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