 'safe-charsets)))))
238 "Alist of MIME-charset/MULE-charsets.")
239
240;; Correct by construction, but should be unnecessary:
241;; XEmacs hates it.
242(when (and (not (featurep 'xemacs))
243 (fboundp 'coding-system-list)
244 (fboundp 'sort-coding-systems))
245 (setq mm-mime-mule-charset-alist
246 (apply
247 'nconc
248 (mapcar
249 (lambda (cs)
56e09c09
DL
250 (when (and (or (coding-system-get cs :mime-charset) ; Emacs 22
251 (coding-system-get cs 'mime-charset))
95fa1ff7 252 (not (eq t (coding-system-get cs 'safe-charsets))))
56e09c09
DL
253 (list (cons (or (coding-system-get cs :mime-charset)
254 (coding-system-get cs 'mime-charset))
95fa1ff7
SZ
255 (delq 'ascii
256 (coding-system-get cs 'safe-charsets))))))
257 (sort-coding-systems (coding-system-list 'base-only))))))
258
47b63dfa
SZ
259(defvar mm-hack-charsets '(iso-8859-15 iso-2022-jp-2)
260 "A list of special charsets.
261Valid elements include:
262`iso-8859-15' convert ISO-8859-1, -9 to ISO-8859-15 if ISO-8859-15 exists.
263`iso-2022-jp-2' convert ISO-2022-jp to ISO-2022-jp-2 if ISO-2022-jp-2 exists."
264)
265
a1506d29 266(defvar mm-iso-8859-15-compatible
47b63dfa
SZ
267 '((iso-8859-1 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE")
268 (iso-8859-9 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE\xD0\xDD\xDE\xF0\xFD\xFE"))
269 "ISO-8859-15 exchangeable coding systems and inconvertible characters.")
270
271(defvar mm-iso-8859-x-to-15-table
272 (and (fboundp 'coding-system-p)
273 (mm-coding-system-p 'iso-8859-15)
a1506d29 274 (mapcar
47b63dfa
SZ
275 (lambda (cs)
276 (if (mm-coding-system-p (car cs))
a1506d29 277 (let ((c (string-to-char
47b63dfa
SZ
278 (decode-coding-string "\341" (car cs)))))
279 (cons (char-charset c)
280 (cons
a1506d29 281 (- (string-to-char
47b63dfa 282 (decode-coding-string "\341" 'iso-8859-15)) c)
a1506d29 283 (string-to-list (decode-coding-string (car (cdr cs))
47b63dfa
SZ
284 (car cs))))))
285 '(gnus-charset 0)))
286 mm-iso-8859-15-compatible))
287 "A table of the difference character between ISO-8859-X and ISO-8859-15.")
288
289(defvar mm-coding-system-priorities nil
290 "Preferred coding systems for encoding outgoing mails.
291
292More than one suitable coding systems may be found for some texts. By
293default, a coding system with the highest priority is used to encode
294outgoing mails (see `sort-coding-systems'). If this variable is set,
295it overrides the default priority. For example, Japanese users may
296prefer iso-2022-jp to japanese-shift-jis:
297
298\(setq mm-coding-system-priorities
299 '(iso-2022-jp iso-2022-jp-2 japanese-shift-jis utf-8))
300")
301
56e09c09 302;; Why on earth was this broken out? -- fx
1f7d2e14
SZ
303(defvar mm-use-find-coding-systems-region
304 (fboundp 'find-coding-systems-region)
305 "Use `find-coding-systems-region' to find proper coding systems.")
306
c113de23
GM
307;;; Internal variables:
308
309;;; Functions:
310
311(defun mm-mule-charset-to-mime-charset (charset)
1c57d870 312 "Return the MIME charset corresponding to the given Mule CHARSET."
95fa1ff7
SZ
313 (if (fboundp 'find-coding-systems-for-charsets)
314 (let (mime)
315 (dolist (cs (find-coding-systems-for-charsets (list charset)))
316 (unless mime
317 (when cs
56e09c09
DL
318 (setq mime (or (coding-system-get cs :mime-charset)
319 (coding-system-get cs 'mime-charset))))))
95fa1ff7
SZ
320 mime)
321 (let ((alist mm-mime-mule-charset-alist)
322 out)
323 (while alist
324 (when (memq charset (cdar alist))
325 (setq out (caar alist)
326 alist nil))
327 (pop alist))
328 out)))
c113de23
GM
329
330(defun mm-charset-to-coding-system (charset &optional lbt)
331 "Return coding-system corresponding to CHARSET.
332CHARSET is a symbol naming a MIME charset.
333If optional argument LBT (`unix', `dos' or `mac') is specified, it is
334used as the line break code type of the coding system."
335 (when (stringp charset)
336 (setq charset (intern (downcase charset))))
c113de23
GM
337 (when lbt
338 (setq charset (intern (format "%s-%s" charset lbt))))
339 (cond
47b63dfa
SZ
340 ((null charset)
341 charset)
c113de23
GM
342 ;; Running in a non-MULE environment.
343 ((null (mm-get-coding-system-list))
344 charset)
345 ;; ascii
346 ((eq charset 'us-ascii)
347 'ascii)
1c57d870
DL
348 ;; Check to see whether we can handle this charset. (This depends
349 ;; on there being some coding system matching each `mime-charset'
95fa1ff7
SZ
350 ;; property defined, as there should be.)
351 ((and (mm-coding-system-p charset)
352;;; Doing this would potentially weed out incorrect charsets.
353;;; charset
354;;; (eq charset (coding-system-get charset 'mime-charset))
355 )
c113de23 356 charset)
95fa1ff7 357 ;; Translate invalid charsets.
d62d49df
SZ
358 ((let ((cs (cdr (assq charset mm-charset-synonym-alist))))
359 (and cs (mm-coding-system-p charset) cs)))
95fa1ff7
SZ
360 ;; Last resort: search the coding system list for entries which
361 ;; have the right mime-charset in case the canonical name isn't
362 ;; defined (though it should be).
363 ((let (cs)
364 ;; mm-get-coding-system-list returns a list of cs without lbt.
365 ;; Do we need -lbt?
366 (dolist (c (mm-get-coding-system-list))
367 (if (and (null cs)
56e09c09
DL
368 (eq charset (or (coding-system-get c :mime-charset)
369 (coding-system-get c 'mime-charset))))
95fa1ff7
SZ
370 (setq cs c)))
371 cs))))
372
373(defsubst mm-replace-chars-in-string (string from to)
374 (mm-subst-char-in-string from to string))
375
376(eval-and-compile
377 (defvar mm-emacs-mule (and (not (featurep 'xemacs))
378 (boundp 'default-enable-multibyte-characters)
379 default-enable-multibyte-characters
380 (fboundp 'set-buffer-multibyte))
56e09c09 381 "True in Emacs with Mule.")
95fa1ff7
SZ
382
383 (if mm-emacs-mule
384 (defun mm-enable-multibyte ()
385 "Set the multibyte flag of the current buffer.
1c57d870
DL
386Only do this if the default value of `enable-multibyte-characters' is
387non-nil. This is a no-op in XEmacs."
95fa1ff7
SZ
388 (set-buffer-multibyte t))
389 (defalias 'mm-enable-multibyte 'ignore))
c113de23 390
95fa1ff7
SZ
391 (if mm-emacs-mule
392 (defun mm-disable-multibyte ()
393 "Unset the multibyte flag of in the current buffer.
1c57d870 394This is a no-op in XEmacs."
95fa1ff7 395 (set-buffer-multibyte nil))
56e09c09 396 (defalias 'mm-disable-multibyte 'ignore)))
052802c1 397
c113de23
GM
398(defun mm-preferred-coding-system (charset)
399 ;; A typo in some Emacs versions.
47b63dfa
SZ
400 (or (get-charset-property charset 'preferred-coding-system)
401 (get-charset-property charset 'prefered-coding-system)))
c113de23
GM
402
403(defun mm-charset-after (&optional pos)
404 "Return charset of a character in current buffer at position POS.
405If POS is nil, it defauls to the current point.
406If POS is out of range, the value is nil.
407If the charset is `composition', return the actual one."
052802c1
DL
408 (let ((char (char-after pos)) charset)
409 (if (< (mm-char-int char) 128)
410 (setq charset 'ascii)
411 ;; charset-after is fake in some Emacsen.
412 (setq charset (and (fboundp 'char-charset) (char-charset char)))
56e09c09 413 (if (eq charset 'composition) ; Mule 4
052802c1
DL
414 (let ((p (or pos (point))))
415 (cadr (find-charset-region p (1+ p))))
416 (if (and charset (not (memq charset '(ascii eight-bit-control
417 eight-bit-graphic))))
418 charset
419 (or
420 mail-parse-mule-charset ;; cached mule-charset
421 (progn
422 (setq mail-parse-mule-charset
423 (and (boundp 'current-language-environment)
95fa1ff7
SZ
424 (car (last
425 (assq 'charset
426 (assoc current-language-environment
427 language-info-alist))))))
052802c1
DL
428 (if (or (not mail-parse-mule-charset)
429 (eq mail-parse-mule-charset 'ascii))
430 (setq mail-parse-mule-charset
431 (or (car (last (assq mail-parse-charset
432 mm-mime-mule-charset-alist)))
bf3b936f 433 ;; Fixme: don't fix that!
052802c1
DL
434 'latin-iso8859-1)))
435 mail-parse-mule-charset)))))))
c113de23
GM
436
437(defun mm-mime-charset (charset)
1c57d870 438 "Return the MIME charset corresponding to the given Mule CHARSET."
95fa1ff7
SZ
439 (if (eq charset 'unknown)
440 (error "The message contains non-printable characters, please use attachment"))
052802c1 441 (if (and (fboundp 'coding-system-get) (fboundp 'get-charset-property))
c113de23
GM
442 ;; This exists in Emacs 20.
443 (or
444 (and (mm-preferred-coding-system charset)
56e09c09
DL
445 (or (coding-system-get
446 (mm-preferred-coding-system charset) :mime-charset)
447 (coding-system-get
448 (mm-preferred-coding-system charset) 'mime-charset)))
c113de23
GM
449 (and (eq charset 'ascii)
450 'us-ascii)
95fa1ff7 451 (mm-preferred-coding-system charset)
c113de23
GM
452 (mm-mule-charset-to-mime-charset charset))
453 ;; This is for XEmacs.
454 (mm-mule-charset-to-mime-charset charset)))
455
456(defun mm-delete-duplicates (list)
457 "Simple substitute for CL `delete-duplicates', testing with `equal'."
458 (let (result head)
459 (while list
460 (setq head (car list))
461 (setq list (delete head list))
462 (setq result (cons head result)))
463 (nreverse result)))
464
95fa1ff7
SZ
465;; It's not clear whether this is supposed to mean the global or local
466;; setting. I think it's used inconsistently. -- fx
c113de23
GM
467(defsubst mm-multibyte-p ()
468 "Say whether multibyte is enabled."
052802c1
DL
469 (if (and (not (featurep 'xemacs))
470 (boundp 'enable-multibyte-characters))
471 enable-multibyte-characters
472 (featurep 'mule)))
c113de23 473
47b63dfa
SZ
474(defun mm-iso-8859-x-to-15-region (&optional b e)
475 (if (fboundp 'char-charset)
476 (let (charset item c inconvertible)
477 (save-restriction
478 (if e (narrow-to-region b e))
479 (goto-char (point-min))
480 (skip-chars-forward "\0-\177")
481 (while (not (eobp))
a1506d29
JB
482 (cond
483 ((not (setq item (assq (char-charset (setq c (char-after)))
47b63dfa
SZ
484 mm-iso-8859-x-to-15-table)))
485 (forward-char))
486 ((memq c (cdr (cdr item)))
487 (setq inconvertible t)
488 (forward-char))
489 (t
490 (insert (prog1 (+ c (car (cdr item))) (delete-char 1))))
491 (skip-chars-forward "\0-\177"))))
492 (not inconvertible))))
493
494(defun mm-sort-coding-systems-predicate (a b)
495 (> (length (memq a mm-coding-system-priorities))
496 (length (memq b mm-coding-system-priorities))))
497
498(defun mm-find-mime-charset-region (b e &optional hack-charsets)
95fa1ff7 499 "Return the MIME charsets needed to encode the region between B and E.
f0529b5b 500nil means ASCII, a single-element list represents an appropriate MIME
95fa1ff7 501charset, and a longer list means no appropriate charset."
47b63dfa
SZ
502 (let (charsets)
503 ;; The return possibilities of this function are a mess...
504 (or (and (mm-multibyte-p)
1f7d2e14 505 mm-use-find-coding-systems-region
47b63dfa
SZ
506 ;; Find the mime-charset of the most preferred coding
507 ;; system that has one.
508 (let ((systems (find-coding-systems-region b e)))
509 (when mm-coding-system-priorities
a1506d29 510 (setq systems
47b63dfa
SZ
511 (sort systems 'mm-sort-coding-systems-predicate)))
512 ;; Fixme: The `mime-charset' (`x-ctext') of `compound-text'
513 ;; is not in the IANA list.
514 (setq systems (delq 'compound-text systems))
515 (unless (equal systems '(undecided))
516 (while systems
56e09c09
DL
517 (let* ((head (pop systems))
518 (cs (or (coding-system-get head :mime-charset)
519 (coding-system-get head 'mime-charset))))
47b63dfa
SZ
520 (if cs
521 (setq systems nil
522 charsets (list cs))))))
523 charsets))
524 ;; Otherwise we're not multibyte, XEmacs or a single coding
525 ;; system won't cover it.
a1506d29 526 (setq charsets
47b63dfa
SZ
527 (mm-delete-duplicates
528 (mapcar 'mm-mime-charset
529 (delq 'ascii
530 (mm-find-charset-region b e))))))
531 (if (and (memq 'iso-8859-15 charsets)
532 (memq 'iso-8859-15 hack-charsets)
533 (save-excursion (mm-iso-8859-x-to-15-region b e)))
534 (mapcar (lambda (x) (setq charsets (delq (car x) charsets)))
535 mm-iso-8859-15-compatible))
536 (if (and (memq 'iso-2022-jp-2 charsets)
537 (memq 'iso-2022-jp-2 hack-charsets))
538 (setq charsets (delq 'iso-2022-jp charsets)))
539 charsets))
95fa1ff7 540
c113de23
GM
541(defmacro mm-with-unibyte-buffer (&rest forms)
542 "Create a temporary buffer, and evaluate FORMS there like `progn'.
1c57d870
DL
543Use unibyte mode for this."
544 `(let (default-enable-multibyte-characters)
545 (with-temp-buffer ,@forms)))
c113de23
GM
546(put 'mm-with-unibyte-buffer 'lisp-indent-function 0)
547(put 'mm-with-unibyte-buffer 'edebug-form-spec '(body))
548
549(defmacro mm-with-unibyte-current-buffer (&rest forms)
56e09c09 550 "Evaluate FORMS with current buffer temporarily made unibyte.
1c57d870
DL
551Also bind `default-enable-multibyte-characters' to nil.
552Equivalent to `progn' in XEmacs"
95fa1ff7
SZ
553 (let ((multibyte (make-symbol "multibyte"))
554 (buffer (make-symbol "buffer")))
a1506d29 555 `(if mm-emacs-mule
95fa1ff7
SZ
556 (let ((,multibyte enable-multibyte-characters)
557 (,buffer (current-buffer)))
1c57d870
DL
558 (unwind-protect
559 (let (default-enable-multibyte-characters)
560 (set-buffer-multibyte nil)
561 ,@forms)
95fa1ff7 562 (set-buffer ,buffer)
1c57d870 563 (set-buffer-multibyte ,multibyte)))
95fa1ff7 564 (let (default-enable-multibyte-characters)
1c57d870 565 ,@forms))))
c113de23
GM
566(put 'mm-with-unibyte-current-buffer 'lisp-indent-function 0)
567(put 'mm-with-unibyte-current-buffer 'edebug-form-spec '(body))
568
569(defmacro mm-with-unibyte (&rest forms)
1c57d870
DL
570 "Eval the FORMS with the default value of `enable-multibyte-characters' nil, ."
571 `(let (default-enable-multibyte-characters)
572 ,@forms))
c113de23
GM
573(put 'mm-with-unibyte 'lisp-indent-function 0)
574(put 'mm-with-unibyte 'edebug-form-spec '(body))
575
576(defun mm-find-charset-region (b e)
1c57d870 577 "Return a list of Emacs charsets in the region B to E."
c113de23
GM
578 (cond
579 ((and (mm-multibyte-p)
95fa1ff7 580 (fboundp 'find-charset-region))
c113de23 581 ;; Remove composition since the base charsets have been included.
95fa1ff7
SZ
582 ;; Remove eight-bit-*, treat them as ascii.
583 (let ((css (find-charset-region b e)))
584 (mapcar (lambda (cs) (setq css (delq cs css)))
585 '(composition eight-bit-control eight-bit-graphic
586 control-1))
587 css))
052802c1
DL
588 (t
589 ;; We are in a unibyte buffer or XEmacs non-mule, so we futz around a bit.
c113de23
GM
590 (save-excursion
591 (save-restriction
592 (narrow-to-region b e)
593 (goto-char (point-min))
594 (skip-chars-forward "\0-\177")
595 (if (eobp)
596 '(ascii)
052802c1
DL
597 (let (charset)
598 (setq charset
599 (and (boundp 'current-language-environment)
95fa1ff7
SZ
600 (car (last (assq 'charset
601 (assoc current-language-environment
052802c1
DL
602 language-info-alist))))))
603 (if (eq charset 'ascii) (setq charset nil))
604 (or charset
605 (setq charset
606 (car (last (assq mail-parse-charset
607 mm-mime-mule-charset-alist)))))
608 (list 'ascii (or charset 'latin-iso8859-1)))))))))
c113de23
GM
609
610(if (fboundp 'shell-quote-argument)
611 (defalias 'mm-quote-arg 'shell-quote-argument)
612 (defun mm-quote-arg (arg)
613 "Return a version of ARG that is safe to evaluate in a shell."
614 (let ((pos 0) new-pos accum)
615 ;; *** bug: we don't handle newline characters properly
616 (while (setq new-pos (string-match "[]*[;!'`\"$\\& \t{} |()<>]" arg pos))
617 (push (substring arg pos new-pos) accum)
618 (push "\\" accum)
619 (push (list (aref arg new-pos)) accum)
620 (setq pos (1+ new-pos)))
621 (if (= pos 0)
622 arg
623 (apply 'concat (nconc (nreverse accum) (list (substring arg pos))))))))
624
625(defun mm-auto-mode-alist ()
626 "Return an `auto-mode-alist' with only the .gz (etc) thingies."
627 (let ((alist auto-mode-alist)
628 out)
629 (while alist
630 (when (listp (cdar alist))
631 (push (car alist) out))
632 (pop alist))
633 (nreverse out)))
634
635(defvar mm-inhibit-file-name-handlers
244d58ba 636 '(jka-compr-handler image-file-handler)
c113de23
GM
637 "A list of handlers doing (un)compression (etc) thingies.")
638
639(defun mm-insert-file-contents (filename &optional visit beg end replace
640 inhibit)
641 "Like `insert-file-contents', q.v., but only reads in the file.
642A buffer may be modified in several ways after reading into the buffer due
643to advanced Emacs features, such as file-name-handlers, format decoding,
644find-file-hooks, etc.
56e09c09 645If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'.
c113de23
GM
646 This function ensures that none of these modifications will take place."
647 (let ((format-alist nil)
648 (auto-mode-alist (if inhibit nil (mm-auto-mode-alist)))
649 (default-major-mode 'fundamental-mode)
650 (enable-local-variables nil)
95fa1ff7 651 (after-insert-file-functions nil)
c113de23
GM
652 (enable-local-eval nil)
653 (find-file-hooks nil)
95fa1ff7 654 (inhibit-file-name-operation (if inhibit
c113de23
GM
655 'insert-file-contents
656 inhibit-file-name-operation))
657 (inhibit-file-name-handlers
658 (if inhibit
95fa1ff7 659 (append mm-inhibit-file-name-handlers
c113de23
GM
660 inhibit-file-name-handlers)
661 inhibit-file-name-handlers)))
662 (insert-file-contents filename visit beg end replace)))
663
664(defun mm-append-to-file (start end filename &optional codesys inhibit)
665 "Append the contents of the region to the end of file FILENAME.
666When called from a function, expects three arguments,
667START, END and FILENAME. START and END are buffer positions
668saying what text to write.
669Optional fourth argument specifies the coding system to use when
670encoding the file.
671If INHIBIT is non-nil, inhibit mm-inhibit-file-name-handlers."
95fa1ff7
SZ
672 (let ((coding-system-for-write
673 (or codesys mm-text-coding-system-for-write
c113de23 674 mm-text-coding-system))
95fa1ff7 675 (inhibit-file-name-operation (if inhibit
c113de23
GM
676 'append-to-file
677 inhibit-file-name-operation))
678 (inhibit-file-name-handlers
679 (if inhibit
95fa1ff7 680 (append mm-inhibit-file-name-handlers
c113de23
GM
681 inhibit-file-name-handlers)
682 inhibit-file-name-handlers)))
683 (append-to-file start end filename)))
684
95fa1ff7 685(defun mm-write-region (start end filename &optional append visit lockname
c113de23
GM
686 coding-system inhibit)
687
688 "Like `write-region'.
689If INHIBIT is non-nil, inhibit mm-inhibit-file-name-handlers."
95fa1ff7
SZ
690 (let ((coding-system-for-write
691 (or coding-system mm-text-coding-system-for-write
c113de23 692 mm-text-coding-system))
95fa1ff7 693 (inhibit-file-name-operation (if inhibit
c113de23
GM
694 'write-region
695 inhibit-file-name-operation))
696 (inhibit-file-name-handlers
697 (if inhibit
95fa1ff7 698 (append mm-inhibit-file-name-handlers
c113de23
GM
699 inhibit-file-name-handlers)
700 inhibit-file-name-handlers)))
701 (write-region start end filename append visit lockname)))
702
95fa1ff7
SZ
703(defun mm-image-load-path (&optional package)
704 (let (dir result)
705 (dolist (path load-path (nreverse result))
706 (if (file-directory-p
707 (setq dir (concat (file-name-directory
708 (directory-file-name path))
709 "etc/" (or package "gnus/"))))
710 (push dir result))
711 (push path result))))
712
3efe5554
SZ
713;; It is not a MIME function, but some MIME functions use it.
714(defalias 'mm-make-temp-file
715 (if (fboundp 'make-temp-file)
716 'make-temp-file
717 (lambda (prefix &optional dir-flag)
718 (let ((file (expand-file-name
719 (make-temp-name prefix)
720 (if (fboundp 'temp-directory)
721 (temp-directory)
722 temporary-file-directory))))
723 (if dir-flag
724 (make-directory file))
725 file))))
726
c113de23
GM
727(provide 'mm-util)
728
ab5796a9 729;;; arch-tag: 94dc5388-825d-4fd1-bfa5-2100aa351238
c113de23 730;;; mm-util.el ends here