1 ;;; mm-util.el --- Utility functions for Mule and low level things
3 ;; Copyright (C) 1998-2014 Free Software Foundation, Inc.
5 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
6 ;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
7 ;; This file is part of GNU Emacs.
9 ;; GNU Emacs is free software: you can redistribute it and/or modify
10 ;; it under the terms of the GNU General Public License as published by
11 ;; the Free Software Foundation, either version 3 of the License, or
12 ;; (at your option) any later version.
14 ;; GNU Emacs is distributed in the hope that it will be useful,
15 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
16 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 ;; GNU General Public License for more details.
19 ;; You should have received a copy of the GNU General Public License
20 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
26 ;; For Emacs <22.2 and XEmacs.
28 (unless (fboundp 'declare-function) (defmacro declare-function (&rest r))))
30 (eval-when-compile (require 'cl))
34 (if (featurep 'xemacs)
35 (unless (ignore-errors
36 (require 'timer-funcs))
40 (defvar mm-mime-mule-charset-alist )
41 ;; Note this is not presently used on Emacs >= 23, which is good,
42 ;; since it means standalone message-mode (which requires mml and
43 ;; hence mml-util) does not load gnus-util.
44 (autoload 'gnus-completing-read "gnus-util")
46 ;; Emulate functions that are not available in every (X)Emacs version.
47 ;; The name of a function is prefixed with mm-, like `mm-char-int' for
48 ;; `char-int' that is a native XEmacs function, not available in Emacs.
49 ;; Gnus programs all should use mm- functions, not the original ones.
53 (let ((nfunc (intern (format "mm-%s" (car elem)))))
54 (if (fboundp (car elem))
55 (defalias nfunc (car elem))
56 (defalias nfunc (cdr elem)))))
57 `(;; `coding-system-list' is not available in XEmacs 21.4 built
58 ;; without the `file-coding' feature.
59 (coding-system-list . ignore)
60 ;; `char-int' is an XEmacs function, not available in Emacs.
62 ;; `coding-system-equal' is an Emacs function, not available in XEmacs.
63 (coding-system-equal . equal)
64 ;; `annotationp' is an XEmacs function, not available in Emacs.
65 (annotationp . ignore)
66 ;; `set-buffer-file-coding-system' is not available in XEmacs 21.4
67 ;; built without the `file-coding' feature.
68 (set-buffer-file-coding-system . ignore)
69 ;; `read-charset' is an Emacs function, not available in XEmacs.
76 (mapcar (lambda (e) (symbol-name (car e)))
77 mm-mime-mule-charset-alist)
79 ;; `subst-char-in-string' is not available in XEmacs 21.4.
81 . ,(lambda (from to string &optional inplace)
82 ;; stolen (and renamed) from nnheader.el
83 "Replace characters in STRING from FROM to TO.
84 Unless optional argument INPLACE is non-nil, return a new string."
85 (let ((string (if inplace string (copy-sequence string)))
88 ;; Replace all occurrences of FROM with TO.
90 (when (= (aref string idx) from)
94 ;; `replace-in-string' is an XEmacs function, not available in Emacs.
96 . ,(lambda (string regexp rep &optional literal)
97 "See `replace-regexp-in-string', only the order of args differs."
98 (replace-regexp-in-string regexp rep string nil literal)))
99 ;; `string-as-unibyte' is an Emacs function, not available in XEmacs.
100 (string-as-unibyte . identity)
101 ;; `string-make-unibyte' is an Emacs function, not available in XEmacs.
102 (string-make-unibyte . identity)
103 ;; string-as-multibyte often doesn't really do what you think it does.
105 ;; (aref (string-as-multibyte "\201") 0) -> 129 (aka ?\201)
106 ;; (aref (string-as-multibyte "\300") 0) -> 192 (aka ?\300)
107 ;; (aref (string-as-multibyte "\300\201") 0) -> 192 (aka ?\300)
108 ;; (aref (string-as-multibyte "\300\201") 1) -> 129 (aka ?\201)
110 ;; (aref (string-as-multibyte "\201\300") 0) -> 2240
111 ;; (aref (string-as-multibyte "\201\300") 1) -> <error>
112 ;; Better use string-to-multibyte or encode-coding-string.
113 ;; If you really need string-as-multibyte somewhere it's usually
114 ;; because you're using the internal emacs-mule representation (maybe
115 ;; because you're using string-as-unibyte somewhere), which is
116 ;; generally a problem in itself.
117 ;; Here is an approximate equivalence table to help think about it:
118 ;; (string-as-multibyte s) ~= (decode-coding-string s 'emacs-mule)
119 ;; (string-to-multibyte s) ~= (decode-coding-string s 'binary)
120 ;; (string-make-multibyte s) ~= (decode-coding-string s locale-coding-system)
121 ;; `string-as-multibyte' is an Emacs function, not available in XEmacs.
122 (string-as-multibyte . identity)
123 ;; `multibyte-string-p' is an Emacs function, not available in XEmacs.
124 (multibyte-string-p . ignore)
125 ;; `insert-byte' is available only in Emacs 23.1 or greater.
126 (insert-byte . insert-char)
127 ;; `multibyte-char-to-unibyte' is an Emacs function, not available
129 (multibyte-char-to-unibyte . identity)
130 ;; `set-buffer-multibyte' is an Emacs function, not available in XEmacs.
131 (set-buffer-multibyte . ignore)
132 ;; `substring-no-properties' is available only in Emacs 22.1 or greater.
133 (substring-no-properties
134 . ,(lambda (string &optional from to)
135 "Return a substring of STRING, without text properties.
136 It starts at index FROM and ending before TO.
137 TO may be nil or omitted; then the substring runs to the end of STRING.
138 If FROM is nil or omitted, the substring starts at the beginning of STRING.
139 If FROM or TO is negative, it counts from the end.
141 With one argument, just copy STRING without its properties."
142 (setq string (substring string (or from 0) to))
143 (set-text-properties 0 (length string) nil string)
145 ;; `line-number-at-pos' is available only in Emacs 22.1 or greater
148 . ,(lambda (&optional pos)
149 "Return (narrowed) buffer line number at position POS.
150 If POS is nil, use current buffer location.
151 Counting starts at (point-min), so the value refers
152 to the contents of the accessible portion of the buffer."
153 (let ((opoint (or pos (point))) start)
155 (goto-char (point-min))
159 (1+ (count-lines start (point))))))))))
161 ;; `special-display-p' is an Emacs function, not available in XEmacs.
162 (defalias 'mm-special-display-p
163 (if (featurep 'emacs)
165 (lambda (buffer-name)
166 "Returns non-nil if a buffer named BUFFER-NAME gets a special frame."
167 (and special-display-function
168 (or (and (member buffer-name special-display-buffer-names) t)
169 (cdr (assoc buffer-name special-display-buffer-names))
171 (dolist (elem special-display-regexps)
173 (string-match elem buffer-name)
177 (string-match (car elem) buffer-name)
178 (throw 'return (cdr elem))))))))))
180 ;; `decode-coding-string', `encode-coding-string', `decode-coding-region'
181 ;; and `encode-coding-region' are available in Emacs and XEmacs built with
182 ;; the `file-coding' feature, but the XEmacs versions treat nil, that is
183 ;; given as the `coding-system' argument, as the `binary' coding system.
185 (if (featurep 'xemacs)
186 (if (featurep 'file-coding)
188 (defun mm-decode-coding-string (str coding-system)
190 (decode-coding-string str coding-system)
192 (defun mm-encode-coding-string (str coding-system)
194 (encode-coding-string str coding-system)
196 (defun mm-decode-coding-region (start end coding-system)
198 (decode-coding-region start end coding-system)))
199 (defun mm-encode-coding-region (start end coding-system)
201 (encode-coding-region start end coding-system))))
202 (defun mm-decode-coding-string (str coding-system) str)
203 (defun mm-encode-coding-string (str coding-system) str)
204 (defalias 'mm-decode-coding-region 'ignore)
205 (defalias 'mm-encode-coding-region 'ignore))
206 (defalias 'mm-decode-coding-string 'decode-coding-string)
207 (defalias 'mm-encode-coding-string 'encode-coding-string)
208 (defalias 'mm-decode-coding-region 'decode-coding-region)
209 (defalias 'mm-encode-coding-region 'encode-coding-region)))
211 ;; `string-to-multibyte' is available only in Emacs.
212 (defalias 'mm-string-to-multibyte (if (featurep 'xemacs)
214 'string-to-multibyte))
216 ;; `char-or-char-int-p' is an XEmacs function, not available in Emacs.
218 (defalias 'mm-char-or-char-int-p
220 ((fboundp 'char-or-char-int-p) 'char-or-char-int-p)
221 ((fboundp 'char-valid-p) 'char-valid-p)
224 ;; `ucs-to-char' is a function that Mule-UCS provides.
226 (if (featurep 'xemacs)
227 (cond ((and (fboundp 'unicode-to-char) ;; XEmacs 21.5.
228 (subrp (symbol-function 'unicode-to-char)))
230 (defalias 'mm-ucs-to-char 'unicode-to-char)
231 (defun mm-ucs-to-char (codepoint)
232 "Convert Unicode codepoint to character."
233 (or (unicode-to-char codepoint) ?#))))
235 (defun mm-ucs-to-char (codepoint)
236 "Convert Unicode codepoint to character."
237 (if (fboundp 'ucs-to-char) ;; Mule-UCS is loaded.
239 (defalias 'mm-ucs-to-char
241 "Convert Unicode codepoint to character."
243 (or (ucs-to-char codepoint) ?#)
245 (mm-ucs-to-char codepoint))
247 (or (int-to-char codepoint) ?#)
250 (defun mm-ucs-to-char (codepoint)
251 "Convert Unicode codepoint to character."
253 (or (int-to-char codepoint) ?#)
255 (if (let ((char (make-char 'japanese-jisx0208 36 34)))
256 (eq char (decode-char 'ucs char)))
258 (defalias 'mm-ucs-to-char 'identity)
259 (defun mm-ucs-to-char (codepoint)
260 "Convert Unicode codepoint to character."
261 (or (decode-char 'ucs codepoint) ?#)))))
263 ;; Fixme: This seems always to be used to read a MIME charset, so it
264 ;; should be re-named and fixed (in Emacs) to offer completion only on
265 ;; proper charset names (base coding systems which have a
266 ;; mime-charset defined). XEmacs doesn't believe in mime-charset;
268 ;; `(or (coding-system-get 'iso-8859-1 'mime-charset)
269 ;; (coding-system-get 'iso-8859-1 :mime-charset))'
270 ;; Actually, there should be an `mm-coding-system-mime-charset'.
272 (defalias 'mm-read-coding-system
273 (if (featurep 'emacs) 'read-coding-system
275 ((fboundp 'read-coding-system)
276 (if (and (featurep 'xemacs)
277 (<= (string-to-number emacs-version) 21.1))
278 (lambda (prompt &optional default-coding-system)
279 (read-coding-system prompt))
280 'read-coding-system))
281 (t (lambda (prompt &optional default-coding-system)
282 "Prompt the user for a coding system."
283 (gnus-completing-read
284 prompt (mapcar (lambda (s) (symbol-name (car s)))
285 mm-mime-mule-charset-alist))))))))
287 (defvar mm-coding-system-list nil)
288 (defun mm-get-coding-system-list ()
289 "Get the coding system list."
290 (or mm-coding-system-list
291 (setq mm-coding-system-list (mm-coding-system-list))))
293 (defun mm-coding-system-p (cs)
294 "Return non-nil if CS is a symbol naming a coding system.
295 In XEmacs, also return non-nil if CS is a coding system object.
296 If CS is available, return CS itself in Emacs, and return a coding
297 system object in XEmacs."
298 (if (fboundp 'find-coding-system)
299 (and cs (find-coding-system cs))
300 (if (fboundp 'coding-system-p)
301 (when (coding-system-p cs)
304 (car (memq cs (mm-get-coding-system-list))))))
306 (defvar mm-charset-synonym-alist
308 ;; Not in XEmacs, but it's not a proper MIME charset anyhow.
309 ,@(unless (mm-coding-system-p 'x-ctext)
310 '((x-ctext . ctext)))
311 ;; ISO-8859-15 is very similar to ISO-8859-1. But it's _different_ in 8
313 ,@(unless (mm-coding-system-p 'iso-8859-15)
314 '((iso-8859-15 . iso-8859-1)))
315 ;; BIG-5HKSCS is similar to, but different than, BIG-5.
316 ,@(unless (mm-coding-system-p 'big5-hkscs)
317 '((big5-hkscs . big5)))
318 ;; A Microsoft misunderstanding.
319 ,@(when (and (not (mm-coding-system-p 'unicode))
320 (mm-coding-system-p 'utf-16-le))
321 '((unicode . utf-16-le)))
322 ;; A Microsoft misunderstanding.
323 ,@(unless (mm-coding-system-p 'ks_c_5601-1987)
324 (if (mm-coding-system-p 'cp949)
325 '((ks_c_5601-1987 . cp949))
326 '((ks_c_5601-1987 . euc-kr))))
327 ;; Windows-31J is Windows Codepage 932.
328 ,@(when (and (not (mm-coding-system-p 'windows-31j))
329 (mm-coding-system-p 'cp932))
330 '((windows-31j . cp932)))
331 ;; Charset name: GBK, Charset aliases: CP936, MS936, windows-936
332 ;; http://www.iana.org/assignments/charset-reg/GBK
333 ;; Emacs 22.1 has cp936, but not gbk, so we alias it:
334 ,@(when (and (not (mm-coding-system-p 'gbk))
335 (mm-coding-system-p 'cp936))
337 ;; UTF8 is a bogus name for UTF-8
338 ,@(when (and (not (mm-coding-system-p 'utf8))
339 (mm-coding-system-p 'utf-8))
341 ;; ISO8859-1 is a bogus name for ISO-8859-1
342 ,@(when (and (not (mm-coding-system-p 'iso8859-1))
343 (mm-coding-system-p 'iso-8859-1))
344 '((iso8859-1 . iso-8859-1)))
345 ;; ISO_8859-1 is a bogus name for ISO-8859-1
346 ,@(when (and (not (mm-coding-system-p 'iso_8859-1))
347 (mm-coding-system-p 'iso-8859-1))
348 '((iso_8859-1 . iso-8859-1)))
350 "A mapping from unknown or invalid charset names to the real charset names.
352 See `mm-codepage-iso-8859-list' and `mm-codepage-ibm-list'.")
354 (defun mm-codepage-setup (number &optional alias)
355 "Create a coding system cpNUMBER.
356 The coding system is created using `codepage-setup'. If ALIAS is
357 non-nil, an alias is created and added to
358 `mm-charset-synonym-alist'. If ALIAS is a string, it's used as
359 the alias. Else windows-NUMBER is used."
361 (let ((completion-ignore-case t)
362 (candidates (if (fboundp 'cp-supported-codepages)
363 (cp-supported-codepages)
364 ;; Removed in Emacs 23 (unicode), so signal an error:
365 (error "`codepage-setup' not present in this Emacs version"))))
366 (list (gnus-completing-read "Setup DOS Codepage" candidates
369 (setq alias (if (stringp alias)
371 (intern (format "windows-%s" number)))))
372 (let* ((cp (intern (format "cp%s" number))))
373 (unless (mm-coding-system-p cp)
374 (if (fboundp 'codepage-setup) ; silence compiler
375 (codepage-setup number)
376 (error "`codepage-setup' not present in this Emacs version")))
378 ;; Don't add alias if setup of cp failed.
379 (mm-coding-system-p cp))
380 (add-to-list 'mm-charset-synonym-alist (cons alias cp)))))
382 (defcustom mm-codepage-iso-8859-list
383 (list 1250 ;; Windows-1250 is a variant of Latin-2 heavily used by Microsoft
384 ;; Outlook users in Czech republic. Use this to allow reading of
386 '(1252 . 1) ;; Windows-1252 is a superset of iso-8859-1 (West
387 ;; Europe). See also `gnus-article-dumbquotes-map'.
388 '(1254 . 9) ;; Windows-1254 is a superset of iso-8859-9 (Turkish).
389 '(1255 . 8));; Windows-1255 is a superset of iso-8859-8 (Hebrew).
390 "A list of Windows codepage numbers and iso-8859 charset numbers.
392 If an element is a number corresponding to a supported windows
393 codepage, appropriate entries to `mm-charset-synonym-alist' are
394 added by `mm-setup-codepage-iso-8859'. An element may also be a
395 cons cell where the car is a codepage number and the cdr is the
396 corresponding number of an iso-8859 charset."
397 :type '(list (set :inline t
398 (const 1250 :tag "Central and East European")
399 (const (1252 . 1) :tag "West European")
400 (const (1254 . 9) :tag "Turkish")
401 (const (1255 . 8) :tag "Hebrew"))
405 (integer :tag "Windows codepage number")
406 (cons (integer :tag "Windows codepage number")
407 (integer :tag "iso-8859 charset number")))))
408 :version "22.1" ;; Gnus 5.10.9
411 (defcustom mm-codepage-ibm-list
412 (list 437 ;; (US etc.)
416 863 ;; (Canadian French)
421 866 ;; (Cyrillic - Russian)
426 ;; In Emacs 23 (unicode), cp... and ibm... are aliases.
427 ;; Cf. http://thread.gmane.org/v9lkng5nwy.fsf@marauder.physik.uni-ulm.de
428 "List of IBM codepage numbers.
430 The codepage mappings slightly differ between IBM and other vendors.
431 See \"ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/IBM/README.TXT\".
433 If an element is a number corresponding to a supported windows
434 codepage, appropriate entries to `mm-charset-synonym-alist' are
435 added by `mm-setup-codepage-ibm'."
436 :type '(list (set :inline t
437 (const 437 :tag "US etc.")
438 (const 860 :tag "Portugal")
439 (const 861 :tag "Iceland")
440 (const 862 :tag "Israel")
441 (const 863 :tag "Canadian French")
442 (const 865 :tag "Nordic")
444 (const 850 :tag "Latin 1")
445 (const 855 :tag "Cyrillic")
446 (const 866 :tag "Cyrillic - Russian")
447 (const 857 :tag "Turkish")
448 (const 864 :tag "Arabic")
449 (const 869 :tag "Greek")
450 (const 874 :tag "Thai"))
453 (integer :tag "Codepage number")))
454 :version "22.1" ;; Gnus 5.10.9
457 (defun mm-setup-codepage-iso-8859 (&optional list)
458 "Add appropriate entries to `mm-charset-synonym-alist'.
459 Unless LIST is given, `mm-codepage-iso-8859-list' is used."
461 (setq list mm-codepage-iso-8859-list))
463 (let (cp windows iso)
465 (setq cp (intern (format "cp%d" (car i)))
466 windows (intern (format "windows-%d" (car i)))
467 iso (intern (format "iso-8859-%d" (cdr i))))
468 (setq cp (intern (format "cp%d" i))
469 windows (intern (format "windows-%d" i))))
470 (unless (mm-coding-system-p windows)
471 (if (mm-coding-system-p cp)
472 (add-to-list 'mm-charset-synonym-alist (cons windows cp))
473 (add-to-list 'mm-charset-synonym-alist (cons windows iso)))))))
475 (defun mm-setup-codepage-ibm (&optional list)
476 "Add appropriate entries to `mm-charset-synonym-alist'.
477 Unless LIST is given, `mm-codepage-ibm-list' is used."
479 (setq list mm-codepage-ibm-list))
480 (dolist (number list)
481 (let ((ibm (intern (format "ibm%d" number)))
482 (cp (intern (format "cp%d" number))))
483 (when (and (not (mm-coding-system-p ibm))
484 (mm-coding-system-p cp))
485 (add-to-list 'mm-charset-synonym-alist (cons ibm cp))))))
488 (mm-setup-codepage-iso-8859)
489 (mm-setup-codepage-ibm)
491 ;; Note: this has to be defined before `mm-charset-to-coding-system'.
492 (defcustom mm-charset-eval-alist
493 (if (featurep 'xemacs)
494 nil ;; I don't know what would be useful for XEmacs.
495 '(;; Emacs 22 provides autoloads for 1250-1258
496 ;; (i.e. `mm-codepage-setup' does nothing).
497 (windows-1250 . (mm-codepage-setup 1250 t))
498 (windows-1251 . (mm-codepage-setup 1251 t))
499 (windows-1253 . (mm-codepage-setup 1253 t))
500 (windows-1257 . (mm-codepage-setup 1257 t))))
501 "An alist of (CHARSET . FORM) pairs.
502 If an article is encoded in an unknown CHARSET, FORM is
503 evaluated. This allows to load additional libraries providing
504 charsets on demand. If supported by your Emacs version, you
505 could use `autoload-coding-system' here."
506 :version "22.1" ;; Gnus 5.10.9
507 :type '(list (set :inline t
508 (const (windows-1250 . (mm-codepage-setup 1250 t)))
509 (const (windows-1251 . (mm-codepage-setup 1251 t)))
510 (const (windows-1253 . (mm-codepage-setup 1253 t)))
511 (const (windows-1257 . (mm-codepage-setup 1257 t)))
512 (const (cp850 . (mm-codepage-setup 850 nil))))
515 (cons (symbol :tag "charset")
516 (symbol :tag "form"))))
518 (put 'mm-charset-eval-alist 'risky-local-variable t)
520 (defvar mm-charset-override-alist)
522 ;; Note: this function has to be defined before `mm-charset-override-alist'
523 ;; since it will use this function in order to determine its default value
524 ;; when loading mm-util.elc.
525 (defun mm-charset-to-coding-system (charset &optional lbt
526 allow-override silent)
527 "Return coding-system corresponding to CHARSET.
528 CHARSET is a symbol naming a MIME charset.
529 If optional argument LBT (`unix', `dos' or `mac') is specified, it is
530 used as the line break code type of the coding system.
532 If ALLOW-OVERRIDE is given, use `mm-charset-override-alist' to
533 map undesired charset names to their replacement. This should
534 only be used for decoding, not for encoding.
536 A non-nil value of SILENT means don't issue a warning even if CHARSET
538 ;; OVERRIDE is used (only) in `mm-decode-body' and `mm-decode-string'.
539 (when (stringp charset)
540 (setq charset (intern (downcase charset))))
542 (setq charset (intern (format "%s-%s" charset lbt))))
546 ;; Running in a non-MULE environment.
547 ((or (null (mm-get-coding-system-list))
548 (not (fboundp 'coding-system-get)))
550 ;; Check override list quite early. Should only used for decoding, not for
553 (let ((cs (cdr (assq charset mm-charset-override-alist))))
554 (and cs (mm-coding-system-p cs) cs))))
556 ((or (eq charset 'us-ascii)
557 (string-match "ansi.x3.4" (symbol-name charset)))
559 ;; Check to see whether we can handle this charset. (This depends
560 ;; on there being some coding system matching each `mime-charset'
561 ;; property defined, as there should be.)
562 ((and (mm-coding-system-p charset)
563 ;;; Doing this would potentially weed out incorrect charsets.
565 ;;; (eq charset (coding-system-get charset 'mime-charset))
568 ;; Use coding system Emacs knows.
569 ((and (fboundp 'coding-system-from-name)
570 (coding-system-from-name charset)))
571 ;; Eval expressions from `mm-charset-eval-alist'
572 ((let* ((el (assq charset mm-charset-eval-alist))
579 (condition-case nil (eval form) (error nil))
580 ;; (message "Failed to eval `%s'" form))
581 (mm-coding-system-p cs)
582 (message "Added charset `%s' via `mm-charset-eval-alist'" cs))
584 ;; Translate invalid charsets.
585 ((let ((cs (cdr (assq charset mm-charset-synonym-alist))))
587 (mm-coding-system-p cs)
589 ;; "Using synonym `%s' from `mm-charset-synonym-alist' for `%s'"
592 ;; Last resort: search the coding system list for entries which
593 ;; have the right mime-charset in case the canonical name isn't
594 ;; defined (though it should be).
596 ;; mm-get-coding-system-list returns a list of cs without lbt.
598 (dolist (c (mm-get-coding-system-list))
600 (eq charset (or (coding-system-get c :mime-charset)
601 (coding-system-get c 'mime-charset))))
603 (unless (or silent cs)
604 ;; Warn the user about unknown charset:
605 (if (fboundp 'gnus-message)
606 (gnus-message 7 "Unknown charset: %s" charset)
607 (message "Unknown charset: %s" charset)))
610 ;; Note: `mm-charset-to-coding-system' has to be defined before this.
611 (defcustom mm-charset-override-alist
612 ;; Note: pairs that cannot be used in the Emacs version currently running
615 (iso-8859-1 . windows-1252)
616 (iso-8859-8 . windows-1255)
617 (iso-8859-9 . windows-1254))
618 "A mapping from undesired charset names to their replacement.
620 You may add pairs like (iso-8859-1 . windows-1252) here,
621 i.e. treat iso-8859-1 as windows-1252. windows-1252 is a
622 superset of iso-8859-1."
629 (mapcar (lambda (pair)
630 (if (mm-charset-to-coding-system (cdr pair)
634 (iso-8859-1 . windows-1252)
635 (iso-8859-8 . windows-1255)
636 (iso-8859-9 . windows-1254)
637 (undecided . windows-1252)))))
638 (val (copy-sequence (default-value 'mm-charset-override-alist)))
641 (push (if (and (prog1
642 (setq pair (assq (caar val) defaults))
643 (setq defaults (delq pair defaults)))
644 (equal (car val) pair))
647 (const :format "(%v" ,(caar val))
648 (symbol :size 3 :format " . %v)\n" ,(cdar val))))
650 (setq val (cdr val)))
652 (push `(const ,(pop defaults)) rest))
655 `(set :inline t :format "%v" ,@(nreverse rest))
656 `(repeat :inline t :tag "Other options"
658 (symbol :size 3 :format "(%v")
659 (symbol :size 3 :format " . %v)\n")))))))
660 ;; Remove pairs that cannot be used in the Emacs version currently
661 ;; running. Note that this section will be evaluated when loading
663 :set (lambda (symbol value)
666 (mapcar (lambda (pair)
667 (if (mm-charset-to-coding-system (cdr pair)
671 :version "22.1" ;; Gnus 5.10.9
674 (defvar mm-binary-coding-system
676 ((mm-coding-system-p 'binary) 'binary)
677 ((mm-coding-system-p 'no-conversion) 'no-conversion)
679 "100% binary coding system.")
681 (defvar mm-text-coding-system
682 (or (if (memq system-type '(windows-nt ms-dos))
683 (and (mm-coding-system-p 'raw-text-dos) 'raw-text-dos)
684 (and (mm-coding-system-p 'raw-text) 'raw-text))
685 mm-binary-coding-system)
686 "Text-safe coding system (For removing ^M).")
688 (defvar mm-text-coding-system-for-write nil
689 "Text coding system for write.")
691 (defvar mm-auto-save-coding-system
693 ((mm-coding-system-p 'utf-8-emacs) ; Mule 7
694 (if (memq system-type '(windows-nt ms-dos))
695 (if (mm-coding-system-p 'utf-8-emacs-dos)
696 'utf-8-emacs-dos mm-binary-coding-system)
698 ((mm-coding-system-p 'emacs-mule)
699 (if (memq system-type '(windows-nt ms-dos))
700 (if (mm-coding-system-p 'emacs-mule-dos)
701 'emacs-mule-dos mm-binary-coding-system)
703 ((mm-coding-system-p 'escape-quoted) 'escape-quoted)
704 (t mm-binary-coding-system))
705 "Coding system of auto save file.")
707 (defvar mm-universal-coding-system mm-auto-save-coding-system
708 "The universal coding system.")
710 ;; Fixme: some of the cars here aren't valid MIME charsets. That
711 ;; should only matter with XEmacs, though.
712 (defvar mm-mime-mule-charset-alist
714 (iso-8859-1 latin-iso8859-1)
715 (iso-8859-2 latin-iso8859-2)
716 (iso-8859-3 latin-iso8859-3)
717 (iso-8859-4 latin-iso8859-4)
718 (iso-8859-5 cyrillic-iso8859-5)
719 ;; Non-mule (X)Emacs uses the last mule-charset for 8bit characters.
720 ;; The fake mule-charset, gnus-koi8-r, tells Gnus that the default
721 ;; charset is koi8-r, not iso-8859-5.
722 (koi8-r cyrillic-iso8859-5 gnus-koi8-r)
723 (iso-8859-6 arabic-iso8859-6)
724 (iso-8859-7 greek-iso8859-7)
725 (iso-8859-8 hebrew-iso8859-8)
726 (iso-8859-9 latin-iso8859-9)
727 (iso-8859-14 latin-iso8859-14)
728 (iso-8859-15 latin-iso8859-15)
729 (viscii vietnamese-viscii-lower)
730 (iso-2022-jp latin-jisx0201 japanese-jisx0208 japanese-jisx0208-1978)
731 (euc-kr korean-ksc5601)
732 (gb2312 chinese-gb2312)
734 (gb18030 gb18030-2-byte
735 gb18030-4-byte-bmp gb18030-4-byte-smp
736 gb18030-4-byte-ext-1 gb18030-4-byte-ext-2)
737 (big5 chinese-big5-1 chinese-big5-2)
739 (thai-tis620 thai-tis620)
740 (windows-1251 cyrillic-iso8859-5)
741 (iso-2022-7bit ethiopic arabic-1-column arabic-2-column)
742 (iso-2022-jp-2 latin-iso8859-1 greek-iso8859-7
743 latin-jisx0201 japanese-jisx0208-1978
744 chinese-gb2312 japanese-jisx0208
745 korean-ksc5601 japanese-jisx0212)
746 (iso-2022-int-1 latin-iso8859-1 greek-iso8859-7
747 latin-jisx0201 japanese-jisx0208-1978
748 chinese-gb2312 japanese-jisx0208
749 korean-ksc5601 japanese-jisx0212
750 chinese-cns11643-1 chinese-cns11643-2)
751 (iso-2022-int-1 latin-iso8859-1 latin-iso8859-2
752 cyrillic-iso8859-5 greek-iso8859-7
753 latin-jisx0201 japanese-jisx0208-1978
754 chinese-gb2312 japanese-jisx0208
755 korean-ksc5601 japanese-jisx0212
756 chinese-cns11643-1 chinese-cns11643-2
757 chinese-cns11643-3 chinese-cns11643-4
758 chinese-cns11643-5 chinese-cns11643-6
760 (iso-2022-jp-3 latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
761 japanese-jisx0213-1 japanese-jisx0213-2)
762 (shift_jis latin-jisx0201 katakana-jisx0201 japanese-jisx0208)
763 ,(cond ((fboundp 'unicode-precedence-list)
764 (cons 'utf-8 (delq 'ascii (mapcar 'charset-name
765 (unicode-precedence-list)))))
766 ((or (not (fboundp 'charsetp)) ;; non-Mule case
767 (charsetp 'unicode-a)
768 (not (mm-coding-system-p 'mule-utf-8)))
769 '(utf-8 unicode-a unicode-b unicode-c unicode-d unicode-e))
770 (t ;; If we have utf-8 we're in Mule 5+.
773 (coding-system-get 'mule-utf-8 'safe-charsets))))))
774 "Alist of MIME-charset/MULE-charsets.")
776 (defun mm-enrich-utf-8-by-mule-ucs ()
777 "Make the `utf-8' MIME charset usable by the Mule-UCS package.
778 This function will run when the `un-define' module is loaded under
779 XEmacs, and fill the `utf-8' entry in `mm-mime-mule-charset-alist'
780 with Mule charsets. It is completely useless for Emacs."
781 (when (boundp 'unicode-basic-translation-charset-order-list)
787 'unicode-basic-translation-charset-order-list))))
788 (elem (assq 'utf-8 mm-mime-mule-charset-alist)))
791 (setq mm-mime-mule-charset-alist
792 (nconc mm-mime-mule-charset-alist
793 (list (cons 'utf-8 val))))))
796 ;; Correct by construction, but should be unnecessary for Emacs:
797 (if (featurep 'xemacs)
798 (eval-after-load "un-define" '(mm-enrich-utf-8-by-mule-ucs))
799 (when (and (fboundp 'coding-system-list)
800 (fboundp 'sort-coding-systems))
801 (let ((css (sort-coding-systems (coding-system-list 'base-only)))
805 mime (or (coding-system-get cs :mime-charset); Emacs 23 (unicode)
806 (coding-system-get cs 'mime-charset)))
808 (not (eq t (setq mule
809 (coding-system-get cs 'safe-charsets))))
810 (not (assq mime alist)))
811 (push (cons mime (delq 'ascii mule)) alist)))
812 (setq mm-mime-mule-charset-alist (nreverse alist)))))
814 (defvar mm-hack-charsets '(iso-8859-15 iso-2022-jp-2)
815 "A list of special charsets.
816 Valid elements include:
817 `iso-8859-15' convert ISO-8859-1, -9 to ISO-8859-15 if ISO-8859-15 exists.
818 `iso-2022-jp-2' convert ISO-2022-jp to ISO-2022-jp-2 if ISO-2022-jp-2 exists."
821 (defvar mm-iso-8859-15-compatible
822 '((iso-8859-1 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE")
823 (iso-8859-9 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE\xD0\xDD\xDE\xF0\xFD\xFE"))
824 "ISO-8859-15 exchangeable coding systems and inconvertible characters.")
826 (defvar mm-iso-8859-x-to-15-table
827 (and (fboundp 'coding-system-p)
828 (mm-coding-system-p 'iso-8859-15)
831 (if (mm-coding-system-p (car cs))
832 (let ((c (string-to-char
833 (decode-coding-string "\341" (car cs)))))
834 (cons (char-charset c)
837 (decode-coding-string "\341" 'iso-8859-15)) c)
838 (string-to-list (decode-coding-string (car (cdr cs))
841 mm-iso-8859-15-compatible))
842 "A table of the difference character between ISO-8859-X and ISO-8859-15.")
844 (defcustom mm-coding-system-priorities
845 (let ((lang (if (boundp 'current-language-environment)
846 (symbol-value 'current-language-environment))))
847 (cond (;; XEmacs without Mule but with `file-coding'.
849 ;; In XEmacs 21.5 it may be the one like "Japanese (UTF-8)".
850 ((string-match "\\`Japanese" lang)
851 ;; Japanese users prefer iso-2022-jp to others usually used
852 ;; for `buffer-file-coding-system', however iso-8859-1 should
853 ;; be used when there are only ASCII and Latin-1 characters.
854 '(iso-8859-1 iso-2022-jp utf-8))))
855 "Preferred coding systems for encoding outgoing messages.
857 More than one suitable coding system may be found for some text.
858 By default, the coding system with the highest priority is used
859 to encode outgoing messages (see `sort-coding-systems'). If this
860 variable is set, it overrides the default priority."
862 :type '(repeat (symbol :tag "Coding system"))
866 (defvar mm-use-find-coding-systems-region
867 (fboundp 'find-coding-systems-region)
868 "Use `find-coding-systems-region' to find proper coding systems.
870 Setting it to nil is useful on Emacsen supporting Unicode if sending
871 mail with multiple parts is preferred to sending a Unicode one.")
873 (defvar mm-extra-numeric-entities
876 (cons (car item) (mm-ucs-to-char (cdr item))))
877 '((#x80 . #x20AC) (#x82 . #x201A) (#x83 . #x0192) (#x84 . #x201E)
878 (#x85 . #x2026) (#x86 . #x2020) (#x87 . #x2021) (#x88 . #x02C6)
879 (#x89 . #x2030) (#x8A . #x0160) (#x8B . #x2039) (#x8C . #x0152)
880 (#x8E . #x017D) (#x91 . #x2018) (#x92 . #x2019) (#x93 . #x201C)
881 (#x94 . #x201D) (#x95 . #x2022) (#x96 . #x2013) (#x97 . #x2014)
882 (#x98 . #x02DC) (#x99 . #x2122) (#x9A . #x0161) (#x9B . #x203A)
883 (#x9C . #x0153) (#x9E . #x017E) (#x9F . #x0178)))
884 "*Alist of extra numeric entities and characters other than ISO 10646.
885 This table is used for decoding extra numeric entities to characters,
886 like \"€\" to the euro sign, mainly in html messages.")
888 ;;; Internal variables:
892 (defun mm-mule-charset-to-mime-charset (charset)
893 "Return the MIME charset corresponding to the given Mule CHARSET."
894 (if (and (fboundp 'find-coding-systems-for-charsets)
895 (fboundp 'sort-coding-systems))
896 (let ((css (sort (sort-coding-systems
897 (find-coding-systems-for-charsets (list charset)))
898 'mm-sort-coding-systems-predicate))
900 (while (and (not mime)
902 (when (setq cs (pop css))
903 (setq mime (or (coding-system-get cs :mime-charset)
904 (coding-system-get cs 'mime-charset)))))
906 (let ((alist (mapcar (lambda (cs)
907 (assq cs mm-mime-mule-charset-alist))
908 (sort (mapcar 'car mm-mime-mule-charset-alist)
909 'mm-sort-coding-systems-predicate)))
912 (when (memq charset (cdar alist))
913 (setq out (caar alist)
919 (if (featurep 'xemacs)
920 (defalias 'mm-enable-multibyte 'ignore)
921 (defun mm-enable-multibyte ()
922 "Set the multibyte flag of the current buffer.
923 Only do this if the default value of `enable-multibyte-characters' is
924 non-nil. This is a no-op in XEmacs."
925 (set-buffer-multibyte 'to)))
927 (if (featurep 'xemacs)
928 (defalias 'mm-disable-multibyte 'ignore)
929 (defun mm-disable-multibyte ()
930 "Unset the multibyte flag of in the current buffer.
931 This is a no-op in XEmacs."
932 (set-buffer-multibyte nil))))
934 (defun mm-preferred-coding-system (charset)
935 ;; A typo in some Emacs versions.
936 (or (get-charset-property charset 'preferred-coding-system)
937 (get-charset-property charset 'prefered-coding-system)))
939 ;; Mule charsets shouldn't be used.
940 (defsubst mm-guess-charset ()
941 "Guess Mule charset from the language environment."
943 mail-parse-mule-charset ;; cached mule-charset
945 (setq mail-parse-mule-charset
946 (and (boundp 'current-language-environment)
949 (assoc current-language-environment
950 language-info-alist))))))
951 (if (or (not mail-parse-mule-charset)
952 (eq mail-parse-mule-charset 'ascii))
953 (setq mail-parse-mule-charset
954 (or (car (last (assq mail-parse-charset
955 mm-mime-mule-charset-alist)))
958 mail-parse-mule-charset)))
960 (defun mm-charset-after (&optional pos)
961 "Return charset of a character in current buffer at position POS.
962 If POS is nil, it defaults to the current point.
963 If POS is out of range, the value is nil.
964 If the charset is `composition', return the actual one."
965 (let ((char (char-after pos)) charset)
966 (if (< (mm-char-int char) 128)
967 (setq charset 'ascii)
968 ;; charset-after is fake in some Emacsen.
969 (setq charset (and (fboundp 'char-charset) (char-charset char)))
970 (if (eq charset 'composition) ; Mule 4
971 (let ((p (or pos (point))))
972 (cadr (find-charset-region p (1+ p))))
973 (if (and charset (not (memq charset '(ascii eight-bit-control
974 eight-bit-graphic))))
976 (mm-guess-charset))))))
978 (defun mm-mime-charset (charset)
979 "Return the MIME charset corresponding to the given Mule CHARSET."
980 (if (eq charset 'unknown)
981 (error "The message contains non-printable characters, please use attachment"))
982 (if (and (fboundp 'coding-system-get) (fboundp 'get-charset-property))
984 (and (mm-preferred-coding-system charset)
985 (or (coding-system-get
986 (mm-preferred-coding-system charset) :mime-charset)
988 (mm-preferred-coding-system charset) 'mime-charset)))
989 (and (eq charset 'ascii)
991 (mm-preferred-coding-system charset)
992 (mm-mule-charset-to-mime-charset charset))
993 ;; This is for XEmacs.
994 (mm-mule-charset-to-mime-charset charset)))
996 ;; `delete-dups' is not available in XEmacs 21.4.
997 (if (fboundp 'delete-dups)
998 (defalias 'mm-delete-duplicates 'delete-dups)
999 (defun mm-delete-duplicates (list)
1000 "Destructively remove `equal' duplicates from LIST.
1001 Store the result in LIST and return it. LIST must be a proper list.
1002 Of several `equal' occurrences of an element in LIST, the first
1005 This is a compatibility function for Emacsen without `delete-dups'."
1006 ;; Code from `subr.el' in Emacs 22:
1009 (setcdr tail (delete (car tail) (cdr tail)))
1010 (setq tail (cdr tail))))
1013 ;; Fixme: This is used in places when it should be testing the
1014 ;; default multibyteness. See mm-default-multibyte-p.
1016 (if (and (not (featurep 'xemacs))
1017 (boundp 'enable-multibyte-characters))
1018 (defun mm-multibyte-p ()
1019 "Non-nil if multibyte is enabled in the current buffer."
1020 enable-multibyte-characters)
1021 (defun mm-multibyte-p () (featurep 'mule))))
1023 (defun mm-default-multibyte-p ()
1024 "Return non-nil if the session is multibyte.
1025 This affects whether coding conversion should be attempted generally."
1026 (if (featurep 'mule)
1027 (if (boundp 'enable-multibyte-characters)
1028 (default-value 'enable-multibyte-characters)
1031 (defun mm-iso-8859-x-to-15-region (&optional b e)
1032 (if (fboundp 'char-charset)
1033 (let (charset item c inconvertible)
1035 (if e (narrow-to-region b e))
1036 (goto-char (point-min))
1037 (skip-chars-forward "\0-\177")
1040 ((not (setq item (assq (char-charset (setq c (char-after)))
1041 mm-iso-8859-x-to-15-table)))
1043 ((memq c (cdr (cdr item)))
1044 (setq inconvertible t)
1047 (insert-before-markers (prog1 (+ c (car (cdr item)))
1049 (skip-chars-forward "\0-\177")))
1050 (not inconvertible))))
1052 (defun mm-sort-coding-systems-predicate (a b)
1054 (mapcar (lambda (cs)
1055 ;; Note: invalid entries are dropped silently
1056 (and (setq cs (mm-coding-system-p cs))
1057 (coding-system-base cs)))
1058 mm-coding-system-priorities)))
1059 (and (setq a (mm-coding-system-p a))
1060 (if (setq b (mm-coding-system-p b))
1061 (> (length (memq (coding-system-base a) priorities))
1062 (length (memq (coding-system-base b) priorities)))
1066 (autoload 'latin-unity-massage-name "latin-unity")
1067 (autoload 'latin-unity-maybe-remap "latin-unity")
1068 (autoload 'latin-unity-representations-feasible-region "latin-unity")
1069 (autoload 'latin-unity-representations-present-region "latin-unity"))
1071 (defvar latin-unity-coding-systems)
1072 (defvar latin-unity-ucs-list)
1074 (defun mm-xemacs-find-mime-charset-1 (begin end)
1075 "Determine which MIME charset to use to send region as message.
1076 This uses the XEmacs-specific latin-unity package to better handle the
1077 case where identical characters from diverse ISO-8859-? character sets
1078 can be encoded using a single one of the corresponding coding systems.
1080 It treats `mm-coding-system-priorities' as the list of preferred
1081 coding systems; a useful example setting for this list in Western
1082 Europe would be '(iso-8859-1 iso-8859-15 utf-8), which would default
1083 to the very standard Latin 1 coding system, and only move to coding
1084 systems that are less supported as is necessary to encode the
1085 characters that exist in the buffer.
1087 Latin Unity doesn't know about those non-ASCII Roman characters that
1088 are available in various East Asian character sets. As such, its
1089 behavior if you have a JIS 0212 LATIN SMALL LETTER A WITH ACUTE in a
1090 buffer and it can otherwise be encoded as Latin 1, won't be ideal.
1091 But this is very much a corner case, so don't worry about it."
1092 (let ((systems mm-coding-system-priorities) csets psets curset)
1094 ;; Load the Latin Unity library, if available.
1095 (when (and (not (featurep 'latin-unity)) (locate-library "latin-unity"))
1096 (require 'latin-unity))
1098 ;; Now, can we use it?
1099 (if (featurep 'latin-unity)
1101 (setq csets (latin-unity-representations-feasible-region begin end)
1102 psets (latin-unity-representations-present-region begin end))
1106 ;; Pass back the first coding system in the preferred list
1107 ;; that can encode the whole region.
1108 (dolist (curset systems)
1109 (setq curset (latin-unity-massage-name 'buffer-default curset))
1111 ;; If the coding system is a universal coding system, then
1112 ;; it can certainly encode all the characters in the region.
1113 (if (memq curset latin-unity-ucs-list)
1114 (throw 'done (list curset)))
1116 ;; If a coding system isn't universal, and isn't in
1117 ;; the list that latin unity knows about, we can't
1118 ;; decide whether to use it here. Leave that until later
1119 ;; in `mm-find-mime-charset-region' function, whence we
1120 ;; have been called.
1121 (unless (memq curset latin-unity-coding-systems)
1124 ;; Right, we know about this coding system, and it may
1125 ;; conceivably be able to encode all the characters in
1127 (if (latin-unity-maybe-remap begin end curset csets psets t)
1128 (throw 'done (list curset))))
1130 ;; Can't encode using anything from the
1131 ;; `mm-coding-system-priorities' list.
1132 ;; Leave `mm-find-mime-charset' to do most of the work.
1135 ;; Right, latin unity isn't available; let `mm-find-charset-region'
1136 ;; take its default action, which equally applies to GNU Emacs.
1139 (defmacro mm-xemacs-find-mime-charset (begin end)
1140 (when (featurep 'xemacs)
1141 `(and (featurep 'mule) (mm-xemacs-find-mime-charset-1 ,begin ,end))))
1143 (declare-function mm-delete-duplicates "mm-util" (list))
1145 (defun mm-find-mime-charset-region (b e &optional hack-charsets)
1146 "Return the MIME charsets needed to encode the region between B and E.
1147 nil means ASCII, a single-element list represents an appropriate MIME
1148 charset, and a longer list means no appropriate charset."
1150 ;; The return possibilities of this function are a mess...
1151 (or (and (mm-multibyte-p)
1152 mm-use-find-coding-systems-region
1153 ;; Find the mime-charset of the most preferred coding
1154 ;; system that has one.
1155 (let ((systems (find-coding-systems-region b e)))
1156 (when mm-coding-system-priorities
1158 (sort systems 'mm-sort-coding-systems-predicate)))
1159 (setq systems (delq 'compound-text systems))
1160 (unless (equal systems '(undecided))
1162 (let* ((head (pop systems))
1163 (cs (or (coding-system-get head :mime-charset)
1164 (coding-system-get head 'mime-charset))))
1165 ;; The mime-charset (`x-ctext') of
1166 ;; `compound-text' is not in the IANA list. We
1167 ;; shouldn't normally use anything here with a
1168 ;; mime-charset having an `x-' prefix.
1169 ;; Fixme: Allow this to be overridden, since
1170 ;; there is existing use of x-ctext.
1171 ;; Also people apparently need the coding system
1172 ;; `iso-2022-jp-3' (which Mule-UCS defines with
1173 ;; mime-charset, though it's not valid).
1175 (not (string-match "^[Xx]-" (symbol-name cs)))
1176 ;; UTF-16 of any variety is invalid for
1177 ;; text parts and, unfortunately, has
1178 ;; mime-charset defined both in Mule-UCS
1179 ;; and versions of Emacs. (The name
1180 ;; might be `mule-utf-16...' or
1182 (not (string-match "utf-16" (symbol-name cs))))
1184 charsets (list cs))))))
1186 ;; If we're XEmacs, and some coding system is appropriate,
1187 ;; mm-xemacs-find-mime-charset will return an appropriate list.
1188 ;; Otherwise, we'll get nil, and the next setq will get invoked.
1189 (setq charsets (mm-xemacs-find-mime-charset b e))
1191 ;; Fixme: won't work for unibyte Emacs 23:
1193 ;; We're not multibyte, or a single coding system won't cover it.
1195 (mm-delete-duplicates
1196 (mapcar 'mm-mime-charset
1198 (mm-find-charset-region b e))))))
1199 (if (and (> (length charsets) 1)
1200 (memq 'iso-8859-15 charsets)
1201 (memq 'iso-8859-15 hack-charsets)
1202 (save-excursion (mm-iso-8859-x-to-15-region b e)))
1203 (dolist (x mm-iso-8859-15-compatible)
1204 (setq charsets (delq (car x) charsets))))
1205 (if (and (memq 'iso-2022-jp-2 charsets)
1206 (memq 'iso-2022-jp-2 hack-charsets))
1207 (setq charsets (delq 'iso-2022-jp charsets)))
1208 ;; Attempt to reduce the number of charsets if utf-8 is available.
1209 (if (and (featurep 'xemacs)
1210 (> (length charsets) 1)
1211 (mm-coding-system-p 'utf-8))
1212 (let ((mm-coding-system-priorities
1213 (cons 'utf-8 mm-coding-system-priorities)))
1215 (mm-delete-duplicates
1216 (mapcar 'mm-mime-charset
1218 (mm-find-charset-region b e)))))))
1221 (defmacro mm-with-unibyte-buffer (&rest forms)
1222 "Create a temporary buffer, and evaluate FORMS there like `progn'.
1223 Use unibyte mode for this."
1225 (mm-disable-multibyte)
1227 (put 'mm-with-unibyte-buffer 'lisp-indent-function 0)
1228 (put 'mm-with-unibyte-buffer 'edebug-form-spec '(body))
1230 (defmacro mm-with-multibyte-buffer (&rest forms)
1231 "Create a temporary buffer, and evaluate FORMS there like `progn'.
1232 Use multibyte mode for this."
1234 (mm-enable-multibyte)
1236 (put 'mm-with-multibyte-buffer 'lisp-indent-function 0)
1237 (put 'mm-with-multibyte-buffer 'edebug-form-spec '(body))
1239 (defmacro mm-with-unibyte-current-buffer (&rest forms)
1240 "Evaluate FORMS with current buffer temporarily made unibyte.
1241 Equivalent to `progn' in XEmacs.
1243 Note: We recommend not using this macro any more; there should be
1244 better ways to do a similar thing. The previous version of this macro
1245 bound the default value of `enable-multibyte-characters' to nil while
1246 evaluating FORMS but it is no longer done. So, some programs assuming
1247 it if any may malfunction."
1248 (if (featurep 'xemacs)
1250 (let ((multibyte (make-symbol "multibyte")))
1251 `(let ((,multibyte enable-multibyte-characters))
1253 (set-buffer-multibyte nil))
1257 (set-buffer-multibyte t)))))))
1258 (put 'mm-with-unibyte-current-buffer 'lisp-indent-function 0)
1259 (put 'mm-with-unibyte-current-buffer 'edebug-form-spec '(body))
1261 (defun mm-find-charset-region (b e)
1262 "Return a list of Emacs charsets in the region B to E."
1264 ((and (mm-multibyte-p)
1265 (fboundp 'find-charset-region))
1266 ;; Remove composition since the base charsets have been included.
1267 ;; Remove eight-bit-*, treat them as ascii.
1268 (let ((css (find-charset-region b e)))
1270 '(composition eight-bit-control eight-bit-graphic control-1)
1272 (setq css (delq cs css)))))
1274 ;; We are in a unibyte buffer or XEmacs non-mule, so we futz around a bit.
1277 (narrow-to-region b e)
1278 (goto-char (point-min))
1279 (skip-chars-forward "\0-\177")
1284 (and (boundp 'current-language-environment)
1285 (car (last (assq 'charset
1286 (assoc current-language-environment
1287 language-info-alist))))))
1288 (if (eq charset 'ascii) (setq charset nil))
1291 (car (last (assq mail-parse-charset
1292 mm-mime-mule-charset-alist)))))
1293 (list 'ascii (or charset 'latin-iso8859-1)))))))))
1295 (defun mm-auto-mode-alist ()
1296 "Return an `auto-mode-alist' with only the .gz (etc) thingies."
1297 (let ((alist auto-mode-alist)
1300 (when (listp (cdar alist))
1301 (push (car alist) out))
1305 (defvar mm-inhibit-file-name-handlers
1306 '(jka-compr-handler image-file-handler epa-file-handler)
1307 "A list of handlers doing (un)compression (etc) thingies.")
1309 (defun mm-insert-file-contents (filename &optional visit beg end replace
1311 "Like `insert-file-contents', but only reads in the file.
1312 A buffer may be modified in several ways after reading into the buffer due
1313 to advanced Emacs features, such as file-name-handlers, format decoding,
1314 `find-file-hooks', etc.
1315 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'.
1316 This function ensures that none of these modifications will take place."
1317 (letf* ((format-alist nil)
1318 (auto-mode-alist (if inhibit nil (mm-auto-mode-alist)))
1319 ((default-value 'major-mode) 'fundamental-mode)
1320 (enable-local-variables nil)
1321 (after-insert-file-functions nil)
1322 (enable-local-eval nil)
1323 (inhibit-file-name-operation (if inhibit
1324 'insert-file-contents
1325 inhibit-file-name-operation))
1326 (inhibit-file-name-handlers
1328 (append mm-inhibit-file-name-handlers
1329 inhibit-file-name-handlers)
1330 inhibit-file-name-handlers))
1331 (ffh (if (boundp 'find-file-hook)
1334 (val (symbol-value ffh)))
1337 (insert-file-contents filename visit beg end replace)
1340 (defun mm-append-to-file (start end filename &optional codesys inhibit)
1341 "Append the contents of the region to the end of file FILENAME.
1342 When called from a function, expects three arguments,
1343 START, END and FILENAME. START and END are buffer positions
1344 saying what text to write.
1345 Optional fourth argument specifies the coding system to use when
1347 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1348 (let ((coding-system-for-write
1349 (or codesys mm-text-coding-system-for-write
1350 mm-text-coding-system))
1351 (inhibit-file-name-operation (if inhibit
1353 inhibit-file-name-operation))
1354 (inhibit-file-name-handlers
1356 (append mm-inhibit-file-name-handlers
1357 inhibit-file-name-handlers)
1358 inhibit-file-name-handlers)))
1359 (write-region start end filename t 'no-message)
1360 (message "Appended to %s" filename)))
1362 (defun mm-write-region (start end filename &optional append visit lockname
1363 coding-system inhibit)
1365 "Like `write-region'.
1366 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1367 (let ((coding-system-for-write
1368 (or coding-system mm-text-coding-system-for-write
1369 mm-text-coding-system))
1370 (inhibit-file-name-operation (if inhibit
1372 inhibit-file-name-operation))
1373 (inhibit-file-name-handlers
1375 (append mm-inhibit-file-name-handlers
1376 inhibit-file-name-handlers)
1377 inhibit-file-name-handlers)))
1378 (write-region start end filename append visit lockname)))
1380 (autoload 'gmm-write-region "gmm-utils")
1381 (declare-function help-function-arglist "help-fns"
1382 (def &optional preserve-names))
1384 ;; It is not a MIME function, but some MIME functions use it.
1385 (if (and (fboundp 'make-temp-file)
1387 (let ((def (if (fboundp 'compiled-function-arglist) ;; XEmacs
1388 (eval (list 'compiled-function-arglist
1389 (symbol-function 'make-temp-file)))
1391 (help-function-arglist 'make-temp-file t))))
1392 (and (>= (length def) 4)
1393 (eq (nth 3 def) 'suffix)))))
1394 (defalias 'mm-make-temp-file 'make-temp-file)
1395 ;; Stolen (and modified for XEmacs) from Emacs 22.
1396 (defun mm-make-temp-file (prefix &optional dir-flag suffix)
1397 "Create a temporary file.
1398 The returned file name (created by appending some random characters at the end
1399 of PREFIX, and expanding against `temporary-file-directory' if necessary),
1400 is guaranteed to point to a newly created empty file.
1401 You can then use `write-region' to write new data into the file.
1403 If DIR-FLAG is non-nil, create a new empty directory instead of a file.
1405 If SUFFIX is non-nil, add that at the end of the file name."
1406 (let ((umask (default-file-modes))
1410 ;; Create temp files with strict access rights. It's easy to
1411 ;; loosen them later, whereas it's impossible to close the
1412 ;; time-window of loose permissions otherwise.
1413 (set-default-file-modes 448)
1414 (while (condition-case err
1420 (if (fboundp 'temp-directory)
1423 temporary-file-directory))))
1425 (setq file (concat file suffix)))
1427 (make-directory file)
1428 ;; NOTE: This is unsafe if Emacs 20
1429 ;; users and XEmacs users don't use
1430 ;; a secure temp directory.
1431 (gmm-write-region "" nil file nil 'silent
1434 (file-already-exists t)
1435 ;; The XEmacs version of `make-directory' issues
1437 (file-error (or (and (featurep 'xemacs)
1438 (file-exists-p file))
1439 (signal (car err) (cdr err)))))
1440 ;; the file was somehow created by someone else between
1441 ;; `make-temp-name' and `write-region', let's try again.
1445 (set-default-file-modes umask)))))
1447 (defvar mm-image-load-path-cache nil)
1449 (defun mm-image-load-path (&optional package)
1450 (if (and mm-image-load-path-cache
1451 (equal load-path (car mm-image-load-path-cache)))
1452 (cdr mm-image-load-path-cache)
1454 (dolist (path load-path)
1457 (setq dir (concat (file-name-directory
1458 (directory-file-name path))
1459 "etc/images/" (or package "gnus/")))))
1461 (setq result (nreverse result)
1462 mm-image-load-path-cache (cons load-path result))
1465 ;; Fixme: This doesn't look useful where it's used.
1466 (if (fboundp 'detect-coding-region)
1467 (defun mm-detect-coding-region (start end)
1468 "Like `detect-coding-region' except returning the best one."
1469 (let ((coding-systems
1470 (detect-coding-region start end)))
1471 (or (car-safe coding-systems)
1473 (defun mm-detect-coding-region (start end)
1474 (let ((point (point)))
1476 (skip-chars-forward "\0-\177" end)
1478 (if (eq (point) end) 'ascii (mm-guess-charset))
1479 (goto-char point)))))
1481 (declare-function mm-detect-coding-region "mm-util" (start end))
1483 (if (fboundp 'coding-system-get)
1484 (defun mm-detect-mime-charset-region (start end)
1485 "Detect MIME charset of the text in the region between START and END."
1486 (let ((cs (mm-detect-coding-region start end)))
1487 (or (coding-system-get cs :mime-charset)
1488 (coding-system-get cs 'mime-charset))))
1489 (defun mm-detect-mime-charset-region (start end)
1490 "Detect MIME charset of the text in the region between START and END."
1491 (let ((cs (mm-detect-coding-region start end)))
1495 (unless (fboundp 'coding-system-to-mime-charset)
1496 (defalias 'coding-system-to-mime-charset 'ignore)))
1498 (defun mm-coding-system-to-mime-charset (coding-system)
1499 "Return the MIME charset corresponding to CODING-SYSTEM.
1500 To make this function work with XEmacs, the APEL package is required."
1502 (or (and (fboundp 'coding-system-get)
1503 (or (coding-system-get coding-system :mime-charset)
1504 (coding-system-get coding-system 'mime-charset)))
1505 (and (featurep 'xemacs)
1506 (or (and (fboundp 'coding-system-to-mime-charset)
1507 (not (eq (symbol-function 'coding-system-to-mime-charset)
1509 (and (condition-case nil
1512 (fboundp 'coding-system-to-mime-charset)))
1513 (coding-system-to-mime-charset coding-system)))))
1515 (defvar jka-compr-acceptable-retval-list)
1516 (declare-function jka-compr-make-temp-name "jka-compr" (&optional local))
1518 (defun mm-decompress-buffer (filename &optional inplace force)
1519 "Decompress buffer's contents, depending on jka-compr.
1520 Only when FORCE is t or `auto-compression-mode' is enabled and FILENAME
1521 agrees with `jka-compr-compression-info-list', decompression is done.
1522 Signal an error if FORCE is neither nil nor t and compressed data are
1523 not decompressed because `auto-compression-mode' is disabled.
1524 If INPLACE is nil, return decompressed data or nil without modifying
1525 the buffer. Otherwise, replace the buffer's contents with the
1526 decompressed data. The buffer's multibyteness must be turned off."
1529 (prog1 t (require 'jka-compr))
1530 (and (fboundp 'jka-compr-installed-p)
1531 (jka-compr-installed-p))))
1532 (let ((info (jka-compr-get-compression-info filename)))
1534 (unless (or (memq force (list nil t))
1535 (jka-compr-installed-p))
1537 (let ((prog (jka-compr-info-uncompress-program info))
1538 (args (jka-compr-info-uncompress-args info))
1539 (msg (format "%s %s..."
1540 (jka-compr-info-uncompress-message info)
1542 (err-file (jka-compr-make-temp-name))
1543 (cur (current-buffer))
1544 (coding-system-for-read mm-binary-coding-system)
1545 (coding-system-for-write mm-binary-coding-system)
1548 (mm-with-unibyte-buffer
1549 (insert-buffer-substring cur)
1552 (unless (memq (apply 'call-process-region
1553 (point-min) (point-max)
1554 prog t (list t err-file) nil args)
1555 jka-compr-acceptable-retval-list)
1557 (insert (mapconcat 'identity
1560 (insert-file-contents err-file)
1566 (format "Error while executing \"%s %s < %s\""
1567 prog (mapconcat 'identity args " ")
1569 (setq retval (buffer-string)))
1571 (setq err-msg (error-message-string err)))))
1572 (when (file-exists-p err-file)
1573 (ignore-errors (delete-file err-file)))
1576 (delete-region (point-min) (point-max))
1579 (message "%s" (or err-msg (concat msg "done")))
1583 (unless (fboundp 'coding-system-name)
1584 (defalias 'coding-system-name 'ignore))
1585 (unless (fboundp 'find-file-coding-system-for-read-from-filename)
1586 (defalias 'find-file-coding-system-for-read-from-filename 'ignore))
1587 (unless (fboundp 'find-operation-coding-system)
1588 (defalias 'find-operation-coding-system 'ignore)))
1590 (defun mm-find-buffer-file-coding-system (&optional filename)
1591 "Find coding system used to decode the contents of the current buffer.
1592 This function looks for the coding system magic cookie or examines the
1593 coding system specified by `file-coding-system-alist' being associated
1594 with FILENAME which defaults to `buffer-file-name'. Data compressed by
1595 gzip, bzip2, etc. are allowed."
1597 (setq filename buffer-file-name))
1599 (let ((decomp (unless ;; Not worth it to examine charset of tar files.
1602 "\\.\\(?:tar\\.[^.]+\\|tbz\\|tgz\\)\\'"
1604 (mm-decompress-buffer filename nil t))))
1606 (set-buffer (generate-new-buffer " *temp*"))
1607 (mm-disable-multibyte)
1609 (setq filename (file-name-sans-extension filename)))
1610 (goto-char (point-min))
1613 ((boundp 'set-auto-coding-function) ;; Emacs
1615 (or (funcall (symbol-value 'set-auto-coding-function)
1616 filename (- (point-max) (point-min)))
1617 (car (find-operation-coding-system 'insert-file-contents
1619 (let (auto-coding-alist)
1621 (funcall (symbol-value 'set-auto-coding-function)
1622 nil (- (point-max) (point-min)))
1624 ((and (featurep 'xemacs) (featurep 'file-coding)) ;; XEmacs
1625 (let ((case-fold-search t)
1626 (end (point-at-eol))
1629 (and (re-search-forward "-\\*-+[\t ]*" end t)
1631 (setq start (match-end 0))
1632 (re-search-forward "[\t ]*-+\\*-" end t))
1634 (setq end (match-beginning 0))
1636 (or (looking-at "coding:[\t ]*\\([^\t ;]+\\)")
1638 "[\t ;]+coding:[\t ]*\\([^\t ;]+\\)"
1640 (find-coding-system (setq codesys
1641 (intern (match-string 1))))
1643 (and (re-search-forward "^[\t ]*;+[\t ]*Local[\t ]+Variables:"
1646 (setq start (match-end 0))
1647 (re-search-forward "^[\t ]*;+[\t ]*End:" nil t))
1649 (setq end (match-beginning 0))
1652 "^[\t ]*;+[\t ]*coding:[\t ]*\\([^\t\n\r ]+\\)"
1654 (find-coding-system (setq codesys
1655 (intern (match-string 1))))
1658 (goto-char (point-min))
1659 (setq case-fold-search nil)
1660 (re-search-forward "^;;;coding system: "
1661 ;;(+ (point-min) 3000) t))
1663 (looking-at "[^\t\n\r ]+")
1665 (setq codesys (intern (match-string 0))))
1669 (find-file-coding-system-for-read-from-filename
1671 (coding-system-name (coding-system-base codesys)))))))
1673 (kill-buffer (current-buffer)))))))
1677 ;;; mm-util.el ends here