1 ;;; mm-util.el --- Utility functions for Mule and low level things
3 ;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
4 ;; 2005, 2006, 2007, 2008 Free Software Foundation, Inc.
6 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
7 ;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
8 ;; This file is part of GNU Emacs.
10 ;; GNU Emacs is free software: you can redistribute it and/or modify
11 ;; it under the terms of the GNU General Public License as published by
12 ;; the Free Software Foundation, either version 3 of the License, or
13 ;; (at your option) any later version.
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.
20 ;; You should have received a copy of the GNU General Public License
21 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
29 (unless (fboundp 'declare-function) (defmacro declare-function (&rest r))))
31 (eval-when-compile (require 'cl))
35 (if (featurep 'xemacs)
36 (unless (ignore-errors
37 (require 'timer-funcs))
41 (defvar mm-mime-mule-charset-alist )
46 (let ((nfunc (intern (format "mm-%s" (car elem)))))
47 (if (fboundp (car elem))
48 (defalias nfunc (car elem))
49 (defalias nfunc (cdr elem)))))
50 '((coding-system-list . ignore)
52 (coding-system-equal . equal)
53 (annotationp . ignore)
54 (set-buffer-file-coding-system . ignore)
61 (mapcar (lambda (e) (list (symbol-name (car e))))
62 mm-mime-mule-charset-alist)
65 . (lambda (from to string &optional inplace)
66 ;; stolen (and renamed) from nnheader.el
67 "Replace characters in STRING from FROM to TO.
68 Unless optional argument INPLACE is non-nil, return a new string."
69 (let ((string (if inplace string (copy-sequence string)))
72 ;; Replace all occurrences of FROM with TO.
74 (when (= (aref string idx) from)
79 . (lambda (string regexp rep &optional literal)
80 "See `replace-regexp-in-string', only the order of args differs."
81 (replace-regexp-in-string regexp rep string nil literal)))
82 (string-as-unibyte . identity)
83 (string-make-unibyte . identity)
84 ;; string-as-multibyte often doesn't really do what you think it does.
86 ;; (aref (string-as-multibyte "\201") 0) -> 129 (aka ?\201)
87 ;; (aref (string-as-multibyte "\300") 0) -> 192 (aka ?\300)
88 ;; (aref (string-as-multibyte "\300\201") 0) -> 192 (aka ?\300)
89 ;; (aref (string-as-multibyte "\300\201") 1) -> 129 (aka ?\201)
91 ;; (aref (string-as-multibyte "\201\300") 0) -> 2240
92 ;; (aref (string-as-multibyte "\201\300") 1) -> <error>
93 ;; Better use string-to-multibyte or encode-coding-string.
94 ;; If you really need string-as-multibyte somewhere it's usually
95 ;; because you're using the internal emacs-mule representation (maybe
96 ;; because you're using string-as-unibyte somewhere), which is
97 ;; generally a problem in itself.
98 ;; Here is an approximate equivalence table to help think about it:
99 ;; (string-as-multibyte s) ~= (decode-coding-string s 'emacs-mule)
100 ;; (string-to-multibyte s) ~= (decode-coding-string s 'binary)
101 ;; (string-make-multibyte s) ~= (decode-coding-string s locale-coding-system)
102 (string-as-multibyte . identity)
103 (multibyte-string-p . ignore)
104 (insert-byte . insert-char)
105 (multibyte-char-to-unibyte . identity)
106 (set-buffer-multibyte . ignore)
108 . (lambda (buffer-name)
109 "Returns non-nil if a buffer named BUFFER-NAME gets a special frame."
110 (and special-display-function
111 (or (and (member buffer-name special-display-buffer-names) t)
112 (cdr (assoc buffer-name special-display-buffer-names))
114 (dolist (elem special-display-regexps)
116 (string-match elem buffer-name)
120 (string-match (car elem) buffer-name)
121 (throw 'return (cdr elem))))))))))))
124 (if (featurep 'xemacs)
125 (if (featurep 'file-coding)
126 ;; Don't modify string if CODING-SYSTEM is nil.
128 (defun mm-decode-coding-string (str coding-system)
130 (decode-coding-string str coding-system)
132 (defun mm-encode-coding-string (str coding-system)
134 (encode-coding-string str coding-system)
136 (defun mm-decode-coding-region (start end coding-system)
138 (decode-coding-region start end coding-system)))
139 (defun mm-encode-coding-region (start end coding-system)
141 (encode-coding-region start end coding-system))))
142 (defun mm-decode-coding-string (str coding-system) str)
143 (defun mm-encode-coding-string (str coding-system) str)
144 (defalias 'mm-decode-coding-region 'ignore)
145 (defalias 'mm-encode-coding-region 'ignore))
146 (defalias 'mm-decode-coding-string 'decode-coding-string)
147 (defalias 'mm-encode-coding-string 'encode-coding-string)
148 (defalias 'mm-decode-coding-region 'decode-coding-region)
149 (defalias 'mm-encode-coding-region 'encode-coding-region)))
151 (defalias 'mm-string-to-multibyte
155 ((fboundp 'string-to-multibyte)
156 'string-to-multibyte)
159 "Return a multibyte string with the same individual chars as string."
161 (lambda (ch) (mm-string-as-multibyte (char-to-string ch)))
165 (defalias 'mm-char-or-char-int-p
167 ((fboundp 'char-or-char-int-p) 'char-or-char-int-p)
168 ((fboundp 'char-valid-p) 'char-valid-p)
171 ;; Fixme: This seems always to be used to read a MIME charset, so it
172 ;; should be re-named and fixed (in Emacs) to offer completion only on
173 ;; proper charset names (base coding systems which have a
174 ;; mime-charset defined). XEmacs doesn't believe in mime-charset;
176 ;; `(or (coding-system-get 'iso-8859-1 'mime-charset)
177 ;; (coding-system-get 'iso-8859-1 :mime-charset))'
178 ;; Actually, there should be an `mm-coding-system-mime-charset'.
180 (defalias 'mm-read-coding-system
182 ((fboundp 'read-coding-system)
183 (if (and (featurep 'xemacs)
184 (<= (string-to-number emacs-version) 21.1))
185 (lambda (prompt &optional default-coding-system)
186 (read-coding-system prompt))
187 'read-coding-system))
188 (t (lambda (prompt &optional default-coding-system)
189 "Prompt the user for a coding system."
191 prompt (mapcar (lambda (s) (list (symbol-name (car s))))
192 mm-mime-mule-charset-alist)))))))
194 (defvar mm-coding-system-list nil)
195 (defun mm-get-coding-system-list ()
196 "Get the coding system list."
197 (or mm-coding-system-list
198 (setq mm-coding-system-list (mm-coding-system-list))))
200 (defun mm-coding-system-p (cs)
201 "Return non-nil if CS is a symbol naming a coding system.
202 In XEmacs, also return non-nil if CS is a coding system object.
203 If CS is available, return CS itself in Emacs, and return a coding
204 system object in XEmacs."
205 (if (fboundp 'find-coding-system)
206 (and cs (find-coding-system cs))
207 (if (fboundp 'coding-system-p)
208 (when (coding-system-p cs)
211 (car (memq cs (mm-get-coding-system-list))))))
213 (defun mm-codepage-setup (number &optional alias)
214 "Create a coding system cpNUMBER.
215 The coding system is created using `codepage-setup'. If ALIAS is
216 non-nil, an alias is created and added to
217 `mm-charset-synonym-alist'. If ALIAS is a string, it's used as
218 the alias. Else windows-NUMBER is used."
220 (let ((completion-ignore-case t)
221 (candidates (if (fboundp 'cp-supported-codepages)
222 (cp-supported-codepages)
223 ;; Removed in Emacs 23 (unicode), so signal an error:
224 (error "`codepage-setup' not present in this Emacs version."))))
225 (list (completing-read "Setup DOS Codepage: (default 437) " candidates
226 nil t nil nil "437"))))
228 (setq alias (if (stringp alias)
230 (intern (format "windows-%s" number)))))
231 (let* ((cp (intern (format "cp%s" number))))
232 (unless (mm-coding-system-p cp)
233 (if (fboundp 'codepage-setup) ; silence compiler
234 (codepage-setup number)
235 (error "`codepage-setup' not present in this Emacs version.")))
237 ;; Don't add alias if setup of cp failed.
238 (mm-coding-system-p cp))
239 (add-to-list 'mm-charset-synonym-alist (cons alias cp)))))
241 (defvar mm-charset-synonym-alist
243 ;; Not in XEmacs, but it's not a proper MIME charset anyhow.
244 ,@(unless (mm-coding-system-p 'x-ctext)
245 '((x-ctext . ctext)))
246 ;; ISO-8859-15 is very similar to ISO-8859-1. But it's _different_ in 8
248 ,@(unless (mm-coding-system-p 'iso-8859-15)
249 '((iso-8859-15 . iso-8859-1)))
250 ;; BIG-5HKSCS is similar to, but different than, BIG-5.
251 ,@(unless (mm-coding-system-p 'big5-hkscs)
252 '((big5-hkscs . big5)))
253 ;; A Microsoft misunderstanding.
254 ,@(when (and (not (mm-coding-system-p 'unicode))
255 (mm-coding-system-p 'utf-16-le))
256 '((unicode . utf-16-le)))
257 ;; A Microsoft misunderstanding.
258 ,@(unless (mm-coding-system-p 'ks_c_5601-1987)
259 (if (mm-coding-system-p 'cp949)
260 '((ks_c_5601-1987 . cp949))
261 '((ks_c_5601-1987 . euc-kr))))
262 ;; Windows-31J is Windows Codepage 932.
263 ,@(when (and (not (mm-coding-system-p 'windows-31j))
264 (mm-coding-system-p 'cp932))
265 '((windows-31j . cp932)))
266 ;; Charset name: GBK, Charset aliases: CP936, MS936, windows-936
267 ;; http://www.iana.org/assignments/charset-reg/GBK
268 ;; Emacs 22.1 has cp936, but not gbk, so we alias it:
269 ,@(when (and (not (mm-coding-system-p 'gbk))
270 (mm-coding-system-p 'cp936))
272 ;; ISO8859-1 is a bogus name for ISO-8859-1
273 ,@(when (and (not (mm-coding-system-p 'iso8859-1))
274 (mm-coding-system-p 'iso-8859-1))
275 '((iso8859-1 . iso-8859-1)))
277 "A mapping from unknown or invalid charset names to the real charset names.
279 See `mm-codepage-iso-8859-list' and `mm-codepage-ibm-list'.")
281 (defcustom mm-codepage-iso-8859-list
282 (list 1250 ;; Windows-1250 is a variant of Latin-2 heavily used by Microsoft
283 ;; Outlook users in Czech republic. Use this to allow reading of
284 ;; their e-mails. cp1250 should be defined by M-x codepage-setup
286 '(1252 . 1) ;; Windows-1252 is a superset of iso-8859-1 (West
287 ;; Europe). See also `gnus-article-dumbquotes-map'.
288 '(1254 . 9) ;; Windows-1254 is a superset of iso-8859-9 (Turkish).
289 '(1255 . 8));; Windows-1255 is a superset of iso-8859-8 (Hebrew).
290 "A list of Windows codepage numbers and iso-8859 charset numbers.
292 If an element is a number corresponding to a supported windows
293 codepage, appropriate entries to `mm-charset-synonym-alist' are
294 added by `mm-setup-codepage-iso-8859'. An element may also be a
295 cons cell where the car is a codepage number and the cdr is the
296 corresponding number of an iso-8859 charset."
297 :type '(list (set :inline t
298 (const 1250 :tag "Central and East European")
299 (const (1252 . 1) :tag "West European")
300 (const (1254 . 9) :tag "Turkish")
301 (const (1255 . 8) :tag "Hebrew"))
305 (integer :tag "Windows codepage number")
306 (cons (integer :tag "Windows codepage number")
307 (integer :tag "iso-8859 charset number")))))
308 :version "22.1" ;; Gnus 5.10.9
311 (defcustom mm-codepage-ibm-list
312 (list 437 ;; (US etc.)
316 863 ;; (Canadian French)
321 866 ;; (Cyrillic - Russian)
326 ;; In Emacs 23 (unicode), cp... and ibm... are aliases.
327 ;; Cf. http://thread.gmane.org/v9lkng5nwy.fsf@marauder.physik.uni-ulm.de
328 "List of IBM codepage numbers.
330 The codepage mappings slighly differ between IBM and other vendors.
331 See \"ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/IBM/README.TXT\".
333 If an element is a number corresponding to a supported windows
334 codepage, appropriate entries to `mm-charset-synonym-alist' are
335 added by `mm-setup-codepage-ibm'."
336 :type '(list (set :inline t
337 (const 437 :tag "US etc.")
338 (const 860 :tag "Portugal")
339 (const 861 :tag "Iceland")
340 (const 862 :tag "Israel")
341 (const 863 :tag "Canadian French")
342 (const 865 :tag "Nordic")
344 (const 850 :tag "Latin 1")
345 (const 855 :tag "Cyrillic")
346 (const 866 :tag "Cyrillic - Russian")
347 (const 857 :tag "Turkish")
348 (const 864 :tag "Arabic")
349 (const 869 :tag "Greek")
350 (const 874 :tag "Thai"))
353 (integer :tag "Codepage number")))
354 :version "22.1" ;; Gnus 5.10.9
357 (defun mm-setup-codepage-iso-8859 (&optional list)
358 "Add appropriate entries to `mm-charset-synonym-alist'.
359 Unless LIST is given, `mm-codepage-iso-8859-list' is used."
361 (setq list mm-codepage-iso-8859-list))
363 (let (cp windows iso)
365 (setq cp (intern (format "cp%d" (car i)))
366 windows (intern (format "windows-%d" (car i)))
367 iso (intern (format "iso-8859-%d" (cdr i))))
368 (setq cp (intern (format "cp%d" i))
369 windows (intern (format "windows-%d" i))))
370 (unless (mm-coding-system-p windows)
371 (if (mm-coding-system-p cp)
372 (add-to-list 'mm-charset-synonym-alist (cons windows cp))
373 (add-to-list 'mm-charset-synonym-alist (cons windows iso)))))))
375 (defun mm-setup-codepage-ibm (&optional list)
376 "Add appropriate entries to `mm-charset-synonym-alist'.
377 Unless LIST is given, `mm-codepage-ibm-list' is used."
379 (setq list mm-codepage-ibm-list))
380 (dolist (number list)
381 (let ((ibm (intern (format "ibm%d" number)))
382 (cp (intern (format "cp%d" number))))
383 (when (and (not (mm-coding-system-p ibm))
384 (mm-coding-system-p cp))
385 (add-to-list 'mm-charset-synonym-alist (cons ibm cp))))))
388 (mm-setup-codepage-iso-8859)
389 (mm-setup-codepage-ibm)
391 (defcustom mm-charset-override-alist
392 '((iso-8859-1 . windows-1252)
393 (iso-8859-8 . windows-1255)
394 (iso-8859-9 . windows-1254))
395 "A mapping from undesired charset names to their replacement.
397 You may add pairs like (iso-8859-1 . windows-1252) here,
398 i.e. treat iso-8859-1 as windows-1252. windows-1252 is a
399 superset of iso-8859-1."
406 (mapcar (lambda (pair)
407 (if (mm-charset-to-coding-system (cdr pair))
410 (iso-8859-1 . windows-1252)
411 (iso-8859-8 . windows-1255)
412 (iso-8859-9 . windows-1254)
413 (undecided . windows-1252)))))
414 (val (copy-sequence (default-value 'mm-charset-override-alist)))
417 (push (if (and (prog1
418 (setq pair (assq (caar val) defaults))
419 (setq defaults (delq pair defaults)))
420 (equal (car val) pair))
423 (const :format "(%v" ,(caar val))
424 (symbol :size 3 :format " . %v)\n" ,(cdar val))))
426 (setq val (cdr val)))
428 (push `(const ,(pop defaults)) rest))
431 `(set :inline t :format "%v" ,@(nreverse rest))
432 `(repeat :inline t :tag "Other options"
434 (symbol :size 3 :format "(%v")
435 (symbol :size 3 :format " . %v)\n")))))))
436 :version "22.1" ;; Gnus 5.10.9
439 (defcustom mm-charset-eval-alist
440 (if (featurep 'xemacs)
441 nil ;; I don't know what would be useful for XEmacs.
442 '(;; Emacs 21 offers 1250 1251 1253 1257. Emacs 22 provides autoloads for
443 ;; 1250-1258 (i.e. `mm-codepage-setup' does nothing).
444 (windows-1250 . (mm-codepage-setup 1250 t))
445 (windows-1251 . (mm-codepage-setup 1251 t))
446 (windows-1253 . (mm-codepage-setup 1253 t))
447 (windows-1257 . (mm-codepage-setup 1257 t))))
448 "An alist of (CHARSET . FORM) pairs.
449 If an article is encoded in an unknown CHARSET, FORM is
450 evaluated. This allows to load additional libraries providing
451 charsets on demand. If supported by your Emacs version, you
452 could use `autoload-coding-system' here."
453 :version "22.1" ;; Gnus 5.10.9
454 :type '(list (set :inline t
455 (const (windows-1250 . (mm-codepage-setup 1250 t)))
456 (const (windows-1251 . (mm-codepage-setup 1251 t)))
457 (const (windows-1253 . (mm-codepage-setup 1253 t)))
458 (const (windows-1257 . (mm-codepage-setup 1257 t)))
459 (const (cp850 . (mm-codepage-setup 850 nil))))
462 (cons (symbol :tag "charset")
463 (symbol :tag "form"))))
465 (put 'mm-charset-eval-alist 'risky-local-variable t)
467 (defvar mm-binary-coding-system
469 ((mm-coding-system-p 'binary) 'binary)
470 ((mm-coding-system-p 'no-conversion) 'no-conversion)
472 "100% binary coding system.")
474 (defvar mm-text-coding-system
475 (or (if (memq system-type '(windows-nt ms-dos ms-windows))
476 (and (mm-coding-system-p 'raw-text-dos) 'raw-text-dos)
477 (and (mm-coding-system-p 'raw-text) 'raw-text))
478 mm-binary-coding-system)
479 "Text-safe coding system (For removing ^M).")
481 (defvar mm-text-coding-system-for-write nil
482 "Text coding system for write.")
484 (defvar mm-auto-save-coding-system
486 ((mm-coding-system-p 'utf-8-emacs) ; Mule 7
487 (if (memq system-type '(windows-nt ms-dos ms-windows))
488 (if (mm-coding-system-p 'utf-8-emacs-dos)
489 'utf-8-emacs-dos mm-binary-coding-system)
491 ((mm-coding-system-p 'emacs-mule)
492 (if (memq system-type '(windows-nt ms-dos ms-windows))
493 (if (mm-coding-system-p 'emacs-mule-dos)
494 'emacs-mule-dos mm-binary-coding-system)
496 ((mm-coding-system-p 'escape-quoted) 'escape-quoted)
497 (t mm-binary-coding-system))
498 "Coding system of auto save file.")
500 (defvar mm-universal-coding-system mm-auto-save-coding-system
501 "The universal coding system.")
503 ;; Fixme: some of the cars here aren't valid MIME charsets. That
504 ;; should only matter with XEmacs, though.
505 (defvar mm-mime-mule-charset-alist
507 (iso-8859-1 latin-iso8859-1)
508 (iso-8859-2 latin-iso8859-2)
509 (iso-8859-3 latin-iso8859-3)
510 (iso-8859-4 latin-iso8859-4)
511 (iso-8859-5 cyrillic-iso8859-5)
512 ;; Non-mule (X)Emacs uses the last mule-charset for 8bit characters.
513 ;; The fake mule-charset, gnus-koi8-r, tells Gnus that the default
514 ;; charset is koi8-r, not iso-8859-5.
515 (koi8-r cyrillic-iso8859-5 gnus-koi8-r)
516 (iso-8859-6 arabic-iso8859-6)
517 (iso-8859-7 greek-iso8859-7)
518 (iso-8859-8 hebrew-iso8859-8)
519 (iso-8859-9 latin-iso8859-9)
520 (iso-8859-14 latin-iso8859-14)
521 (iso-8859-15 latin-iso8859-15)
522 (viscii vietnamese-viscii-lower)
523 (iso-2022-jp latin-jisx0201 japanese-jisx0208 japanese-jisx0208-1978)
524 (euc-kr korean-ksc5601)
525 (gb2312 chinese-gb2312)
527 (gb18030 gb18030-2-byte
528 gb18030-4-byte-bmp gb18030-4-byte-smp
529 gb18030-4-byte-ext-1 gb18030-4-byte-ext-2)
530 (big5 chinese-big5-1 chinese-big5-2)
532 (thai-tis620 thai-tis620)
533 (windows-1251 cyrillic-iso8859-5)
534 (iso-2022-7bit ethiopic arabic-1-column arabic-2-column)
535 (iso-2022-jp-2 latin-iso8859-1 greek-iso8859-7
536 latin-jisx0201 japanese-jisx0208-1978
537 chinese-gb2312 japanese-jisx0208
538 korean-ksc5601 japanese-jisx0212)
539 (iso-2022-int-1 latin-iso8859-1 greek-iso8859-7
540 latin-jisx0201 japanese-jisx0208-1978
541 chinese-gb2312 japanese-jisx0208
542 korean-ksc5601 japanese-jisx0212
543 chinese-cns11643-1 chinese-cns11643-2)
544 (iso-2022-int-1 latin-iso8859-1 latin-iso8859-2
545 cyrillic-iso8859-5 greek-iso8859-7
546 latin-jisx0201 japanese-jisx0208-1978
547 chinese-gb2312 japanese-jisx0208
548 korean-ksc5601 japanese-jisx0212
549 chinese-cns11643-1 chinese-cns11643-2
550 chinese-cns11643-3 chinese-cns11643-4
551 chinese-cns11643-5 chinese-cns11643-6
553 (iso-2022-jp-3 latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
554 japanese-jisx0213-1 japanese-jisx0213-2)
555 (shift_jis latin-jisx0201 katakana-jisx0201 japanese-jisx0208)
556 ,(cond ((fboundp 'unicode-precedence-list)
557 (cons 'utf-8 (delq 'ascii (mapcar 'charset-name
558 (unicode-precedence-list)))))
559 ((or (not (fboundp 'charsetp)) ;; non-Mule case
560 (charsetp 'unicode-a)
561 (not (mm-coding-system-p 'mule-utf-8)))
562 '(utf-8 unicode-a unicode-b unicode-c unicode-d unicode-e))
563 (t ;; If we have utf-8 we're in Mule 5+.
566 (coding-system-get 'mule-utf-8 'safe-charsets))))))
567 "Alist of MIME-charset/MULE-charsets.")
569 (defun mm-enrich-utf-8-by-mule-ucs ()
570 "Make the `utf-8' MIME charset usable by the Mule-UCS package.
571 This function will run when the `un-define' module is loaded under
572 XEmacs, and fill the `utf-8' entry in `mm-mime-mule-charset-alist'
573 with Mule charsets. It is completely useless for Emacs."
574 (when (boundp 'unicode-basic-translation-charset-order-list)
580 'unicode-basic-translation-charset-order-list))))
581 (elem (assq 'utf-8 mm-mime-mule-charset-alist)))
584 (setq mm-mime-mule-charset-alist
585 (nconc mm-mime-mule-charset-alist
586 (list (cons 'utf-8 val))))))
589 ;; Correct by construction, but should be unnecessary for Emacs:
590 (if (featurep 'xemacs)
591 (eval-after-load "un-define" '(mm-enrich-utf-8-by-mule-ucs))
592 (when (and (fboundp 'coding-system-list)
593 (fboundp 'sort-coding-systems))
594 (let ((css (sort-coding-systems (coding-system-list 'base-only)))
598 mime (or (coding-system-get cs :mime-charset); Emacs 23 (unicode)
599 (coding-system-get cs 'mime-charset)))
601 (not (eq t (setq mule
602 (coding-system-get cs 'safe-charsets))))
603 (not (assq mime alist)))
604 (push (cons mime (delq 'ascii mule)) alist)))
605 (setq mm-mime-mule-charset-alist (nreverse alist)))))
607 (defvar mm-hack-charsets '(iso-8859-15 iso-2022-jp-2)
608 "A list of special charsets.
609 Valid elements include:
610 `iso-8859-15' convert ISO-8859-1, -9 to ISO-8859-15 if ISO-8859-15 exists.
611 `iso-2022-jp-2' convert ISO-2022-jp to ISO-2022-jp-2 if ISO-2022-jp-2 exists."
614 (defvar mm-iso-8859-15-compatible
615 '((iso-8859-1 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE")
616 (iso-8859-9 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE\xD0\xDD\xDE\xF0\xFD\xFE"))
617 "ISO-8859-15 exchangeable coding systems and inconvertible characters.")
619 (defvar mm-iso-8859-x-to-15-table
620 (and (fboundp 'coding-system-p)
621 (mm-coding-system-p 'iso-8859-15)
624 (if (mm-coding-system-p (car cs))
625 (let ((c (string-to-char
626 (decode-coding-string "\341" (car cs)))))
627 (cons (char-charset c)
630 (decode-coding-string "\341" 'iso-8859-15)) c)
631 (string-to-list (decode-coding-string (car (cdr cs))
634 mm-iso-8859-15-compatible))
635 "A table of the difference character between ISO-8859-X and ISO-8859-15.")
637 (defcustom mm-coding-system-priorities
638 (if (boundp 'current-language-environment)
639 (let ((lang (symbol-value 'current-language-environment)))
640 (cond ((string= lang "Japanese")
641 ;; Japanese users prefer iso-2022-jp to euc-japan or
642 ;; shift_jis, however iso-8859-1 should be used when
643 ;; there are only ASCII text and Latin-1 characters.
644 '(iso-8859-1 iso-2022-jp iso-2022-jp-2 shift_jis utf-8)))))
645 "Preferred coding systems for encoding outgoing messages.
647 More than one suitable coding system may be found for some text.
648 By default, the coding system with the highest priority is used
649 to encode outgoing messages (see `sort-coding-systems'). If this
650 variable is set, it overrides the default priority."
652 :type '(repeat (symbol :tag "Coding system"))
656 (defvar mm-use-find-coding-systems-region
657 (fboundp 'find-coding-systems-region)
658 "Use `find-coding-systems-region' to find proper coding systems.
660 Setting it to nil is useful on Emacsen supporting Unicode if sending
661 mail with multiple parts is preferred to sending a Unicode one.")
663 ;;; Internal variables:
667 (defun mm-mule-charset-to-mime-charset (charset)
668 "Return the MIME charset corresponding to the given Mule CHARSET."
669 (if (and (fboundp 'find-coding-systems-for-charsets)
670 (fboundp 'sort-coding-systems))
671 (let ((css (sort (sort-coding-systems
672 (find-coding-systems-for-charsets (list charset)))
673 'mm-sort-coding-systems-predicate))
675 (while (and (not mime)
677 (when (setq cs (pop css))
678 (setq mime (or (coding-system-get cs :mime-charset)
679 (coding-system-get cs 'mime-charset)))))
681 (let ((alist (mapcar (lambda (cs)
682 (assq cs mm-mime-mule-charset-alist))
683 (sort (mapcar 'car mm-mime-mule-charset-alist)
684 'mm-sort-coding-systems-predicate)))
687 (when (memq charset (cdar alist))
688 (setq out (caar alist)
693 (defun mm-charset-to-coding-system (charset &optional lbt
695 "Return coding-system corresponding to CHARSET.
696 CHARSET is a symbol naming a MIME charset.
697 If optional argument LBT (`unix', `dos' or `mac') is specified, it is
698 used as the line break code type of the coding system.
700 If ALLOW-OVERRIDE is given, use `mm-charset-override-alist' to
701 map undesired charset names to their replacement. This should
702 only be used for decoding, not for encoding."
703 ;; OVERRIDE is used (only) in `mm-decode-body' and `mm-decode-string'.
704 (when (stringp charset)
705 (setq charset (intern (downcase charset))))
707 (setq charset (intern (format "%s-%s" charset lbt))))
711 ;; Running in a non-MULE environment.
712 ((or (null (mm-get-coding-system-list))
713 (not (fboundp 'coding-system-get)))
715 ;; Check override list quite early. Should only used for decoding, not for
718 (let ((cs (cdr (assq charset mm-charset-override-alist))))
719 (and cs (mm-coding-system-p cs) cs))))
721 ((eq charset 'us-ascii)
723 ;; Check to see whether we can handle this charset. (This depends
724 ;; on there being some coding system matching each `mime-charset'
725 ;; property defined, as there should be.)
726 ((and (mm-coding-system-p charset)
727 ;;; Doing this would potentially weed out incorrect charsets.
729 ;;; (eq charset (coding-system-get charset 'mime-charset))
732 ;; Eval expressions from `mm-charset-eval-alist'
733 ((let* ((el (assq charset mm-charset-eval-alist))
740 (condition-case nil (eval form) (error nil))
741 ;; (message "Failed to eval `%s'" form))
742 (mm-coding-system-p cs)
743 (message "Added charset `%s' via `mm-charset-eval-alist'" cs))
745 ;; Translate invalid charsets.
746 ((let ((cs (cdr (assq charset mm-charset-synonym-alist))))
748 (mm-coding-system-p cs)
750 ;; "Using synonym `%s' from `mm-charset-synonym-alist' for `%s'"
753 ;; Last resort: search the coding system list for entries which
754 ;; have the right mime-charset in case the canonical name isn't
755 ;; defined (though it should be).
757 ;; mm-get-coding-system-list returns a list of cs without lbt.
759 (dolist (c (mm-get-coding-system-list))
761 (eq charset (or (coding-system-get c :mime-charset)
762 (coding-system-get c 'mime-charset))))
765 ;; Warn the user about unknown charset:
766 (if (fboundp 'gnus-message)
767 (gnus-message 7 "Unknown charset: %s" charset)
768 (message "Unknown charset: %s" charset)))
772 (defvar mm-emacs-mule (and (not (featurep 'xemacs))
773 (boundp 'default-enable-multibyte-characters)
774 default-enable-multibyte-characters
775 (fboundp 'set-buffer-multibyte))
776 "True in Emacs with Mule.")
779 (defun mm-enable-multibyte ()
780 "Set the multibyte flag of the current buffer.
781 Only do this if the default value of `enable-multibyte-characters' is
782 non-nil. This is a no-op in XEmacs."
783 (set-buffer-multibyte 'to))
784 (defalias 'mm-enable-multibyte 'ignore))
787 (defun mm-disable-multibyte ()
788 "Unset the multibyte flag of in the current buffer.
789 This is a no-op in XEmacs."
790 (set-buffer-multibyte nil))
791 (defalias 'mm-disable-multibyte 'ignore)))
793 (defun mm-preferred-coding-system (charset)
794 ;; A typo in some Emacs versions.
795 (or (get-charset-property charset 'preferred-coding-system)
796 (get-charset-property charset 'prefered-coding-system)))
798 ;; Mule charsets shouldn't be used.
799 (defsubst mm-guess-charset ()
800 "Guess Mule charset from the language environment."
802 mail-parse-mule-charset ;; cached mule-charset
804 (setq mail-parse-mule-charset
805 (and (boundp 'current-language-environment)
808 (assoc current-language-environment
809 language-info-alist))))))
810 (if (or (not mail-parse-mule-charset)
811 (eq mail-parse-mule-charset 'ascii))
812 (setq mail-parse-mule-charset
813 (or (car (last (assq mail-parse-charset
814 mm-mime-mule-charset-alist)))
817 mail-parse-mule-charset)))
819 (defun mm-charset-after (&optional pos)
820 "Return charset of a character in current buffer at position POS.
821 If POS is nil, it defauls to the current point.
822 If POS is out of range, the value is nil.
823 If the charset is `composition', return the actual one."
824 (let ((char (char-after pos)) charset)
825 (if (< (mm-char-int char) 128)
826 (setq charset 'ascii)
827 ;; charset-after is fake in some Emacsen.
828 (setq charset (and (fboundp 'char-charset) (char-charset char)))
829 (if (eq charset 'composition) ; Mule 4
830 (let ((p (or pos (point))))
831 (cadr (find-charset-region p (1+ p))))
832 (if (and charset (not (memq charset '(ascii eight-bit-control
833 eight-bit-graphic))))
835 (mm-guess-charset))))))
837 (defun mm-mime-charset (charset)
838 "Return the MIME charset corresponding to the given Mule CHARSET."
839 (if (eq charset 'unknown)
840 (error "The message contains non-printable characters, please use attachment"))
841 (if (and (fboundp 'coding-system-get) (fboundp 'get-charset-property))
842 ;; This exists in Emacs 20.
844 (and (mm-preferred-coding-system charset)
845 (or (coding-system-get
846 (mm-preferred-coding-system charset) :mime-charset)
848 (mm-preferred-coding-system charset) 'mime-charset)))
849 (and (eq charset 'ascii)
851 (mm-preferred-coding-system charset)
852 (mm-mule-charset-to-mime-charset charset))
853 ;; This is for XEmacs.
854 (mm-mule-charset-to-mime-charset charset)))
856 (if (fboundp 'delete-dups)
857 (defalias 'mm-delete-duplicates 'delete-dups)
858 (defun mm-delete-duplicates (list)
859 "Destructively remove `equal' duplicates from LIST.
860 Store the result in LIST and return it. LIST must be a proper list.
861 Of several `equal' occurrences of an element in LIST, the first
864 This is a compatibility function for Emacsen without `delete-dups'."
865 ;; Code from `subr.el' in Emacs 22:
868 (setcdr tail (delete (car tail) (cdr tail)))
869 (setq tail (cdr tail))))
872 ;; Fixme: This is used in places when it should be testing the
873 ;; default multibyteness. See mm-default-multibyte-p.
875 (if (and (not (featurep 'xemacs))
876 (boundp 'enable-multibyte-characters))
877 (defun mm-multibyte-p ()
878 "Non-nil if multibyte is enabled in the current buffer."
879 enable-multibyte-characters)
880 (defun mm-multibyte-p () (featurep 'mule))))
882 (defun mm-default-multibyte-p ()
883 "Return non-nil if the session is multibyte.
884 This affects whether coding conversion should be attempted generally."
886 (if (boundp 'default-enable-multibyte-characters)
887 default-enable-multibyte-characters
890 (defun mm-iso-8859-x-to-15-region (&optional b e)
891 (if (fboundp 'char-charset)
892 (let (charset item c inconvertible)
894 (if e (narrow-to-region b e))
895 (goto-char (point-min))
896 (skip-chars-forward "\0-\177")
899 ((not (setq item (assq (char-charset (setq c (char-after)))
900 mm-iso-8859-x-to-15-table)))
902 ((memq c (cdr (cdr item)))
903 (setq inconvertible t)
906 (insert-before-markers (prog1 (+ c (car (cdr item)))
908 (skip-chars-forward "\0-\177")))
909 (not inconvertible))))
911 (defun mm-sort-coding-systems-predicate (a b)
914 ;; Note: invalid entries are dropped silently
915 (and (setq cs (mm-coding-system-p cs))
916 (coding-system-base cs)))
917 mm-coding-system-priorities)))
918 (and (setq a (mm-coding-system-p a))
919 (if (setq b (mm-coding-system-p b))
920 (> (length (memq (coding-system-base a) priorities))
921 (length (memq (coding-system-base b) priorities)))
925 (autoload 'latin-unity-massage-name "latin-unity")
926 (autoload 'latin-unity-maybe-remap "latin-unity")
927 (autoload 'latin-unity-representations-feasible-region "latin-unity")
928 (autoload 'latin-unity-representations-present-region "latin-unity"))
930 (defvar latin-unity-coding-systems)
931 (defvar latin-unity-ucs-list)
933 (defun mm-xemacs-find-mime-charset-1 (begin end)
934 "Determine which MIME charset to use to send region as message.
935 This uses the XEmacs-specific latin-unity package to better handle the
936 case where identical characters from diverse ISO-8859-? character sets
937 can be encoded using a single one of the corresponding coding systems.
939 It treats `mm-coding-system-priorities' as the list of preferred
940 coding systems; a useful example setting for this list in Western
941 Europe would be '(iso-8859-1 iso-8859-15 utf-8), which would default
942 to the very standard Latin 1 coding system, and only move to coding
943 systems that are less supported as is necessary to encode the
944 characters that exist in the buffer.
946 Latin Unity doesn't know about those non-ASCII Roman characters that
947 are available in various East Asian character sets. As such, its
948 behavior if you have a JIS 0212 LATIN SMALL LETTER A WITH ACUTE in a
949 buffer and it can otherwise be encoded as Latin 1, won't be ideal.
950 But this is very much a corner case, so don't worry about it."
951 (let ((systems mm-coding-system-priorities) csets psets curset)
953 ;; Load the Latin Unity library, if available.
954 (when (and (not (featurep 'latin-unity)) (locate-library "latin-unity"))
955 (require 'latin-unity))
957 ;; Now, can we use it?
958 (if (featurep 'latin-unity)
960 (setq csets (latin-unity-representations-feasible-region begin end)
961 psets (latin-unity-representations-present-region begin end))
965 ;; Pass back the first coding system in the preferred list
966 ;; that can encode the whole region.
967 (dolist (curset systems)
968 (setq curset (latin-unity-massage-name 'buffer-default curset))
970 ;; If the coding system is a universal coding system, then
971 ;; it can certainly encode all the characters in the region.
972 (if (memq curset latin-unity-ucs-list)
973 (throw 'done (list curset)))
975 ;; If a coding system isn't universal, and isn't in
976 ;; the list that latin unity knows about, we can't
977 ;; decide whether to use it here. Leave that until later
978 ;; in `mm-find-mime-charset-region' function, whence we
980 (unless (memq curset latin-unity-coding-systems)
983 ;; Right, we know about this coding system, and it may
984 ;; conceivably be able to encode all the characters in
986 (if (latin-unity-maybe-remap begin end curset csets psets t)
987 (throw 'done (list curset))))
989 ;; Can't encode using anything from the
990 ;; `mm-coding-system-priorities' list.
991 ;; Leave `mm-find-mime-charset' to do most of the work.
994 ;; Right, latin unity isn't available; let `mm-find-charset-region'
995 ;; take its default action, which equally applies to GNU Emacs.
998 (defmacro mm-xemacs-find-mime-charset (begin end)
999 (when (featurep 'xemacs)
1000 `(and (featurep 'mule) (mm-xemacs-find-mime-charset-1 ,begin ,end))))
1002 (declare-function mm-delete-duplicates "mm-util" (list))
1004 (defun mm-find-mime-charset-region (b e &optional hack-charsets)
1005 "Return the MIME charsets needed to encode the region between B and E.
1006 nil means ASCII, a single-element list represents an appropriate MIME
1007 charset, and a longer list means no appropriate charset."
1009 ;; The return possibilities of this function are a mess...
1010 (or (and (mm-multibyte-p)
1011 mm-use-find-coding-systems-region
1012 ;; Find the mime-charset of the most preferred coding
1013 ;; system that has one.
1014 (let ((systems (find-coding-systems-region b e)))
1015 (when mm-coding-system-priorities
1017 (sort systems 'mm-sort-coding-systems-predicate)))
1018 (setq systems (delq 'compound-text systems))
1019 (unless (equal systems '(undecided))
1021 (let* ((head (pop systems))
1022 (cs (or (coding-system-get head :mime-charset)
1023 (coding-system-get head 'mime-charset))))
1024 ;; The mime-charset (`x-ctext') of
1025 ;; `compound-text' is not in the IANA list. We
1026 ;; shouldn't normally use anything here with a
1027 ;; mime-charset having an `x-' prefix.
1028 ;; Fixme: Allow this to be overridden, since
1029 ;; there is existing use of x-ctext.
1030 ;; Also people apparently need the coding system
1031 ;; `iso-2022-jp-3' (which Mule-UCS defines with
1032 ;; mime-charset, though it's not valid).
1034 (not (string-match "^[Xx]-" (symbol-name cs)))
1035 ;; UTF-16 of any variety is invalid for
1036 ;; text parts and, unfortunately, has
1037 ;; mime-charset defined both in Mule-UCS
1038 ;; and versions of Emacs. (The name
1039 ;; might be `mule-utf-16...' or
1041 (not (string-match "utf-16" (symbol-name cs))))
1043 charsets (list cs))))))
1045 ;; If we're XEmacs, and some coding system is appropriate,
1046 ;; mm-xemacs-find-mime-charset will return an appropriate list.
1047 ;; Otherwise, we'll get nil, and the next setq will get invoked.
1048 (setq charsets (mm-xemacs-find-mime-charset b e))
1050 ;; Fixme: won't work for unibyte Emacs 23:
1052 ;; We're not multibyte, or a single coding system won't cover it.
1054 (mm-delete-duplicates
1055 (mapcar 'mm-mime-charset
1057 (mm-find-charset-region b e))))))
1058 (if (and (> (length charsets) 1)
1059 (memq 'iso-8859-15 charsets)
1060 (memq 'iso-8859-15 hack-charsets)
1061 (save-excursion (mm-iso-8859-x-to-15-region b e)))
1062 (dolist (x mm-iso-8859-15-compatible)
1063 (setq charsets (delq (car x) charsets))))
1064 (if (and (memq 'iso-2022-jp-2 charsets)
1065 (memq 'iso-2022-jp-2 hack-charsets))
1066 (setq charsets (delq 'iso-2022-jp charsets)))
1067 ;; Attempt to reduce the number of charsets if utf-8 is available.
1068 (if (and (featurep 'xemacs)
1069 (> (length charsets) 1)
1070 (mm-coding-system-p 'utf-8))
1071 (let ((mm-coding-system-priorities
1072 (cons 'utf-8 mm-coding-system-priorities)))
1074 (mm-delete-duplicates
1075 (mapcar 'mm-mime-charset
1077 (mm-find-charset-region b e)))))))
1080 (defmacro mm-with-unibyte-buffer (&rest forms)
1081 "Create a temporary buffer, and evaluate FORMS there like `progn'.
1082 Use unibyte mode for this."
1084 (mm-disable-multibyte)
1086 (put 'mm-with-unibyte-buffer 'lisp-indent-function 0)
1087 (put 'mm-with-unibyte-buffer 'edebug-form-spec '(body))
1089 (defmacro mm-with-multibyte-buffer (&rest forms)
1090 "Create a temporary buffer, and evaluate FORMS there like `progn'.
1091 Use multibyte mode for this."
1093 (mm-enable-multibyte)
1095 (put 'mm-with-multibyte-buffer 'lisp-indent-function 0)
1096 (put 'mm-with-multibyte-buffer 'edebug-form-spec '(body))
1098 (defmacro mm-with-unibyte-current-buffer (&rest forms)
1099 "Evaluate FORMS with current buffer temporarily made unibyte.
1100 Also bind `default-enable-multibyte-characters' to nil.
1101 Equivalent to `progn' in XEmacs
1103 NOTE: Use this macro with caution in multibyte buffers (it is not
1104 worth using this macro in unibyte buffers of course). Use of
1105 `(set-buffer-multibyte t)', which is run finally, is generally
1106 harmful since it is likely to modify existing data in the buffer.
1107 For instance, it converts \"\\300\\255\" into \"\\255\" in
1108 Emacs 23 (unicode)."
1109 (let ((multibyte (make-symbol "multibyte"))
1110 (buffer (make-symbol "buffer")))
1112 (let ((,multibyte enable-multibyte-characters)
1113 (,buffer (current-buffer)))
1115 (let (default-enable-multibyte-characters)
1116 (set-buffer-multibyte nil)
1118 (set-buffer ,buffer)
1119 (set-buffer-multibyte ,multibyte)))
1120 (let (default-enable-multibyte-characters)
1122 (put 'mm-with-unibyte-current-buffer 'lisp-indent-function 0)
1123 (put 'mm-with-unibyte-current-buffer 'edebug-form-spec '(body))
1125 (defun mm-find-charset-region (b e)
1126 "Return a list of Emacs charsets in the region B to E."
1128 ((and (mm-multibyte-p)
1129 (fboundp 'find-charset-region))
1130 ;; Remove composition since the base charsets have been included.
1131 ;; Remove eight-bit-*, treat them as ascii.
1132 (let ((css (find-charset-region b e)))
1134 '(composition eight-bit-control eight-bit-graphic control-1)
1136 (setq css (delq cs css)))))
1138 ;; We are in a unibyte buffer or XEmacs non-mule, so we futz around a bit.
1141 (narrow-to-region b e)
1142 (goto-char (point-min))
1143 (skip-chars-forward "\0-\177")
1148 (and (boundp 'current-language-environment)
1149 (car (last (assq 'charset
1150 (assoc current-language-environment
1151 language-info-alist))))))
1152 (if (eq charset 'ascii) (setq charset nil))
1155 (car (last (assq mail-parse-charset
1156 mm-mime-mule-charset-alist)))))
1157 (list 'ascii (or charset 'latin-iso8859-1)))))))))
1159 (defun mm-auto-mode-alist ()
1160 "Return an `auto-mode-alist' with only the .gz (etc) thingies."
1161 (let ((alist auto-mode-alist)
1164 (when (listp (cdar alist))
1165 (push (car alist) out))
1169 (defvar mm-inhibit-file-name-handlers
1170 '(jka-compr-handler image-file-handler epa-file-handler)
1171 "A list of handlers doing (un)compression (etc) thingies.")
1173 (defun mm-insert-file-contents (filename &optional visit beg end replace
1175 "Like `insert-file-contents', but only reads in the file.
1176 A buffer may be modified in several ways after reading into the buffer due
1177 to advanced Emacs features, such as file-name-handlers, format decoding,
1178 `find-file-hooks', etc.
1179 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'.
1180 This function ensures that none of these modifications will take place."
1181 (let* ((format-alist nil)
1182 (auto-mode-alist (if inhibit nil (mm-auto-mode-alist)))
1183 (default-major-mode 'fundamental-mode)
1184 (enable-local-variables nil)
1185 (after-insert-file-functions nil)
1186 (enable-local-eval nil)
1187 (inhibit-file-name-operation (if inhibit
1188 'insert-file-contents
1189 inhibit-file-name-operation))
1190 (inhibit-file-name-handlers
1192 (append mm-inhibit-file-name-handlers
1193 inhibit-file-name-handlers)
1194 inhibit-file-name-handlers))
1195 (ffh (if (boundp 'find-file-hook)
1198 (val (symbol-value ffh)))
1201 (insert-file-contents filename visit beg end replace)
1204 (defun mm-append-to-file (start end filename &optional codesys inhibit)
1205 "Append the contents of the region to the end of file FILENAME.
1206 When called from a function, expects three arguments,
1207 START, END and FILENAME. START and END are buffer positions
1208 saying what text to write.
1209 Optional fourth argument specifies the coding system to use when
1211 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1212 (let ((coding-system-for-write
1213 (or codesys mm-text-coding-system-for-write
1214 mm-text-coding-system))
1215 (inhibit-file-name-operation (if inhibit
1217 inhibit-file-name-operation))
1218 (inhibit-file-name-handlers
1220 (append mm-inhibit-file-name-handlers
1221 inhibit-file-name-handlers)
1222 inhibit-file-name-handlers)))
1223 (write-region start end filename t 'no-message)
1224 (message "Appended to %s" filename)))
1226 (defun mm-write-region (start end filename &optional append visit lockname
1227 coding-system inhibit)
1229 "Like `write-region'.
1230 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1231 (let ((coding-system-for-write
1232 (or coding-system mm-text-coding-system-for-write
1233 mm-text-coding-system))
1234 (inhibit-file-name-operation (if inhibit
1236 inhibit-file-name-operation))
1237 (inhibit-file-name-handlers
1239 (append mm-inhibit-file-name-handlers
1240 inhibit-file-name-handlers)
1241 inhibit-file-name-handlers)))
1242 (write-region start end filename append visit lockname)))
1244 (autoload 'gmm-write-region "gmm-utils")
1246 ;; It is not a MIME function, but some MIME functions use it.
1247 (if (and (fboundp 'make-temp-file)
1249 (let ((def (symbol-function 'make-temp-file)))
1250 (and (byte-code-function-p def)
1251 (setq def (if (fboundp 'compiled-function-arglist)
1253 (eval (list 'compiled-function-arglist def))
1256 (eq (nth 3 def) 'suffix)))))
1257 (defalias 'mm-make-temp-file 'make-temp-file)
1258 ;; Stolen (and modified for XEmacs) from Emacs 22.
1259 (defun mm-make-temp-file (prefix &optional dir-flag suffix)
1260 "Create a temporary file.
1261 The returned file name (created by appending some random characters at the end
1262 of PREFIX, and expanding against `temporary-file-directory' if necessary),
1263 is guaranteed to point to a newly created empty file.
1264 You can then use `write-region' to write new data into the file.
1266 If DIR-FLAG is non-nil, create a new empty directory instead of a file.
1268 If SUFFIX is non-nil, add that at the end of the file name."
1269 (let ((umask (default-file-modes))
1273 ;; Create temp files with strict access rights. It's easy to
1274 ;; loosen them later, whereas it's impossible to close the
1275 ;; time-window of loose permissions otherwise.
1276 (set-default-file-modes 448)
1277 (while (condition-case err
1283 (if (fboundp 'temp-directory)
1286 temporary-file-directory))))
1288 (setq file (concat file suffix)))
1290 (make-directory file)
1291 ;; NOTE: This is unsafe if Emacs 20
1292 ;; users and XEmacs users don't use
1293 ;; a secure temp directory.
1294 (gmm-write-region "" nil file nil 'silent
1297 (file-already-exists t)
1298 ;; The XEmacs version of `make-directory' issues
1300 (file-error (or (and (featurep 'xemacs)
1301 (file-exists-p file))
1302 (signal (car err) (cdr err)))))
1303 ;; the file was somehow created by someone else between
1304 ;; `make-temp-name' and `write-region', let's try again.
1308 (set-default-file-modes umask)))))
1310 (defun mm-image-load-path (&optional package)
1312 (dolist (path load-path (nreverse result))
1315 (setq dir (concat (file-name-directory
1316 (directory-file-name path))
1317 "etc/images/" (or package "gnus/")))))
1319 (push path result))))
1321 ;; Fixme: This doesn't look useful where it's used.
1322 (if (fboundp 'detect-coding-region)
1323 (defun mm-detect-coding-region (start end)
1324 "Like `detect-coding-region' except returning the best one."
1325 (let ((coding-systems
1326 (detect-coding-region start end)))
1327 (or (car-safe coding-systems)
1329 (defun mm-detect-coding-region (start end)
1330 (let ((point (point)))
1332 (skip-chars-forward "\0-\177" end)
1334 (if (eq (point) end) 'ascii (mm-guess-charset))
1335 (goto-char point)))))
1337 (declare-function mm-detect-coding-region "mm-util" (start end))
1339 (if (fboundp 'coding-system-get)
1340 (defun mm-detect-mime-charset-region (start end)
1341 "Detect MIME charset of the text in the region between START and END."
1342 (let ((cs (mm-detect-coding-region start end)))
1343 (or (coding-system-get cs :mime-charset)
1344 (coding-system-get cs 'mime-charset))))
1345 (defun mm-detect-mime-charset-region (start end)
1346 "Detect MIME charset of the text in the region between START and END."
1347 (let ((cs (mm-detect-coding-region start end)))
1351 (unless (fboundp 'coding-system-to-mime-charset)
1352 (defalias 'coding-system-to-mime-charset 'ignore)))
1354 (defun mm-coding-system-to-mime-charset (coding-system)
1355 "Return the MIME charset corresponding to CODING-SYSTEM.
1356 To make this function work with XEmacs, the APEL package is required."
1358 (or (and (fboundp 'coding-system-get)
1359 (or (coding-system-get coding-system :mime-charset)
1360 (coding-system-get coding-system 'mime-charset)))
1361 (and (featurep 'xemacs)
1362 (or (and (fboundp 'coding-system-to-mime-charset)
1363 (not (eq (symbol-function 'coding-system-to-mime-charset)
1365 (and (condition-case nil
1368 (fboundp 'coding-system-to-mime-charset)))
1369 (coding-system-to-mime-charset coding-system)))))
1372 (require 'jka-compr))
1374 (defun mm-decompress-buffer (filename &optional inplace force)
1375 "Decompress buffer's contents, depending on jka-compr.
1376 Only when FORCE is t or `auto-compression-mode' is enabled and FILENAME
1377 agrees with `jka-compr-compression-info-list', decompression is done.
1378 Signal an error if FORCE is neither nil nor t and compressed data are
1379 not decompressed because `auto-compression-mode' is disabled.
1380 If INPLACE is nil, return decompressed data or nil without modifying
1381 the buffer. Otherwise, replace the buffer's contents with the
1382 decompressed data. The buffer's multibyteness must be turned off."
1385 (prog1 t (require 'jka-compr))
1386 (and (fboundp 'jka-compr-installed-p)
1387 (jka-compr-installed-p))))
1388 (let ((info (jka-compr-get-compression-info filename)))
1390 (unless (or (memq force (list nil t))
1391 (jka-compr-installed-p))
1393 (let ((prog (jka-compr-info-uncompress-program info))
1394 (args (jka-compr-info-uncompress-args info))
1395 (msg (format "%s %s..."
1396 (jka-compr-info-uncompress-message info)
1398 (err-file (jka-compr-make-temp-name))
1399 (cur (current-buffer))
1400 (coding-system-for-read mm-binary-coding-system)
1401 (coding-system-for-write mm-binary-coding-system)
1404 (mm-with-unibyte-buffer
1405 (insert-buffer-substring cur)
1408 (unless (memq (apply 'call-process-region
1409 (point-min) (point-max)
1410 prog t (list t err-file) nil args)
1411 jka-compr-acceptable-retval-list)
1415 (delete "" (split-string
1417 (insert-file-contents err-file)
1423 (format "Error while executing \"%s %s < %s\""
1424 prog (mapconcat 'identity args " ")
1426 (setq retval (buffer-string)))
1428 (setq err-msg (error-message-string err)))))
1429 (when (file-exists-p err-file)
1430 (ignore-errors (jka-compr-delete-temp-file err-file)))
1433 (delete-region (point-min) (point-max))
1436 (message "%s" (or err-msg (concat msg "done")))
1440 (unless (fboundp 'coding-system-name)
1441 (defalias 'coding-system-name 'ignore))
1442 (unless (fboundp 'find-file-coding-system-for-read-from-filename)
1443 (defalias 'find-file-coding-system-for-read-from-filename 'ignore))
1444 (unless (fboundp 'find-operation-coding-system)
1445 (defalias 'find-operation-coding-system 'ignore)))
1447 (defun mm-find-buffer-file-coding-system (&optional filename)
1448 "Find coding system used to decode the contents of the current buffer.
1449 This function looks for the coding system magic cookie or examines the
1450 coding system specified by `file-coding-system-alist' being associated
1451 with FILENAME which defaults to `buffer-file-name'. Data compressed by
1452 gzip, bzip2, etc. are allowed."
1454 (setq filename buffer-file-name))
1456 (let ((decomp (unless ;; No worth to examine charset of tar files.
1459 "\\.\\(?:tar\\.[^.]+\\|tbz\\|tgz\\)\\'"
1461 (mm-decompress-buffer filename nil t))))
1463 (set-buffer (let (default-enable-multibyte-characters)
1464 (generate-new-buffer " *temp*")))
1466 (setq filename (file-name-sans-extension filename)))
1467 (goto-char (point-min))
1470 ((boundp 'set-auto-coding-function) ;; Emacs
1472 (or (funcall (symbol-value 'set-auto-coding-function)
1473 filename (- (point-max) (point-min)))
1474 (car (find-operation-coding-system 'insert-file-contents
1476 (let (auto-coding-alist)
1478 (funcall (symbol-value 'set-auto-coding-function)
1479 nil (- (point-max) (point-min)))
1481 ((and (featurep 'xemacs) (featurep 'file-coding)) ;; XEmacs
1482 (let ((case-fold-search t)
1483 (end (point-at-eol))
1486 (and (re-search-forward "-\\*-+[\t ]*" end t)
1488 (setq start (match-end 0))
1489 (re-search-forward "[\t ]*-+\\*-" end t))
1491 (setq end (match-beginning 0))
1493 (or (looking-at "coding:[\t ]*\\([^\t ;]+\\)")
1495 "[\t ;]+coding:[\t ]*\\([^\t ;]+\\)"
1497 (find-coding-system (setq codesys
1498 (intern (match-string 1))))
1500 (and (re-search-forward "^[\t ]*;+[\t ]*Local[\t ]+Variables:"
1503 (setq start (match-end 0))
1504 (re-search-forward "^[\t ]*;+[\t ]*End:" nil t))
1506 (setq end (match-beginning 0))
1509 "^[\t ]*;+[\t ]*coding:[\t ]*\\([^\t\n\r ]+\\)"
1511 (find-coding-system (setq codesys
1512 (intern (match-string 1))))
1515 (goto-char (point-min))
1516 (setq case-fold-search nil)
1517 (re-search-forward "^;;;coding system: "
1518 ;;(+ (point-min) 3000) t))
1520 (looking-at "[^\t\n\r ]+")
1522 (setq codesys (intern (match-string 0))))
1526 (find-file-coding-system-for-read-from-filename
1528 (coding-system-name (coding-system-base codesys)))))))
1530 (kill-buffer (current-buffer)))))))
1534 ;; arch-tag: 94dc5388-825d-4fd1-bfa5-2100aa351238
1535 ;;; mm-util.el ends here