code.delx.au - gnu-emacs/blob - lisp/gnus/mm-util.el

   1 ;;; mm-util.el --- Utility functions for Mule and low level things
   2
   3 ;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   4 ;;   2005, 2006, 2007 Free Software Foundation, Inc.
   5
   6 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
   7 ;;      MORIOKA Tomohiko <morioka@jaist.ac.jp>
   8 ;; This file is part of GNU Emacs.
   9
  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, or (at your option)
  13 ;; any later version.
  14
  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.
  19
  20 ;; You should have received a copy of the GNU General Public License
  21 ;; along with GNU Emacs; see the file COPYING.  If not, write to the
  22 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  23 ;; Boston, MA 02110-1301, USA.
  24
  25 ;;; Commentary:
  26
  27 ;;; Code:
  28
  29 (eval-when-compile (require 'cl))
  30 (require 'mail-prsvr)
  31
  32 (eval-and-compile
  33   (if (featurep 'xemacs)
  34       (unless (ignore-errors
  35                 (require 'timer-funcs))
  36         (require 'timer))
  37     (require 'timer)))
  38
  39 (eval-and-compile
  40   (mapc
  41    (lambda (elem)
  42      (let ((nfunc (intern (format "mm-%s" (car elem)))))
  43        (if (fboundp (car elem))
  44            (defalias nfunc (car elem))
  45          (defalias nfunc (cdr elem)))))
  46    '((coding-system-list . ignore)
  47      (char-int . identity)
  48      (coding-system-equal . equal)
  49      (annotationp . ignore)
  50      (set-buffer-file-coding-system . ignore)
  51      (read-charset
  52       . (lambda (prompt)
  53           "Return a charset."
  54           (intern
  55            (completing-read
  56             prompt
  57             (mapcar (lambda (e) (list (symbol-name (car e))))
  58                     mm-mime-mule-charset-alist)
  59             nil t))))
  60      (subst-char-in-string
  61       . (lambda (from to string &optional inplace)
  62           ;; stolen (and renamed) from nnheader.el
  63           "Replace characters in STRING from FROM to TO.
  64           Unless optional argument INPLACE is non-nil, return a new string."
  65           (let ((string (if inplace string (copy-sequence string)))
  66                 (len (length string))
  67                 (idx 0))
  68             ;; Replace all occurrences of FROM with TO.
  69             (while (< idx len)
  70               (when (= (aref string idx) from)
  71                 (aset string idx to))
  72               (setq idx (1+ idx)))
  73             string)))
  74      (replace-in-string
  75       . (lambda (string regexp rep &optional literal)
  76           "See `replace-regexp-in-string', only the order of args differs."
  77           (replace-regexp-in-string regexp rep string nil literal)))
  78      (string-as-unibyte . identity)
  79      (string-make-unibyte . identity)
  80      ;; string-as-multibyte often doesn't really do what you think it does.
  81      ;; Example:
  82      ;;    (aref (string-as-multibyte "\201") 0) -> 129 (aka ?\201)
  83      ;;    (aref (string-as-multibyte "\300") 0) -> 192 (aka ?\300)
  84      ;;    (aref (string-as-multibyte "\300\201") 0) -> 192 (aka ?\300)
  85      ;;    (aref (string-as-multibyte "\300\201") 1) -> 129 (aka ?\201)
  86      ;; but
  87      ;;    (aref (string-as-multibyte "\201\300") 0) -> 2240
  88      ;;    (aref (string-as-multibyte "\201\300") 1) -> <error>
  89      ;; Better use string-to-multibyte or encode-coding-string.
  90      ;; If you really need string-as-multibyte somewhere it's usually
  91      ;; because you're using the internal emacs-mule representation (maybe
  92      ;; because you're using string-as-unibyte somewhere), which is
  93      ;; generally a problem in itself.
  94      ;; Here is an approximate equivalence table to help think about it:
  95      ;; (string-as-multibyte s)   ~= (decode-coding-string s 'emacs-mule)
  96      ;; (string-to-multibyte s)   ~= (decode-coding-string s 'binary)
  97      ;; (string-make-multibyte s) ~= (decode-coding-string s locale-coding-system)
  98      (string-as-multibyte . identity)
  99      (multibyte-string-p . ignore)
 100      (insert-byte . insert-char)
 101      (multibyte-char-to-unibyte . identity)
 102      (special-display-p
 103       . (lambda (buffer-name)
 104           "Returns non-nil if a buffer named BUFFER-NAME gets a special frame."
 105           (and special-display-function
 106                (or (and (member buffer-name special-display-buffer-names) t)
 107                    (cdr (assoc buffer-name special-display-buffer-names))
 108                    (catch 'return
 109                      (dolist (elem special-display-regexps)
 110                        (and (stringp elem)
 111                             (string-match elem buffer-name)
 112                             (throw 'return t))
 113                        (and (consp elem)
 114                             (stringp (car elem))
 115                             (string-match (car elem) buffer-name)
 116                             (throw 'return (cdr elem))))))))))))
 117
 118 (eval-and-compile
 119   (if (featurep 'xemacs)
 120       (if (featurep 'file-coding)
 121           ;; Don't modify string if CODING-SYSTEM is nil.
 122           (progn
 123             (defun mm-decode-coding-string (str coding-system)
 124               (if coding-system
 125                   (decode-coding-string str coding-system)
 126                 str))
 127             (defun mm-encode-coding-string (str coding-system)
 128               (if coding-system
 129                   (encode-coding-string str coding-system)
 130                 str))
 131             (defun mm-decode-coding-region (start end coding-system)
 132               (if coding-system
 133                   (decode-coding-region start end coding-system)))
 134             (defun mm-encode-coding-region (start end coding-system)
 135               (if coding-system
 136                   (encode-coding-region start end coding-system))))
 137         (defun mm-decode-coding-string (str coding-system) str)
 138         (defun mm-encode-coding-string (str coding-system) str)
 139         (defalias 'mm-decode-coding-region 'ignore)
 140         (defalias 'mm-encode-coding-region 'ignore))
 141     (defalias 'mm-decode-coding-string 'decode-coding-string)
 142     (defalias 'mm-encode-coding-string 'encode-coding-string)
 143     (defalias 'mm-decode-coding-region 'decode-coding-region)
 144     (defalias 'mm-encode-coding-region 'encode-coding-region)))
 145
 146 (defalias 'mm-string-to-multibyte
 147   (cond
 148    ((featurep 'xemacs)
 149     'identity)
 150    ((fboundp 'string-to-multibyte)
 151     'string-to-multibyte)
 152    (t
 153     (lambda (string)
 154       "Return a multibyte string with the same individual chars as string."
 155       (mapconcat
 156        (lambda (ch) (mm-string-as-multibyte (char-to-string ch)))
 157        string "")))))
 158
 159 (eval-and-compile
 160   (defalias 'mm-char-or-char-int-p
 161     (cond
 162      ((fboundp 'char-or-char-int-p) 'char-or-char-int-p)
 163      ((fboundp 'char-valid-p) 'char-valid-p)
 164      (t 'identity))))
 165
 166 ;; Fixme:  This seems always to be used to read a MIME charset, so it
 167 ;; should be re-named and fixed (in Emacs) to offer completion only on
 168 ;; proper charset names (base coding systems which have a
 169 ;; mime-charset defined).  XEmacs doesn't believe in mime-charset;
 170 ;; test with
 171 ;;   `(or (coding-system-get 'iso-8859-1 'mime-charset)
 172 ;;        (coding-system-get 'iso-8859-1 :mime-charset))'
 173 ;; Actually, there should be an `mm-coding-system-mime-charset'.
 174 (eval-and-compile
 175   (defalias 'mm-read-coding-system
 176     (cond
 177      ((fboundp 'read-coding-system)
 178       (if (and (featurep 'xemacs)
 179                (<= (string-to-number emacs-version) 21.1))
 180           (lambda (prompt &optional default-coding-system)
 181             (read-coding-system prompt))
 182         'read-coding-system))
 183      (t (lambda (prompt &optional default-coding-system)
 184           "Prompt the user for a coding system."
 185           (completing-read
 186            prompt (mapcar (lambda (s) (list (symbol-name (car s))))
 187                           mm-mime-mule-charset-alist)))))))
 188
 189 (defvar mm-coding-system-list nil)
 190 (defun mm-get-coding-system-list ()
 191   "Get the coding system list."
 192   (or mm-coding-system-list
 193       (setq mm-coding-system-list (mm-coding-system-list))))
 194
 195 (defun mm-coding-system-p (cs)
 196   "Return non-nil if CS is a symbol naming a coding system.
 197 In XEmacs, also return non-nil if CS is a coding system object.
 198 If CS is available, return CS itself in Emacs, and return a coding
 199 system object in XEmacs."
 200   (if (fboundp 'find-coding-system)
 201       (and cs (find-coding-system cs))
 202     (if (fboundp 'coding-system-p)
 203         (when (coding-system-p cs)
 204           cs)
 205       ;; no-MULE XEmacs:
 206       (car (memq cs (mm-get-coding-system-list))))))
 207
 208 (defun mm-codepage-setup (number &optional alias)
 209   "Create a coding system cpNUMBER.
 210 The coding system is created using `codepage-setup'.  If ALIAS is
 211 non-nil, an alias is created and added to
 212 `mm-charset-synonym-alist'.  If ALIAS is a string, it's used as
 213 the alias.  Else windows-NUMBER is used."
 214   (interactive
 215    (let ((completion-ignore-case t)
 216          (candidates (cp-supported-codepages)))
 217      (list (completing-read "Setup DOS Codepage: (default 437) " candidates
 218                             nil t nil nil "437"))))
 219   (when alias
 220     (setq alias (if (stringp alias)
 221                     (intern alias)
 222                   (intern (format "windows-%s" number)))))
 223   (let* ((cp (intern (format "cp%s" number))))
 224     (unless (mm-coding-system-p cp)
 225       (codepage-setup number))
 226     (when (and alias
 227                ;; Don't add alias if setup of cp failed.
 228                (mm-coding-system-p cp))
 229       (add-to-list 'mm-charset-synonym-alist (cons alias cp)))))
 230
 231 (defvar mm-charset-synonym-alist
 232   `(
 233     ;; Not in XEmacs, but it's not a proper MIME charset anyhow.
 234     ,@(unless (mm-coding-system-p 'x-ctext)
 235         '((x-ctext . ctext)))
 236     ;; ISO-8859-15 is very similar to ISO-8859-1.  But it's _different_ in 8
 237     ;; positions!
 238     ,@(unless (mm-coding-system-p 'iso-8859-15)
 239         '((iso-8859-15 . iso-8859-1)))
 240     ;; BIG-5HKSCS is similar to, but different than, BIG-5.
 241     ,@(unless (mm-coding-system-p 'big5-hkscs)
 242         '((big5-hkscs . big5)))
 243     ;; A Microsoft misunderstanding.
 244     ,@(when (and (not (mm-coding-system-p 'unicode))
 245                  (mm-coding-system-p 'utf-16-le))
 246         '((unicode . utf-16-le)))
 247     ;; A Microsoft misunderstanding.
 248     ,@(unless (mm-coding-system-p 'ks_c_5601-1987)
 249         (if (mm-coding-system-p 'cp949)
 250             '((ks_c_5601-1987 . cp949))
 251           '((ks_c_5601-1987 . euc-kr))))
 252     ;; Windows-31J is Windows Codepage 932.
 253     ,@(when (and (not (mm-coding-system-p 'windows-31j))
 254                  (mm-coding-system-p 'cp932))
 255         '((windows-31j . cp932)))
 256     ;; Charset name: GBK, Charset aliases: CP936, MS936, windows-936
 257     ;; http://www.iana.org/assignments/charset-reg/GBK
 258     ;; Emacs 22.1 has cp936, but not gbk, so we alias it:
 259     ,@(when (and (not (mm-coding-system-p 'gbk))
 260                  (mm-coding-system-p 'cp936))
 261         '((gbk . cp936)))
 262     ;; ISO8859-1 is a bogus name for ISO-8859-1
 263     ,@(when (and (not (mm-coding-system-p 'iso8859-1))
 264                  (mm-coding-system-p 'iso-8859-1))
 265         '((iso8859-1 . iso-8859-1)))
 266     )
 267   "A mapping from unknown or invalid charset names to the real charset names.
 268
 269 See `mm-codepage-iso-8859-list' and `mm-codepage-ibm-list'.")
 270
 271 (defcustom mm-codepage-iso-8859-list
 272   (list 1250 ;; Windows-1250 is a variant of Latin-2 heavily used by Microsoft
 273         ;; Outlook users in Czech republic.  Use this to allow reading of
 274         ;; their e-mails.  cp1250 should be defined by M-x codepage-setup
 275         ;; (Emacs 21).
 276         '(1252 . 1) ;; Windows-1252 is a superset of iso-8859-1 (West
 277                     ;; Europe).  See also `gnus-article-dumbquotes-map'.
 278         '(1254 . 9) ;; Windows-1254 is a superset of iso-8859-9 (Turkish).
 279         '(1255 . 8));; Windows-1255 is a superset of iso-8859-8 (Hebrew).
 280   "A list of Windows codepage numbers and iso-8859 charset numbers.
 281
 282 If an element is a number corresponding to a supported windows
 283 codepage, appropriate entries to `mm-charset-synonym-alist' are
 284 added by `mm-setup-codepage-iso-8859'.  An element may also be a
 285 cons cell where the car is a codepage number and the cdr is the
 286 corresponding number of an iso-8859 charset."
 287   :type '(list (set :inline t
 288                     (const 1250 :tag "Central and East European")
 289                     (const (1252 . 1) :tag "West European")
 290                     (const (1254 . 9) :tag "Turkish")
 291                     (const (1255 . 8) :tag "Hebrew"))
 292                (repeat :inline t
 293                        :tag "Other options"
 294                        (choice
 295                         (integer :tag "Windows codepage number")
 296                         (cons (integer :tag "Windows codepage number")
 297                               (integer :tag "iso-8859 charset  number")))))
 298   :version "22.1" ;; Gnus 5.10.9
 299   :group 'mime)
 300
 301 (defcustom mm-codepage-ibm-list
 302   (list 437 ;; (US etc.)
 303         860 ;; (Portugal)
 304         861 ;; (Iceland)
 305         862 ;; (Israel)
 306         863 ;; (Canadian French)
 307         865 ;; (Nordic)
 308         852 ;;
 309         850 ;; (Latin 1)
 310         855 ;; (Cyrillic)
 311         866 ;; (Cyrillic - Russian)
 312         857 ;; (Turkish)
 313         864 ;; (Arabic)
 314         869 ;; (Greek)
 315         874);; (Thai)
 316   ;; In Emacs 23 (unicode), cp... and ibm... are aliases.
 317   ;; Cf. http://thread.gmane.org/v9lkng5nwy.fsf@marauder.physik.uni-ulm.de
 318   "List of IBM codepage numbers.
 319
 320 The codepage mappings slighly differ between IBM and other vendors.
 321 See \"ftp://ftp.unicode.org/Public/MAPPINGS/VENDORS/IBM/README.TXT\".
 322
 323 If an element is a number corresponding to a supported windows
 324 codepage, appropriate entries to `mm-charset-synonym-alist' are
 325 added by `mm-setup-codepage-ibm'."
 326   :type '(list (set :inline t
 327                     (const 437 :tag "US etc.")
 328                     (const 860 :tag "Portugal")
 329                     (const 861 :tag "Iceland")
 330                     (const 862 :tag "Israel")
 331                     (const 863 :tag "Canadian French")
 332                     (const 865 :tag "Nordic")
 333                     (const 852)
 334                     (const 850 :tag "Latin 1")
 335                     (const 855 :tag "Cyrillic")
 336                     (const 866 :tag "Cyrillic - Russian")
 337                     (const 857 :tag "Turkish")
 338                     (const 864 :tag "Arabic")
 339                     (const 869 :tag "Greek")
 340                     (const 874 :tag "Thai"))
 341                (repeat :inline t
 342                        :tag "Other options"
 343                        (integer :tag "Codepage number")))
 344   :version "22.1" ;; Gnus 5.10.9
 345   :group 'mime)
 346
 347 (defun mm-setup-codepage-iso-8859 (&optional list)
 348   "Add appropriate entries to `mm-charset-synonym-alist'.
 349 Unless LIST is given, `mm-codepage-iso-8859-list' is used."
 350   (unless list
 351     (setq list mm-codepage-iso-8859-list))
 352   (dolist (i list)
 353     (let (cp windows iso)
 354       (if (consp i)
 355           (setq cp (intern (format "cp%d" (car i)))
 356                 windows (intern (format "windows-%d" (car i)))
 357                 iso (intern (format "iso-8859-%d" (cdr i))))
 358         (setq cp (intern (format "cp%d" i))
 359               windows (intern (format "windows-%d" i))))
 360       (unless (mm-coding-system-p windows)
 361         (if (mm-coding-system-p cp)
 362             (add-to-list 'mm-charset-synonym-alist (cons windows cp))
 363           (add-to-list 'mm-charset-synonym-alist (cons windows iso)))))))
 364
 365 (defun mm-setup-codepage-ibm (&optional list)
 366   "Add appropriate entries to `mm-charset-synonym-alist'.
 367 Unless LIST is given, `mm-codepage-ibm-list' is used."
 368   (unless list
 369     (setq list mm-codepage-ibm-list))
 370   (dolist (number list)
 371     (let ((ibm (intern (format "ibm%d" number)))
 372           (cp  (intern (format "cp%d" number))))
 373       (when (and (not (mm-coding-system-p ibm))
 374                  (mm-coding-system-p cp))
 375         (add-to-list 'mm-charset-synonym-alist (cons ibm cp))))))
 376
 377 ;; Initialize:
 378 (mm-setup-codepage-iso-8859)
 379 (mm-setup-codepage-ibm)
 380
 381 (defcustom mm-charset-override-alist
 382   '((iso-8859-1 . windows-1252)
 383     (iso-8859-8 . windows-1255)
 384     (iso-8859-9 . windows-1254))
 385   "A mapping from undesired charset names to their replacement.
 386
 387 You may add pairs like (iso-8859-1 . windows-1252) here,
 388 i.e. treat iso-8859-1 as windows-1252.  windows-1252 is a
 389 superset of iso-8859-1."
 390   :type '(list (set :inline t
 391                     (const (iso-8859-1 . windows-1252))
 392                     (const (iso-8859-8 . windows-1255))
 393                     (const (iso-8859-9 . windows-1254))
 394                     (const (undecided  . windows-1252)))
 395                (repeat :inline t
 396                        :tag "Other options"
 397                        (cons (symbol :tag "From charset")
 398                              (symbol :tag "To charset"))))
 399   :version "22.1" ;; Gnus 5.10.9
 400   :group 'mime)
 401
 402 (defcustom mm-charset-eval-alist
 403   (if (featurep 'xemacs)
 404       nil ;; I don't know what would be useful for XEmacs.
 405     '(;; Emacs 21 offers 1250 1251 1253 1257.  Emacs 22 provides autoloads for
 406       ;; 1250-1258 (i.e. `mm-codepage-setup' does nothing).
 407       (windows-1250 . (mm-codepage-setup 1250 t))
 408       (windows-1251 . (mm-codepage-setup 1251 t))
 409       (windows-1253 . (mm-codepage-setup 1253 t))
 410       (windows-1257 . (mm-codepage-setup 1257 t))))
 411   "An alist of (CHARSET . FORM) pairs.
 412 If an article is encoded in an unknown CHARSET, FORM is
 413 evaluated.  This allows to load additional libraries providing
 414 charsets on demand.  If supported by your Emacs version, you
 415 could use `autoload-coding-system' here."
 416   :version "22.1" ;; Gnus 5.10.9
 417   :type '(list (set :inline t
 418                     (const (windows-1250 . (mm-codepage-setup 1250 t)))
 419                     (const (windows-1251 . (mm-codepage-setup 1251 t)))
 420                     (const (windows-1253 . (mm-codepage-setup 1253 t)))
 421                     (const (windows-1257 . (mm-codepage-setup 1257 t)))
 422                     (const (cp850 . (mm-codepage-setup 850 nil))))
 423                (repeat :inline t
 424                        :tag "Other options"
 425                        (cons (symbol :tag "charset")
 426                              (symbol :tag "form"))))
 427   :group 'mime)
 428 (put 'mm-charset-eval-alist 'risky-local-variable t)
 429
 430 (defvar mm-binary-coding-system
 431   (cond
 432    ((mm-coding-system-p 'binary) 'binary)
 433    ((mm-coding-system-p 'no-conversion) 'no-conversion)
 434    (t nil))
 435   "100% binary coding system.")
 436
 437 (defvar mm-text-coding-system
 438   (or (if (memq system-type '(windows-nt ms-dos ms-windows))
 439           (and (mm-coding-system-p 'raw-text-dos) 'raw-text-dos)
 440         (and (mm-coding-system-p 'raw-text) 'raw-text))
 441       mm-binary-coding-system)
 442   "Text-safe coding system (For removing ^M).")
 443
 444 (defvar mm-text-coding-system-for-write nil
 445   "Text coding system for write.")
 446
 447 (defvar mm-auto-save-coding-system
 448   (cond
 449    ((mm-coding-system-p 'utf-8-emacs)   ; Mule 7
 450     (if (memq system-type '(windows-nt ms-dos ms-windows))
 451         (if (mm-coding-system-p 'utf-8-emacs-dos)
 452             'utf-8-emacs-dos mm-binary-coding-system)
 453       'utf-8-emacs))
 454    ((mm-coding-system-p 'emacs-mule)
 455     (if (memq system-type '(windows-nt ms-dos ms-windows))
 456         (if (mm-coding-system-p 'emacs-mule-dos)
 457             'emacs-mule-dos mm-binary-coding-system)
 458       'emacs-mule))
 459    ((mm-coding-system-p 'escape-quoted) 'escape-quoted)
 460    (t mm-binary-coding-system))
 461   "Coding system of auto save file.")
 462
 463 (defvar mm-universal-coding-system mm-auto-save-coding-system
 464   "The universal coding system.")
 465
 466 ;; Fixme: some of the cars here aren't valid MIME charsets.  That
 467 ;; should only matter with XEmacs, though.
 468 (defvar mm-mime-mule-charset-alist
 469   `((us-ascii ascii)
 470     (iso-8859-1 latin-iso8859-1)
 471     (iso-8859-2 latin-iso8859-2)
 472     (iso-8859-3 latin-iso8859-3)
 473     (iso-8859-4 latin-iso8859-4)
 474     (iso-8859-5 cyrillic-iso8859-5)
 475     ;; Non-mule (X)Emacs uses the last mule-charset for 8bit characters.
 476     ;; The fake mule-charset, gnus-koi8-r, tells Gnus that the default
 477     ;; charset is koi8-r, not iso-8859-5.
 478     (koi8-r cyrillic-iso8859-5 gnus-koi8-r)
 479     (iso-8859-6 arabic-iso8859-6)
 480     (iso-8859-7 greek-iso8859-7)
 481     (iso-8859-8 hebrew-iso8859-8)
 482     (iso-8859-9 latin-iso8859-9)
 483     (iso-8859-14 latin-iso8859-14)
 484     (iso-8859-15 latin-iso8859-15)
 485     (viscii vietnamese-viscii-lower)
 486     (iso-2022-jp latin-jisx0201 japanese-jisx0208 japanese-jisx0208-1978)
 487     (euc-kr korean-ksc5601)
 488     (gb2312 chinese-gb2312)
 489     (gbk chinese-gbk)
 490     (gb18030 gb18030-2-byte
 491              gb18030-4-byte-bmp gb18030-4-byte-smp
 492              gb18030-4-byte-ext-1 gb18030-4-byte-ext-2)
 493     (big5 chinese-big5-1 chinese-big5-2)
 494     (tibetan tibetan)
 495     (thai-tis620 thai-tis620)
 496     (windows-1251 cyrillic-iso8859-5)
 497     (iso-2022-7bit ethiopic arabic-1-column arabic-2-column)
 498     (iso-2022-jp-2 latin-iso8859-1 greek-iso8859-7
 499                    latin-jisx0201 japanese-jisx0208-1978
 500                    chinese-gb2312 japanese-jisx0208
 501                    korean-ksc5601 japanese-jisx0212)
 502     (iso-2022-int-1 latin-iso8859-1 greek-iso8859-7
 503                     latin-jisx0201 japanese-jisx0208-1978
 504                     chinese-gb2312 japanese-jisx0208
 505                     korean-ksc5601 japanese-jisx0212
 506                     chinese-cns11643-1 chinese-cns11643-2)
 507     (iso-2022-int-1 latin-iso8859-1 latin-iso8859-2
 508                     cyrillic-iso8859-5 greek-iso8859-7
 509                     latin-jisx0201 japanese-jisx0208-1978
 510                     chinese-gb2312 japanese-jisx0208
 511                     korean-ksc5601 japanese-jisx0212
 512                     chinese-cns11643-1 chinese-cns11643-2
 513                     chinese-cns11643-3 chinese-cns11643-4
 514                     chinese-cns11643-5 chinese-cns11643-6
 515                     chinese-cns11643-7)
 516     (iso-2022-jp-3 latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 517                    japanese-jisx0213-1 japanese-jisx0213-2)
 518     (shift_jis latin-jisx0201 katakana-jisx0201 japanese-jisx0208)
 519     ,(cond ((fboundp 'unicode-precedence-list)
 520             (cons 'utf-8 (delq 'ascii (mapcar 'charset-name
 521                                               (unicode-precedence-list)))))
 522            ((or (not (fboundp 'charsetp)) ;; non-Mule case
 523                 (charsetp 'unicode-a)
 524                 (not (mm-coding-system-p 'mule-utf-8)))
 525             '(utf-8 unicode-a unicode-b unicode-c unicode-d unicode-e))
 526            (t ;; If we have utf-8 we're in Mule 5+.
 527             (append '(utf-8)
 528                     (delete 'ascii
 529                             (coding-system-get 'mule-utf-8 'safe-charsets))))))
 530   "Alist of MIME-charset/MULE-charsets.")
 531
 532 (defun mm-enrich-utf-8-by-mule-ucs ()
 533   "Make the `utf-8' MIME charset usable by the Mule-UCS package.
 534 This function will run when the `un-define' module is loaded under
 535 XEmacs, and fill the `utf-8' entry in `mm-mime-mule-charset-alist'
 536 with Mule charsets.  It is completely useless for Emacs."
 537   (when (boundp 'unicode-basic-translation-charset-order-list)
 538     (condition-case nil
 539         (let ((val (delq
 540                     'ascii
 541                     (copy-sequence
 542                      (symbol-value
 543                       'unicode-basic-translation-charset-order-list))))
 544               (elem (assq 'utf-8 mm-mime-mule-charset-alist)))
 545           (if elem
 546               (setcdr elem val)
 547             (setq mm-mime-mule-charset-alist
 548                   (nconc mm-mime-mule-charset-alist
 549                          (list (cons 'utf-8 val))))))
 550       (error))))
 551
 552 ;; Correct by construction, but should be unnecessary for Emacs:
 553 (if (featurep 'xemacs)
 554     (eval-after-load "un-define" '(mm-enrich-utf-8-by-mule-ucs))
 555   (when (and (fboundp 'coding-system-list)
 556              (fboundp 'sort-coding-systems))
 557     (let ((css (sort-coding-systems (coding-system-list 'base-only)))
 558           cs mime mule alist)
 559       (while css
 560         (setq cs (pop css)
 561               mime (or (coding-system-get cs :mime-charset); Emacs 23 (unicode)
 562                        (coding-system-get cs 'mime-charset)))
 563         (when (and mime
 564                    (not (eq t (setq mule
 565                                     (coding-system-get cs 'safe-charsets))))
 566                    (not (assq mime alist)))
 567           (push (cons mime (delq 'ascii mule)) alist)))
 568       (setq mm-mime-mule-charset-alist (nreverse alist)))))
 569
 570 (defcustom mm-coding-system-priorities
 571   (if (boundp 'current-language-environment)
 572       (let ((lang (symbol-value 'current-language-environment)))
 573         (cond ((string= lang "Japanese")
 574                ;; Japanese users prefer iso-2022-jp to euc-japan or
 575                ;; shift_jis, however iso-8859-1 should be used when
 576                ;; there are only ASCII text and Latin-1 characters.
 577                '(iso-8859-1 iso-2022-jp iso-2022-jp-2 shift_jis utf-8)))))
 578   "Preferred coding systems for encoding outgoing messages.
 579
 580 More than one suitable coding system may be found for some text.
 581 By default, the coding system with the highest priority is used
 582 to encode outgoing messages (see `sort-coding-systems').  If this
 583 variable is set, it overrides the default priority."
 584   :version "21.2"
 585   :type '(repeat (symbol :tag "Coding system"))
 586   :group 'mime)
 587
 588 ;; ??
 589 (defvar mm-use-find-coding-systems-region
 590   (fboundp 'find-coding-systems-region)
 591   "Use `find-coding-systems-region' to find proper coding systems.
 592
 593 Setting it to nil is useful on Emacsen supporting Unicode if sending
 594 mail with multiple parts is preferred to sending a Unicode one.")
 595
 596 ;;; Internal variables:
 597
 598 ;;; Functions:
 599
 600 (defun mm-mule-charset-to-mime-charset (charset)
 601   "Return the MIME charset corresponding to the given Mule CHARSET."
 602   (if (and (fboundp 'find-coding-systems-for-charsets)
 603            (fboundp 'sort-coding-systems))
 604       (let ((css (sort (sort-coding-systems
 605                         (find-coding-systems-for-charsets (list charset)))
 606                        'mm-sort-coding-systems-predicate))
 607             cs mime)
 608         (while (and (not mime)
 609                     css)
 610           (when (setq cs (pop css))
 611             (setq mime (or (coding-system-get cs :mime-charset)
 612                            (coding-system-get cs 'mime-charset)))))
 613         mime)
 614     (let ((alist (mapcar (lambda (cs)
 615                            (assq cs mm-mime-mule-charset-alist))
 616                          (sort (mapcar 'car mm-mime-mule-charset-alist)
 617                                'mm-sort-coding-systems-predicate)))
 618           out)
 619       (while alist
 620         (when (memq charset (cdar alist))
 621           (setq out (caar alist)
 622                 alist nil))
 623         (pop alist))
 624       out)))
 625
 626 (defun mm-charset-to-coding-system (charset &optional lbt
 627                                             allow-override)
 628   "Return coding-system corresponding to CHARSET.
 629 CHARSET is a symbol naming a MIME charset.
 630 If optional argument LBT (`unix', `dos' or `mac') is specified, it is
 631 used as the line break code type of the coding system.
 632
 633 If ALLOW-OVERRIDE is given, use `mm-charset-override-alist' to
 634 map undesired charset names to their replacement.  This should
 635 only be used for decoding, not for encoding."
 636   ;; OVERRIDE is used (only) in `mm-decode-body' and `mm-decode-string'.
 637   (when (stringp charset)
 638     (setq charset (intern (downcase charset))))
 639   (when lbt
 640     (setq charset (intern (format "%s-%s" charset lbt))))
 641   (cond
 642    ((null charset)
 643     charset)
 644    ;; Running in a non-MULE environment.
 645    ((or (null (mm-get-coding-system-list))
 646         (not (fboundp 'coding-system-get)))
 647     charset)
 648    ;; Check override list quite early.  Should only used for decoding, not for
 649    ;; encoding!
 650    ((and allow-override
 651          (let ((cs (cdr (assq charset mm-charset-override-alist))))
 652            (and cs (mm-coding-system-p cs) cs))))
 653    ;; ascii
 654    ((eq charset 'us-ascii)
 655     'ascii)
 656    ;; Check to see whether we can handle this charset.  (This depends
 657    ;; on there being some coding system matching each `mime-charset'
 658    ;; property defined, as there should be.)
 659    ((and (mm-coding-system-p charset)
 660 ;;; Doing this would potentially weed out incorrect charsets.
 661 ;;;      charset
 662 ;;;      (eq charset (coding-system-get charset 'mime-charset))
 663          )
 664     charset)
 665    ;; Eval expressions from `mm-charset-eval-alist'
 666    ((let* ((el (assq charset mm-charset-eval-alist))
 667            (cs (car el))
 668            (form (cdr el)))
 669       (and cs
 670            form
 671            (prog2
 672                ;; Avoid errors...
 673                (condition-case nil (eval form) (error nil))
 674                ;; (message "Failed to eval `%s'" form))
 675                (mm-coding-system-p cs)
 676              (message "Added charset `%s' via `mm-charset-eval-alist'" cs))
 677            cs)))
 678    ;; Translate invalid charsets.
 679    ((let ((cs (cdr (assq charset mm-charset-synonym-alist))))
 680       (and cs
 681            (mm-coding-system-p cs)
 682            ;; (message
 683            ;;  "Using synonym `%s' from `mm-charset-synonym-alist' for `%s'"
 684            ;;  cs charset)
 685            cs)))
 686    ;; Last resort: search the coding system list for entries which
 687    ;; have the right mime-charset in case the canonical name isn't
 688    ;; defined (though it should be).
 689    ((let (cs)
 690       ;; mm-get-coding-system-list returns a list of cs without lbt.
 691       ;; Do we need -lbt?
 692       (dolist (c (mm-get-coding-system-list))
 693         (if (and (null cs)
 694                  (eq charset (or (coding-system-get c :mime-charset)
 695                                  (coding-system-get c 'mime-charset))))
 696             (setq cs c)))
 697       (unless cs
 698         ;; Warn the user about unknown charset:
 699         (if (fboundp 'gnus-message)
 700             (gnus-message 7 "Unknown charset: %s" charset)
 701           (message "Unknown charset: %s" charset)))
 702       cs))))
 703
 704 (eval-and-compile
 705   (defvar mm-emacs-mule (and (not (featurep 'xemacs))
 706                              (boundp 'default-enable-multibyte-characters)
 707                              default-enable-multibyte-characters
 708                              (fboundp 'set-buffer-multibyte))
 709     "True in Emacs with Mule.")
 710
 711   (if mm-emacs-mule
 712       (defun mm-enable-multibyte ()
 713         "Set the multibyte flag of the current buffer.
 714 Only do this if the default value of `enable-multibyte-characters' is
 715 non-nil.  This is a no-op in XEmacs."
 716         (set-buffer-multibyte 'to))
 717     (defalias 'mm-enable-multibyte 'ignore))
 718
 719   (if mm-emacs-mule
 720       (defun mm-disable-multibyte ()
 721         "Unset the multibyte flag of in the current buffer.
 722 This is a no-op in XEmacs."
 723         (set-buffer-multibyte nil))
 724     (defalias 'mm-disable-multibyte 'ignore)))
 725
 726 (defun mm-preferred-coding-system (charset)
 727   ;; A typo in some Emacs versions.
 728   (or (get-charset-property charset 'preferred-coding-system)
 729       (get-charset-property charset 'prefered-coding-system)))
 730
 731 ;; Mule charsets shouldn't be used.
 732 (defsubst mm-guess-charset ()
 733   "Guess Mule charset from the language environment."
 734   (or
 735    mail-parse-mule-charset ;; cached mule-charset
 736    (progn
 737      (setq mail-parse-mule-charset
 738            (and (boundp 'current-language-environment)
 739                 (car (last
 740                       (assq 'charset
 741                             (assoc current-language-environment
 742                                    language-info-alist))))))
 743      (if (or (not mail-parse-mule-charset)
 744              (eq mail-parse-mule-charset 'ascii))
 745          (setq mail-parse-mule-charset
 746                (or (car (last (assq mail-parse-charset
 747                                     mm-mime-mule-charset-alist)))
 748                    ;; default
 749                    'latin-iso8859-1)))
 750      mail-parse-mule-charset)))
 751
 752 (defun mm-charset-after (&optional pos)
 753   "Return charset of a character in current buffer at position POS.
 754 If POS is nil, it defauls to the current point.
 755 If POS is out of range, the value is nil.
 756 If the charset is `composition', return the actual one."
 757   (let ((char (char-after pos)) charset)
 758     (if (< (mm-char-int char) 128)
 759         (setq charset 'ascii)
 760       ;; charset-after is fake in some Emacsen.
 761       (setq charset (and (fboundp 'char-charset) (char-charset char)))
 762       (if (eq charset 'composition)     ; Mule 4
 763           (let ((p (or pos (point))))
 764             (cadr (find-charset-region p (1+ p))))
 765         (if (and charset (not (memq charset '(ascii eight-bit-control
 766                                                     eight-bit-graphic))))
 767             charset
 768           (mm-guess-charset))))))
 769
 770 (defun mm-mime-charset (charset)
 771   "Return the MIME charset corresponding to the given Mule CHARSET."
 772   (if (eq charset 'unknown)
 773       (error "The message contains non-printable characters, please use attachment"))
 774   (if (and (fboundp 'coding-system-get) (fboundp 'get-charset-property))
 775       ;; This exists in Emacs 20.
 776       (or
 777        (and (mm-preferred-coding-system charset)
 778             (or (coding-system-get
 779                  (mm-preferred-coding-system charset) :mime-charset)
 780                 (coding-system-get
 781                  (mm-preferred-coding-system charset) 'mime-charset)))
 782        (and (eq charset 'ascii)
 783             'us-ascii)
 784        (mm-preferred-coding-system charset)
 785        (mm-mule-charset-to-mime-charset charset))
 786     ;; This is for XEmacs.
 787     (mm-mule-charset-to-mime-charset charset)))
 788
 789 (if (fboundp 'delete-dups)
 790     (defalias 'mm-delete-duplicates 'delete-dups)
 791   (defun mm-delete-duplicates (list)
 792     "Destructively remove `equal' duplicates from LIST.
 793 Store the result in LIST and return it.  LIST must be a proper list.
 794 Of several `equal' occurrences of an element in LIST, the first
 795 one is kept.
 796
 797 This is a compatibility function for Emacsen without `delete-dups'."
 798     ;; Code from `subr.el' in Emacs 22:
 799     (let ((tail list))
 800       (while tail
 801         (setcdr tail (delete (car tail) (cdr tail)))
 802         (setq tail (cdr tail))))
 803     list))
 804
 805 ;; Fixme:  This is used in places when it should be testing the
 806 ;; default multibyteness.  See mm-default-multibyte-p.
 807 (eval-and-compile
 808   (if (and (not (featurep 'xemacs))
 809            (boundp 'enable-multibyte-characters))
 810       (defun mm-multibyte-p ()
 811         "Non-nil if multibyte is enabled in the current buffer."
 812         enable-multibyte-characters)
 813     (defun mm-multibyte-p () (featurep 'mule))))
 814
 815 (defun mm-default-multibyte-p ()
 816   "Return non-nil if the session is multibyte.
 817 This affects whether coding conversion should be attempted generally."
 818   (if (featurep 'mule)
 819       (if (boundp 'default-enable-multibyte-characters)
 820           default-enable-multibyte-characters
 821         t)))
 822
 823 (defun mm-sort-coding-systems-predicate (a b)
 824   (let ((priorities
 825          (mapcar (lambda (cs)
 826                    ;; Note: invalid entries are dropped silently
 827                    (and (setq cs (mm-coding-system-p cs))
 828                         (coding-system-base cs)))
 829                  mm-coding-system-priorities)))
 830     (and (setq a (mm-coding-system-p a))
 831          (if (setq b (mm-coding-system-p b))
 832              (> (length (memq (coding-system-base a) priorities))
 833                 (length (memq (coding-system-base b) priorities)))
 834            t))))
 835
 836 (eval-when-compile
 837   (autoload 'latin-unity-massage-name "latin-unity")
 838   (autoload 'latin-unity-maybe-remap "latin-unity")
 839   (autoload 'latin-unity-representations-feasible-region "latin-unity")
 840   (autoload 'latin-unity-representations-present-region "latin-unity")
 841   (defvar latin-unity-coding-systems)
 842   (defvar latin-unity-ucs-list))
 843
 844 (defun mm-xemacs-find-mime-charset-1 (begin end)
 845   "Determine which MIME charset to use to send region as message.
 846 This uses the XEmacs-specific latin-unity package to better handle the
 847 case where identical characters from diverse ISO-8859-? character sets
 848 can be encoded using a single one of the corresponding coding systems.
 849
 850 It treats `mm-coding-system-priorities' as the list of preferred
 851 coding systems; a useful example setting for this list in Western
 852 Europe would be '(iso-8859-1 iso-8859-15 utf-8), which would default
 853 to the very standard Latin 1 coding system, and only move to coding
 854 systems that are less supported as is necessary to encode the
 855 characters that exist in the buffer.
 856
 857 Latin Unity doesn't know about those non-ASCII Roman characters that
 858 are available in various East Asian character sets.  As such, its
 859 behavior if you have a JIS 0212 LATIN SMALL LETTER A WITH ACUTE in a
 860 buffer and it can otherwise be encoded as Latin 1, won't be ideal.
 861 But this is very much a corner case, so don't worry about it."
 862   (let ((systems mm-coding-system-priorities) csets psets curset)
 863
 864     ;; Load the Latin Unity library, if available.
 865     (when (and (not (featurep 'latin-unity)) (locate-library "latin-unity"))
 866       (require 'latin-unity))
 867
 868     ;; Now, can we use it?
 869     (if (featurep 'latin-unity)
 870         (progn
 871           (setq csets (latin-unity-representations-feasible-region begin end)
 872                 psets (latin-unity-representations-present-region begin end))
 873
 874           (catch 'done
 875
 876             ;; Pass back the first coding system in the preferred list
 877             ;; that can encode the whole region.
 878             (dolist (curset systems)
 879               (setq curset (latin-unity-massage-name 'buffer-default curset))
 880
 881               ;; If the coding system is a universal coding system, then
 882               ;; it can certainly encode all the characters in the region.
 883               (if (memq curset latin-unity-ucs-list)
 884                   (throw 'done (list curset)))
 885
 886               ;; If a coding system isn't universal, and isn't in
 887               ;; the list that latin unity knows about, we can't
 888               ;; decide whether to use it here. Leave that until later
 889               ;; in `mm-find-mime-charset-region' function, whence we
 890               ;; have been called.
 891               (unless (memq curset latin-unity-coding-systems)
 892                 (throw 'done nil))
 893
 894               ;; Right, we know about this coding system, and it may
 895               ;; conceivably be able to encode all the characters in
 896               ;; the region.
 897               (if (latin-unity-maybe-remap begin end curset csets psets t)
 898                   (throw 'done (list curset))))
 899
 900             ;; Can't encode using anything from the
 901             ;; `mm-coding-system-priorities' list.
 902             ;; Leave `mm-find-mime-charset' to do most of the work.
 903             nil))
 904
 905       ;; Right, latin unity isn't available; let `mm-find-charset-region'
 906       ;; take its default action, which equally applies to GNU Emacs.
 907       nil)))
 908
 909 (defmacro mm-xemacs-find-mime-charset (begin end)
 910   (when (featurep 'xemacs)
 911     `(and (featurep 'mule) (mm-xemacs-find-mime-charset-1 ,begin ,end))))
 912
 913 (defun mm-find-mime-charset-region (b e &optional hack-charsets)
 914   "Return the MIME charsets needed to encode the region between B and E.
 915 nil means ASCII, a single-element list represents an appropriate MIME
 916 charset, and a longer list means no appropriate charset."
 917   (let (charsets)
 918     ;; The return possibilities of this function are a mess...
 919     (or (and (mm-multibyte-p)
 920              mm-use-find-coding-systems-region
 921              ;; Find the mime-charset of the most preferred coding
 922              ;; system that has one.
 923              (let ((systems (find-coding-systems-region b e)))
 924                (when mm-coding-system-priorities
 925                  (setq systems
 926                        (sort systems 'mm-sort-coding-systems-predicate)))
 927                (setq systems (delq 'compound-text systems))
 928                (unless (equal systems '(undecided))
 929                  (while systems
 930                    (let* ((head (pop systems))
 931                           (cs (or (coding-system-get head :mime-charset)
 932                                   (coding-system-get head 'mime-charset))))
 933                      ;; The mime-charset (`x-ctext') of
 934                      ;; `compound-text' is not in the IANA list.  We
 935                      ;; shouldn't normally use anything here with a
 936                      ;; mime-charset having an `x-' prefix.
 937                      ;; Fixme:  Allow this to be overridden, since
 938                      ;; there is existing use of x-ctext.
 939                      ;; Also people apparently need the coding system
 940                      ;; `iso-2022-jp-3' (which Mule-UCS defines with
 941                      ;; mime-charset, though it's not valid).
 942                      (if (and cs
 943                               (not (string-match "^[Xx]-" (symbol-name cs)))
 944                               ;; UTF-16 of any variety is invalid for
 945                               ;; text parts and, unfortunately, has
 946                               ;; mime-charset defined both in Mule-UCS
 947                               ;; and versions of Emacs.  (The name
 948                               ;; might be `mule-utf-16...'  or
 949                               ;; `utf-16...'.)
 950                               (not (string-match "utf-16" (symbol-name cs))))
 951                          (setq systems nil
 952                                charsets (list cs))))))
 953                charsets))
 954         ;; If we're XEmacs, and some coding system is appropriate,
 955         ;; mm-xemacs-find-mime-charset will return an appropriate list.
 956         ;; Otherwise, we'll get nil, and the next setq will get invoked.
 957         (setq charsets (mm-xemacs-find-mime-charset b e))
 958
 959         ;; Fixme: won't work for unibyte Emacs 23:
 960
 961         ;; We're not multibyte, or a single coding system won't cover it.
 962         (setq charsets
 963               (mm-delete-duplicates
 964                (mapcar 'mm-mime-charset
 965                        (delq 'ascii
 966                              (mm-find-charset-region b e))))))
 967     charsets))
 968
 969 (defmacro mm-with-unibyte-buffer (&rest forms)
 970   "Create a temporary buffer, and evaluate FORMS there like `progn'.
 971 Use unibyte mode for this."
 972   `(let (default-enable-multibyte-characters)
 973      (with-temp-buffer ,@forms)))
 974 (put 'mm-with-unibyte-buffer 'lisp-indent-function 0)
 975 (put 'mm-with-unibyte-buffer 'edebug-form-spec '(body))
 976
 977 (defmacro mm-with-multibyte-buffer (&rest forms)
 978   "Create a temporary buffer, and evaluate FORMS there like `progn'.
 979 Use multibyte mode for this."
 980   `(let ((default-enable-multibyte-characters t))
 981      (with-temp-buffer ,@forms)))
 982 (put 'mm-with-multibyte-buffer 'lisp-indent-function 0)
 983 (put 'mm-with-multibyte-buffer 'edebug-form-spec '(body))
 984
 985 (defmacro mm-with-unibyte-current-buffer (&rest forms)
 986   "Evaluate FORMS with current buffer temporarily made unibyte.
 987 Also bind `default-enable-multibyte-characters' to nil.
 988 Equivalent to `progn' in XEmacs
 989
 990 NOTE: Use this macro with caution in multibyte buffers (it is not
 991 worth using this macro in unibyte buffers of course).  Use of
 992 `(set-buffer-multibyte t)', which is run finally, is generally
 993 harmful since it is likely to modify existing data in the buffer.
 994 For instance, it converts \"\\300\\255\" into \"\\255\" in
 995 Emacs 23 (unicode)."
 996   (let ((multibyte (make-symbol "multibyte"))
 997         (buffer (make-symbol "buffer")))
 998     `(if mm-emacs-mule
 999          (let ((,multibyte enable-multibyte-characters)
1000                (,buffer (current-buffer)))
1001            (unwind-protect
1002                (let (default-enable-multibyte-characters)
1003                  (set-buffer-multibyte nil)
1004                  ,@forms)
1005              (set-buffer ,buffer)
1006              (set-buffer-multibyte ,multibyte)))
1007        (let (default-enable-multibyte-characters)
1008          ,@forms))))
1009 (put 'mm-with-unibyte-current-buffer 'lisp-indent-function 0)
1010 (put 'mm-with-unibyte-current-buffer 'edebug-form-spec '(body))
1011
1012 (defmacro mm-with-unibyte (&rest forms)
1013   "Eval the FORMS with the default value of `enable-multibyte-characters' nil."
1014   `(let (default-enable-multibyte-characters)
1015      ,@forms))
1016 (put 'mm-with-unibyte 'lisp-indent-function 0)
1017 (put 'mm-with-unibyte 'edebug-form-spec '(body))
1018
1019 (defmacro mm-with-multibyte (&rest forms)
1020   "Eval the FORMS with the default value of `enable-multibyte-characters' t."
1021   `(let ((default-enable-multibyte-characters t))
1022      ,@forms))
1023 (put 'mm-with-multibyte 'lisp-indent-function 0)
1024 (put 'mm-with-multibyte 'edebug-form-spec '(body))
1025
1026 (defun mm-find-charset-region (b e)
1027   "Return a list of Emacs charsets in the region B to E."
1028   (cond
1029    ((and (mm-multibyte-p)
1030          (fboundp 'find-charset-region))
1031     ;; Remove composition since the base charsets have been included.
1032     ;; Remove eight-bit-*, treat them as ascii.
1033     (let ((css (find-charset-region b e)))
1034       (dolist (cs
1035                '(composition eight-bit-control eight-bit-graphic control-1)
1036                css)
1037         (setq css (delq cs css)))))
1038    (t
1039     ;; We are in a unibyte buffer or XEmacs non-mule, so we futz around a bit.
1040     (save-excursion
1041       (save-restriction
1042         (narrow-to-region b e)
1043         (goto-char (point-min))
1044         (skip-chars-forward "\0-\177")
1045         (if (eobp)
1046             '(ascii)
1047           (let (charset)
1048             (setq charset
1049                   (and (boundp 'current-language-environment)
1050                        (car (last (assq 'charset
1051                                         (assoc current-language-environment
1052                                                language-info-alist))))))
1053             (if (eq charset 'ascii) (setq charset nil))
1054             (or charset
1055                 (setq charset
1056                       (car (last (assq mail-parse-charset
1057                                        mm-mime-mule-charset-alist)))))
1058             (list 'ascii (or charset 'latin-iso8859-1)))))))))
1059
1060 (defun mm-auto-mode-alist ()
1061   "Return an `auto-mode-alist' with only the .gz (etc) thingies."
1062   (let ((alist auto-mode-alist)
1063         out)
1064     (while alist
1065       (when (listp (cdar alist))
1066         (push (car alist) out))
1067       (pop alist))
1068     (nreverse out)))
1069
1070 (defvar mm-inhibit-file-name-handlers
1071   '(jka-compr-handler image-file-handler epa-file-handler)
1072   "A list of handlers doing (un)compression (etc) thingies.")
1073
1074 (defun mm-insert-file-contents (filename &optional visit beg end replace
1075                                          inhibit)
1076   "Like `insert-file-contents', but only reads in the file.
1077 A buffer may be modified in several ways after reading into the buffer due
1078 to advanced Emacs features, such as file-name-handlers, format decoding,
1079 `find-file-hooks', etc.
1080 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'.
1081   This function ensures that none of these modifications will take place."
1082   (let* ((format-alist nil)
1083          (auto-mode-alist (if inhibit nil (mm-auto-mode-alist)))
1084          (default-major-mode 'fundamental-mode)
1085          (enable-local-variables nil)
1086          (after-insert-file-functions nil)
1087          (enable-local-eval nil)
1088          (inhibit-file-name-operation (if inhibit
1089                                           'insert-file-contents
1090                                         inhibit-file-name-operation))
1091          (inhibit-file-name-handlers
1092           (if inhibit
1093               (append mm-inhibit-file-name-handlers
1094                       inhibit-file-name-handlers)
1095             inhibit-file-name-handlers))
1096          (ffh (if (boundp 'find-file-hook)
1097                   'find-file-hook
1098                 'find-file-hooks))
1099          (val (symbol-value ffh)))
1100     (set ffh nil)
1101     (unwind-protect
1102         (insert-file-contents filename visit beg end replace)
1103       (set ffh val))))
1104
1105 (defun mm-append-to-file (start end filename &optional codesys inhibit)
1106   "Append the contents of the region to the end of file FILENAME.
1107 When called from a function, expects three arguments,
1108 START, END and FILENAME.  START and END are buffer positions
1109 saying what text to write.
1110 Optional fourth argument specifies the coding system to use when
1111 encoding the file.
1112 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1113   (let ((coding-system-for-write
1114          (or codesys mm-text-coding-system-for-write
1115              mm-text-coding-system))
1116         (inhibit-file-name-operation (if inhibit
1117                                          'append-to-file
1118                                        inhibit-file-name-operation))
1119         (inhibit-file-name-handlers
1120          (if inhibit
1121              (append mm-inhibit-file-name-handlers
1122                      inhibit-file-name-handlers)
1123            inhibit-file-name-handlers)))
1124     (write-region start end filename t 'no-message)
1125     (message "Appended to %s" filename)))
1126
1127 (defun mm-write-region (start end filename &optional append visit lockname
1128                               coding-system inhibit)
1129
1130   "Like `write-region'.
1131 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1132   (let ((coding-system-for-write
1133          (or coding-system mm-text-coding-system-for-write
1134              mm-text-coding-system))
1135         (inhibit-file-name-operation (if inhibit
1136                                          'write-region
1137                                        inhibit-file-name-operation))
1138         (inhibit-file-name-handlers
1139          (if inhibit
1140              (append mm-inhibit-file-name-handlers
1141                      inhibit-file-name-handlers)
1142            inhibit-file-name-handlers)))
1143     (write-region start end filename append visit lockname)))
1144
1145 ;; It is not a MIME function, but some MIME functions use it.
1146 (if (and (fboundp 'make-temp-file)
1147          (ignore-errors
1148            (let ((def (symbol-function 'make-temp-file)))
1149              (and (byte-code-function-p def)
1150                   (setq def (if (fboundp 'compiled-function-arglist)
1151                                 ;; XEmacs
1152                                 (eval (list 'compiled-function-arglist def))
1153                               (aref def 0)))
1154                   (>= (length def) 4)
1155                   (eq (nth 3 def) 'suffix)))))
1156     (defalias 'mm-make-temp-file 'make-temp-file)
1157   ;; Stolen (and modified for XEmacs) from Emacs 22.
1158   (defun mm-make-temp-file (prefix &optional dir-flag suffix)
1159     "Create a temporary file.
1160 The returned file name (created by appending some random characters at the end
1161 of PREFIX, and expanding against `temporary-file-directory' if necessary),
1162 is guaranteed to point to a newly created empty file.
1163 You can then use `write-region' to write new data into the file.
1164
1165 If DIR-FLAG is non-nil, create a new empty directory instead of a file.
1166
1167 If SUFFIX is non-nil, add that at the end of the file name."
1168     (let ((umask (default-file-modes))
1169           file)
1170       (unwind-protect
1171           (progn
1172             ;; Create temp files with strict access rights.  It's easy to
1173             ;; loosen them later, whereas it's impossible to close the
1174             ;; time-window of loose permissions otherwise.
1175             (set-default-file-modes 448)
1176             (while (condition-case err
1177                        (progn
1178                          (setq file
1179                                (make-temp-name
1180                                 (expand-file-name
1181                                  prefix
1182                                  (if (fboundp 'temp-directory)
1183                                      ;; XEmacs
1184                                      (temp-directory)
1185                                    temporary-file-directory))))
1186                          (if suffix
1187                              (setq file (concat file suffix)))
1188                          (if dir-flag
1189                              (make-directory file)
1190                            ;; NOTE: This is unsafe if Emacs 20
1191                            ;; users and XEmacs users don't use
1192                            ;; a secure temp directory.
1193                            (gmm-write-region "" nil file nil 'silent
1194                                              nil 'excl))
1195                          nil)
1196                      (file-already-exists t)
1197                      ;; The XEmacs version of `make-directory' issues
1198                      ;; `file-error'.
1199                      (file-error (or (and (featurep 'xemacs)
1200                                           (file-exists-p file))
1201                                      (signal (car err) (cdr err)))))
1202               ;; the file was somehow created by someone else between
1203               ;; `make-temp-name' and `write-region', let's try again.
1204               nil)
1205             file)
1206         ;; Reset the umask.
1207         (set-default-file-modes umask)))))
1208
1209 (defun mm-image-load-path (&optional package)
1210   (let (dir result)
1211     (dolist (path load-path (nreverse result))
1212       (when (and path
1213                  (file-directory-p
1214                   (setq dir (concat (file-name-directory
1215                                      (directory-file-name path))
1216                                     "etc/images/" (or package "gnus/")))))
1217         (push dir result))
1218       (push path result))))
1219
1220 ;; Fixme: This doesn't look useful where it's used.
1221 (if (fboundp 'detect-coding-region)
1222     (defun mm-detect-coding-region (start end)
1223       "Like `detect-coding-region' except returning the best one."
1224       (let ((coding-systems
1225              (detect-coding-region start end)))
1226         (or (car-safe coding-systems)
1227             coding-systems)))
1228   (defun mm-detect-coding-region (start end)
1229     (let ((point (point)))
1230       (goto-char start)
1231       (skip-chars-forward "\0-\177" end)
1232       (prog1
1233           (if (eq (point) end) 'ascii (mm-guess-charset))
1234         (goto-char point)))))
1235
1236 (if (fboundp 'coding-system-get)
1237     (defun mm-detect-mime-charset-region (start end)
1238       "Detect MIME charset of the text in the region between START and END."
1239       (let ((cs (mm-detect-coding-region start end)))
1240         (or (coding-system-get cs :mime-charset)
1241             (coding-system-get cs 'mime-charset))))
1242   (defun mm-detect-mime-charset-region (start end)
1243     "Detect MIME charset of the text in the region between START and END."
1244     (let ((cs (mm-detect-coding-region start end)))
1245       cs)))
1246
1247 (eval-when-compile
1248   (unless (fboundp 'coding-system-to-mime-charset)
1249     (defalias 'coding-system-to-mime-charset 'ignore)))
1250
1251 (defun mm-coding-system-to-mime-charset (coding-system)
1252   "Return the MIME charset corresponding to CODING-SYSTEM.
1253 To make this function work with XEmacs, the APEL package is required."
1254   (when coding-system
1255     (or (and (fboundp 'coding-system-get)
1256              (or (coding-system-get coding-system :mime-charset)
1257                  (coding-system-get coding-system 'mime-charset)))
1258         (and (featurep 'xemacs)
1259              (or (and (fboundp 'coding-system-to-mime-charset)
1260                       (not (eq (symbol-function 'coding-system-to-mime-charset)
1261                                'ignore)))
1262                  (and (condition-case nil
1263                           (require 'mcharset)
1264                         (error nil))
1265                       (fboundp 'coding-system-to-mime-charset)))
1266              (coding-system-to-mime-charset coding-system)))))
1267
1268 (eval-when-compile
1269   (require 'jka-compr))
1270
1271 (defun mm-decompress-buffer (filename &optional inplace force)
1272   "Decompress buffer's contents, depending on jka-compr.
1273 Only when FORCE is t or `auto-compression-mode' is enabled and FILENAME
1274 agrees with `jka-compr-compression-info-list', decompression is done.
1275 Signal an error if FORCE is neither nil nor t and compressed data are
1276 not decompressed because `auto-compression-mode' is disabled.
1277 If INPLACE is nil, return decompressed data or nil without modifying
1278 the buffer.  Otherwise, replace the buffer's contents with the
1279 decompressed data.  The buffer's multibyteness must be turned off."
1280   (when (and filename
1281              (if force
1282                  (prog1 t (require 'jka-compr))
1283                (and (fboundp 'jka-compr-installed-p)
1284                     (jka-compr-installed-p))))
1285     (let ((info (jka-compr-get-compression-info filename)))
1286       (when info
1287         (unless (or (memq force (list nil t))
1288                     (jka-compr-installed-p))
1289           (error ""))
1290         (let ((prog (jka-compr-info-uncompress-program info))
1291               (args (jka-compr-info-uncompress-args info))
1292               (msg (format "%s %s..."
1293                            (jka-compr-info-uncompress-message info)
1294                            filename))
1295               (err-file (jka-compr-make-temp-name))
1296               (cur (current-buffer))
1297               (coding-system-for-read mm-binary-coding-system)
1298               (coding-system-for-write mm-binary-coding-system)
1299               retval err-msg)
1300           (message "%s" msg)
1301           (mm-with-unibyte-buffer
1302             (insert-buffer-substring cur)
1303             (condition-case err
1304                 (progn
1305                   (unless (memq (apply 'call-process-region
1306                                        (point-min) (point-max)
1307                                        prog t (list t err-file) nil args)
1308                                 jka-compr-acceptable-retval-list)
1309                     (erase-buffer)
1310                     (insert (mapconcat
1311                              'identity
1312                              (delete "" (split-string
1313                                          (prog2
1314                                              (insert-file-contents err-file)
1315                                              (buffer-string)
1316                                            (erase-buffer))))
1317                              " ")
1318                             "\n")
1319                     (setq err-msg
1320                           (format "Error while executing \"%s %s < %s\""
1321                                   prog (mapconcat 'identity args " ")
1322                                   filename)))
1323                   (setq retval (buffer-string)))
1324               (error
1325                (setq err-msg (error-message-string err)))))
1326           (when (file-exists-p err-file)
1327             (ignore-errors (jka-compr-delete-temp-file err-file)))
1328           (when inplace
1329             (unless err-msg
1330               (delete-region (point-min) (point-max))
1331               (insert retval))
1332             (setq retval nil))
1333           (message "%s" (or err-msg (concat msg "done")))
1334           retval)))))
1335
1336 (eval-when-compile
1337   (unless (fboundp 'coding-system-name)
1338     (defalias 'coding-system-name 'ignore))
1339   (unless (fboundp 'find-file-coding-system-for-read-from-filename)
1340     (defalias 'find-file-coding-system-for-read-from-filename 'ignore))
1341   (unless (fboundp 'find-operation-coding-system)
1342     (defalias 'find-operation-coding-system 'ignore)))
1343
1344 (defun mm-find-buffer-file-coding-system (&optional filename)
1345   "Find coding system used to decode the contents of the current buffer.
1346 This function looks for the coding system magic cookie or examines the
1347 coding system specified by `file-coding-system-alist' being associated
1348 with FILENAME which defaults to `buffer-file-name'.  Data compressed by
1349 gzip, bzip2, etc. are allowed."
1350   (unless filename
1351     (setq filename buffer-file-name))
1352   (save-excursion
1353     (let ((decomp (unless ;; No worth to examine charset of tar files.
1354                       (and filename
1355                            (string-match
1356                             "\\.\\(?:tar\\.[^.]+\\|tbz\\|tgz\\)\\'"
1357                             filename))
1358                     (mm-decompress-buffer filename nil t))))
1359       (when decomp
1360         (set-buffer (let (default-enable-multibyte-characters)
1361                       (generate-new-buffer " *temp*")))
1362         (insert decomp)
1363         (setq filename (file-name-sans-extension filename)))
1364       (goto-char (point-min))
1365       (prog1
1366           (cond
1367            ((boundp 'set-auto-coding-function) ;; Emacs
1368             (if filename
1369                 (or (funcall (symbol-value 'set-auto-coding-function)
1370                              filename (- (point-max) (point-min)))
1371                     (car (find-operation-coding-system 'insert-file-contents
1372                                                        filename)))
1373               (let (auto-coding-alist)
1374                 (condition-case nil
1375                     (funcall (symbol-value 'set-auto-coding-function)
1376                              nil (- (point-max) (point-min)))
1377                   (error nil)))))
1378            ((featurep 'file-coding) ;; XEmacs
1379             (let ((case-fold-search t)
1380                   (end (point-at-eol))
1381                   codesys start)
1382               (or
1383                (and (re-search-forward "-\\*-+[\t ]*" end t)
1384                     (progn
1385                       (setq start (match-end 0))
1386                       (re-search-forward "[\t ]*-+\\*-" end t))
1387                     (progn
1388                       (setq end (match-beginning 0))
1389                       (goto-char start)
1390                       (or (looking-at "coding:[\t ]*\\([^\t ;]+\\)")
1391                           (re-search-forward
1392                            "[\t ;]+coding:[\t ]*\\([^\t ;]+\\)"
1393                            end t)))
1394                     (find-coding-system (setq codesys
1395                                               (intern (match-string 1))))
1396                     codesys)
1397                (and (re-search-forward "^[\t ]*;+[\t ]*Local[\t ]+Variables:"
1398                                        nil t)
1399                     (progn
1400                       (setq start (match-end 0))
1401                       (re-search-forward "^[\t ]*;+[\t ]*End:" nil t))
1402                     (progn
1403                       (setq end (match-beginning 0))
1404                       (goto-char start)
1405                       (re-search-forward
1406                        "^[\t ]*;+[\t ]*coding:[\t ]*\\([^\t\n\r ]+\\)"
1407                        end t))
1408                     (find-coding-system (setq codesys
1409                                               (intern (match-string 1))))
1410                     codesys)
1411                (and (progn
1412                       (goto-char (point-min))
1413                       (setq case-fold-search nil)
1414                       (re-search-forward "^;;;coding system: "
1415                                          ;;(+ (point-min) 3000) t))
1416                                          nil t))
1417                     (looking-at "[^\t\n\r ]+")
1418                     (find-coding-system
1419                      (setq codesys (intern (match-string 0))))
1420                     codesys)
1421                (and filename
1422                     (setq codesys
1423                           (find-file-coding-system-for-read-from-filename
1424                            filename))
1425                     (coding-system-name (coding-system-base codesys)))))))
1426         (when decomp
1427           (kill-buffer (current-buffer)))))))
1428
1429 (provide 'mm-util)
1430
1431 ;; arch-tag: 94dc5388-825d-4fd1-bfa5-2100aa351238
1432 ;;; mm-util.el ends here