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   (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     (big5 chinese-big5-1 chinese-big5-2)
 478     (tibetan tibetan)
 479     (thai-tis620 thai-tis620)
 480     (windows-1251 cyrillic-iso8859-5)
 481     (iso-2022-7bit ethiopic arabic-1-column arabic-2-column)
 482     (iso-2022-jp-2 latin-iso8859-1 greek-iso8859-7
 483                    latin-jisx0201 japanese-jisx0208-1978
 484                    chinese-gb2312 japanese-jisx0208
 485                    korean-ksc5601 japanese-jisx0212)
 486     (iso-2022-int-1 latin-iso8859-1 greek-iso8859-7
 487                     latin-jisx0201 japanese-jisx0208-1978
 488                     chinese-gb2312 japanese-jisx0208
 489                     korean-ksc5601 japanese-jisx0212
 490                     chinese-cns11643-1 chinese-cns11643-2)
 491     (iso-2022-int-1 latin-iso8859-1 latin-iso8859-2
 492                     cyrillic-iso8859-5 greek-iso8859-7
 493                     latin-jisx0201 japanese-jisx0208-1978
 494                     chinese-gb2312 japanese-jisx0208
 495                     korean-ksc5601 japanese-jisx0212
 496                     chinese-cns11643-1 chinese-cns11643-2
 497                     chinese-cns11643-3 chinese-cns11643-4
 498                     chinese-cns11643-5 chinese-cns11643-6
 499                     chinese-cns11643-7)
 500     (iso-2022-jp-3 latin-jisx0201 japanese-jisx0208-1978 japanese-jisx0208
 501                    japanese-jisx0213-1 japanese-jisx0213-2)
 502     (shift_jis latin-jisx0201 katakana-jisx0201 japanese-jisx0208)
 503     ,(cond ((fboundp 'unicode-precedence-list)
 504             (cons 'utf-8 (delq 'ascii (mapcar 'charset-name
 505                                               (unicode-precedence-list)))))
 506            ((or (not (fboundp 'charsetp)) ;; non-Mule case
 507                 (charsetp 'unicode-a)
 508                 (not (mm-coding-system-p 'mule-utf-8)))
 509             '(utf-8 unicode-a unicode-b unicode-c unicode-d unicode-e))
 510            (t ;; If we have utf-8 we're in Mule 5+.
 511             (append '(utf-8)
 512                     (delete 'ascii
 513                             (coding-system-get 'mule-utf-8 'safe-charsets))))))
 514   "Alist of MIME-charset/MULE-charsets.")
 515
 516 (defun mm-enrich-utf-8-by-mule-ucs ()
 517   "Make the `utf-8' MIME charset usable by the Mule-UCS package.
 518 This function will run when the `un-define' module is loaded under
 519 XEmacs, and fill the `utf-8' entry in `mm-mime-mule-charset-alist'
 520 with Mule charsets.  It is completely useless for Emacs."
 521   (when (boundp 'unicode-basic-translation-charset-order-list)
 522     (condition-case nil
 523         (let ((val (delq
 524                     'ascii
 525                     (copy-sequence
 526                      (symbol-value
 527                       'unicode-basic-translation-charset-order-list))))
 528               (elem (assq 'utf-8 mm-mime-mule-charset-alist)))
 529           (if elem
 530               (setcdr elem val)
 531             (setq mm-mime-mule-charset-alist
 532                   (nconc mm-mime-mule-charset-alist
 533                          (list (cons 'utf-8 val))))))
 534       (error))))
 535
 536 ;; Correct by construction, but should be unnecessary for Emacs:
 537 (if (featurep 'xemacs)
 538     (eval-after-load "un-define" '(mm-enrich-utf-8-by-mule-ucs))
 539   (when (and (fboundp 'coding-system-list)
 540              (fboundp 'sort-coding-systems))
 541     (let ((css (sort-coding-systems (coding-system-list 'base-only)))
 542           cs mime mule alist)
 543       (while css
 544         (setq cs (pop css)
 545               mime (or (coding-system-get cs :mime-charset) ; Emacs 23 (unicode)
 546                        (coding-system-get cs 'mime-charset)))
 547         (when (and mime
 548                    (not (eq t (setq mule
 549                                     (coding-system-get cs 'safe-charsets))))
 550                    (not (assq mime alist)))
 551           (push (cons mime (delq 'ascii mule)) alist)))
 552       (setq mm-mime-mule-charset-alist (nreverse alist)))))
 553
 554 (defvar mm-hack-charsets '(iso-8859-15 iso-2022-jp-2)
 555   "A list of special charsets.
 556 Valid elements include:
 557 `iso-8859-15'    convert ISO-8859-1, -9 to ISO-8859-15 if ISO-8859-15 exists.
 558 `iso-2022-jp-2'  convert ISO-2022-jp to ISO-2022-jp-2 if ISO-2022-jp-2 exists."
 559 )
 560
 561 (defvar mm-iso-8859-15-compatible
 562   '((iso-8859-1 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE")
 563     (iso-8859-9 "\xA4\xA6\xA8\xB4\xB8\xBC\xBD\xBE\xD0\xDD\xDE\xF0\xFD\xFE"))
 564   "ISO-8859-15 exchangeable coding systems and inconvertible characters.")
 565
 566 (defvar mm-iso-8859-x-to-15-table
 567   (and (fboundp 'coding-system-p)
 568        (mm-coding-system-p 'iso-8859-15)
 569        (mapcar
 570         (lambda (cs)
 571           (if (mm-coding-system-p (car cs))
 572               (let ((c (string-to-char
 573                         (decode-coding-string "\341" (car cs)))))
 574                 (cons (char-charset c)
 575                       (cons
 576                        (- (string-to-char
 577                            (decode-coding-string "\341" 'iso-8859-15)) c)
 578                        (string-to-list (decode-coding-string (car (cdr cs))
 579                                                              (car cs))))))
 580             '(gnus-charset 0)))
 581         mm-iso-8859-15-compatible))
 582   "A table of the difference character between ISO-8859-X and ISO-8859-15.")
 583
 584 (defcustom mm-coding-system-priorities
 585   (if (boundp 'current-language-environment)
 586       (let ((lang (symbol-value 'current-language-environment)))
 587         (cond ((string= lang "Japanese")
 588                ;; Japanese users prefer iso-2022-jp to euc-japan or
 589                ;; shift_jis, however iso-8859-1 should be used when
 590                ;; there are only ASCII text and Latin-1 characters.
 591                '(iso-8859-1 iso-2022-jp iso-2022-jp-2 shift_jis utf-8)))))
 592   "Preferred coding systems for encoding outgoing messages.
 593
 594 More than one suitable coding system may be found for some text.
 595 By default, the coding system with the highest priority is used
 596 to encode outgoing messages (see `sort-coding-systems').  If this
 597 variable is set, it overrides the default priority."
 598   :version "21.2"
 599   :type '(repeat (symbol :tag "Coding system"))
 600   :group 'mime)
 601
 602 ;; ??
 603 (defvar mm-use-find-coding-systems-region
 604   (fboundp 'find-coding-systems-region)
 605   "Use `find-coding-systems-region' to find proper coding systems.
 606
 607 Setting it to nil is useful on Emacsen supporting Unicode if sending
 608 mail with multiple parts is preferred to sending a Unicode one.")
 609
 610 ;;; Internal variables:
 611
 612 ;;; Functions:
 613
 614 (defun mm-mule-charset-to-mime-charset (charset)
 615   "Return the MIME charset corresponding to the given Mule CHARSET."
 616   (if (and (fboundp 'find-coding-systems-for-charsets)
 617            (fboundp 'sort-coding-systems))
 618       (let ((css (sort (sort-coding-systems
 619                         (find-coding-systems-for-charsets (list charset)))
 620                        'mm-sort-coding-systems-predicate))
 621             cs mime)
 622         (while (and (not mime)
 623                     css)
 624           (when (setq cs (pop css))
 625             (setq mime (or (coding-system-get cs :mime-charset)
 626                            (coding-system-get cs 'mime-charset)))))
 627         mime)
 628     (let ((alist (mapcar (lambda (cs)
 629                            (assq cs mm-mime-mule-charset-alist))
 630                          (sort (mapcar 'car mm-mime-mule-charset-alist)
 631                                'mm-sort-coding-systems-predicate)))
 632           out)
 633       (while alist
 634         (when (memq charset (cdar alist))
 635           (setq out (caar alist)
 636                 alist nil))
 637         (pop alist))
 638       out)))
 639
 640 (defun mm-charset-to-coding-system (charset &optional lbt
 641                                             allow-override)
 642   "Return coding-system corresponding to CHARSET.
 643 CHARSET is a symbol naming a MIME charset.
 644 If optional argument LBT (`unix', `dos' or `mac') is specified, it is
 645 used as the line break code type of the coding system.
 646
 647 If ALLOW-OVERRIDE is given, use `mm-charset-override-alist' to
 648 map undesired charset names to their replacement.  This should
 649 only be used for decoding, not for encoding."
 650   ;; OVERRIDE is used (only) in `mm-decode-body' and `mm-decode-string'.
 651   (when (stringp charset)
 652     (setq charset (intern (downcase charset))))
 653   (when lbt
 654     (setq charset (intern (format "%s-%s" charset lbt))))
 655   (cond
 656    ((null charset)
 657     charset)
 658    ;; Running in a non-MULE environment.
 659    ((or (null (mm-get-coding-system-list))
 660         (not (fboundp 'coding-system-get)))
 661     charset)
 662    ;; Check override list quite early.  Should only used for decoding, not for
 663    ;; encoding!
 664    ((and allow-override
 665          (let ((cs (cdr (assq charset mm-charset-override-alist))))
 666            (and cs (mm-coding-system-p cs) cs))))
 667    ;; ascii
 668    ((eq charset 'us-ascii)
 669     'ascii)
 670    ;; Check to see whether we can handle this charset.  (This depends
 671    ;; on there being some coding system matching each `mime-charset'
 672    ;; property defined, as there should be.)
 673    ((and (mm-coding-system-p charset)
 674 ;;; Doing this would potentially weed out incorrect charsets.
 675 ;;;      charset
 676 ;;;      (eq charset (coding-system-get charset 'mime-charset))
 677          )
 678     charset)
 679    ;; Eval expressions from `mm-charset-eval-alist'
 680    ((let* ((el (assq charset mm-charset-eval-alist))
 681            (cs (car el))
 682            (form (cdr el)))
 683       (and cs
 684            form
 685            (prog2
 686                ;; Avoid errors...
 687                (condition-case nil (eval form) (error nil))
 688                ;; (message "Failed to eval `%s'" form))
 689                (mm-coding-system-p cs)
 690              (message "Added charset `%s' via `mm-charset-eval-alist'" cs))
 691            cs)))
 692    ;; Translate invalid charsets.
 693    ((let ((cs (cdr (assq charset mm-charset-synonym-alist))))
 694       (and cs
 695            (mm-coding-system-p cs)
 696            ;; (message
 697            ;;  "Using synonym `%s' from `mm-charset-synonym-alist' for `%s'"
 698            ;;  cs charset)
 699            cs)))
 700    ;; Last resort: search the coding system list for entries which
 701    ;; have the right mime-charset in case the canonical name isn't
 702    ;; defined (though it should be).
 703    ((let (cs)
 704       ;; mm-get-coding-system-list returns a list of cs without lbt.
 705       ;; Do we need -lbt?
 706       (dolist (c (mm-get-coding-system-list))
 707         (if (and (null cs)
 708                  (eq charset (or (coding-system-get c :mime-charset)
 709                                  (coding-system-get c 'mime-charset))))
 710             (setq cs c)))
 711       (unless cs
 712         ;; Warn the user about unknown charset:
 713         (if (fboundp 'gnus-message)
 714             (gnus-message 7 "Unknown charset: %s" charset)
 715           (message "Unknown charset: %s" charset)))
 716       cs))))
 717
 718 (defsubst mm-replace-chars-in-string (string from to)
 719   (mm-subst-char-in-string from to string))
 720
 721 (eval-and-compile
 722   (defvar mm-emacs-mule (and (not (featurep 'xemacs))
 723                              (boundp 'default-enable-multibyte-characters)
 724                              default-enable-multibyte-characters
 725                              (fboundp 'set-buffer-multibyte))
 726     "True in Emacs with Mule.")
 727
 728   (if mm-emacs-mule
 729       (defun mm-enable-multibyte ()
 730         "Set the multibyte flag of the current buffer.
 731 Only do this if the default value of `enable-multibyte-characters' is
 732 non-nil.  This is a no-op in XEmacs."
 733         (set-buffer-multibyte 'to))
 734     (defalias 'mm-enable-multibyte 'ignore))
 735
 736   (if mm-emacs-mule
 737       (defun mm-disable-multibyte ()
 738         "Unset the multibyte flag of in the current buffer.
 739 This is a no-op in XEmacs."
 740         (set-buffer-multibyte nil))
 741     (defalias 'mm-disable-multibyte 'ignore)))
 742
 743 (defun mm-preferred-coding-system (charset)
 744   ;; A typo in some Emacs versions.
 745   (or (get-charset-property charset 'preferred-coding-system)
 746       (get-charset-property charset 'prefered-coding-system)))
 747
 748 ;; Mule charsets shouldn't be used.
 749 (defsubst mm-guess-charset ()
 750   "Guess Mule charset from the language environment."
 751   (or
 752    mail-parse-mule-charset ;; cached mule-charset
 753    (progn
 754      (setq mail-parse-mule-charset
 755            (and (boundp 'current-language-environment)
 756                 (car (last
 757                       (assq 'charset
 758                             (assoc current-language-environment
 759                                    language-info-alist))))))
 760      (if (or (not mail-parse-mule-charset)
 761              (eq mail-parse-mule-charset 'ascii))
 762          (setq mail-parse-mule-charset
 763                (or (car (last (assq mail-parse-charset
 764                                     mm-mime-mule-charset-alist)))
 765                    ;; default
 766                    'latin-iso8859-1)))
 767      mail-parse-mule-charset)))
 768
 769 (defun mm-charset-after (&optional pos)
 770   "Return charset of a character in current buffer at position POS.
 771 If POS is nil, it defauls to the current point.
 772 If POS is out of range, the value is nil.
 773 If the charset is `composition', return the actual one."
 774   (let ((char (char-after pos)) charset)
 775     (if (< (mm-char-int char) 128)
 776         (setq charset 'ascii)
 777       ;; charset-after is fake in some Emacsen.
 778       (setq charset (and (fboundp 'char-charset) (char-charset char)))
 779       (if (eq charset 'composition)     ; Mule 4
 780           (let ((p (or pos (point))))
 781             (cadr (find-charset-region p (1+ p))))
 782         (if (and charset (not (memq charset '(ascii eight-bit-control
 783                                                     eight-bit-graphic))))
 784             charset
 785           (mm-guess-charset))))))
 786
 787 (defun mm-mime-charset (charset)
 788   "Return the MIME charset corresponding to the given Mule CHARSET."
 789   (if (eq charset 'unknown)
 790       (error "The message contains non-printable characters, please use attachment"))
 791   (if (and (fboundp 'coding-system-get) (fboundp 'get-charset-property))
 792       ;; This exists in Emacs 20.
 793       (or
 794        (and (mm-preferred-coding-system charset)
 795             (or (coding-system-get
 796                  (mm-preferred-coding-system charset) :mime-charset)
 797                 (coding-system-get
 798                  (mm-preferred-coding-system charset) 'mime-charset)))
 799        (and (eq charset 'ascii)
 800             'us-ascii)
 801        (mm-preferred-coding-system charset)
 802        (mm-mule-charset-to-mime-charset charset))
 803     ;; This is for XEmacs.
 804     (mm-mule-charset-to-mime-charset charset)))
 805
 806 (if (fboundp 'delete-dups)
 807     (defalias 'mm-delete-duplicates 'delete-dups)
 808   (defun mm-delete-duplicates (list)
 809     "Destructively remove `equal' duplicates from LIST.
 810 Store the result in LIST and return it.  LIST must be a proper list.
 811 Of several `equal' occurrences of an element in LIST, the first
 812 one is kept.
 813
 814 This is a compatibility function for Emacsen without `delete-dups'."
 815     ;; Code from `subr.el' in Emacs 22:
 816     (let ((tail list))
 817       (while tail
 818         (setcdr tail (delete (car tail) (cdr tail)))
 819         (setq tail (cdr tail))))
 820     list))
 821
 822 ;; Fixme:  This is used in places when it should be testing the
 823 ;; default multibyteness.  See mm-default-multibyte-p.
 824 (eval-and-compile
 825   (if (and (not (featurep 'xemacs))
 826            (boundp 'enable-multibyte-characters))
 827       (defun mm-multibyte-p ()
 828         "Non-nil if multibyte is enabled in the current buffer."
 829         enable-multibyte-characters)
 830     (defun mm-multibyte-p () (featurep 'mule))))
 831
 832 (defun mm-default-multibyte-p ()
 833   "Return non-nil if the session is multibyte.
 834 This affects whether coding conversion should be attempted generally."
 835   (if (featurep 'mule)
 836       (if (boundp 'default-enable-multibyte-characters)
 837           default-enable-multibyte-characters
 838         t)))
 839
 840 (defun mm-iso-8859-x-to-15-region (&optional b e)
 841   (if (fboundp 'char-charset)
 842       (let (charset item c inconvertible)
 843         (save-restriction
 844           (if e (narrow-to-region b e))
 845           (goto-char (point-min))
 846           (skip-chars-forward "\0-\177")
 847           (while (not (eobp))
 848             (cond
 849              ((not (setq item (assq (char-charset (setq c (char-after)))
 850                                     mm-iso-8859-x-to-15-table)))
 851               (forward-char))
 852              ((memq c (cdr (cdr item)))
 853               (setq inconvertible t)
 854               (forward-char))
 855              (t
 856               (insert-before-markers (prog1 (+ c (car (cdr item)))
 857                                        (delete-char 1)))))
 858             (skip-chars-forward "\0-\177")))
 859         (not inconvertible))))
 860
 861 (defun mm-sort-coding-systems-predicate (a b)
 862   (let ((priorities
 863          (mapcar (lambda (cs)
 864                    ;; Note: invalid entries are dropped silently
 865                    (and (setq cs (mm-coding-system-p cs))
 866                         (coding-system-base cs)))
 867                  mm-coding-system-priorities)))
 868     (and (setq a (mm-coding-system-p a))
 869          (if (setq b (mm-coding-system-p b))
 870              (> (length (memq (coding-system-base a) priorities))
 871                 (length (memq (coding-system-base b) priorities)))
 872            t))))
 873
 874 (eval-when-compile
 875   (autoload 'latin-unity-massage-name "latin-unity")
 876   (autoload 'latin-unity-maybe-remap "latin-unity")
 877   (autoload 'latin-unity-representations-feasible-region "latin-unity")
 878   (autoload 'latin-unity-representations-present-region "latin-unity")
 879   (defvar latin-unity-coding-systems)
 880   (defvar latin-unity-ucs-list))
 881
 882 (defun mm-xemacs-find-mime-charset-1 (begin end)
 883   "Determine which MIME charset to use to send region as message.
 884 This uses the XEmacs-specific latin-unity package to better handle the
 885 case where identical characters from diverse ISO-8859-? character sets
 886 can be encoded using a single one of the corresponding coding systems.
 887
 888 It treats `mm-coding-system-priorities' as the list of preferred
 889 coding systems; a useful example setting for this list in Western
 890 Europe would be '(iso-8859-1 iso-8859-15 utf-8), which would default
 891 to the very standard Latin 1 coding system, and only move to coding
 892 systems that are less supported as is necessary to encode the
 893 characters that exist in the buffer.
 894
 895 Latin Unity doesn't know about those non-ASCII Roman characters that
 896 are available in various East Asian character sets.  As such, its
 897 behavior if you have a JIS 0212 LATIN SMALL LETTER A WITH ACUTE in a
 898 buffer and it can otherwise be encoded as Latin 1, won't be ideal.
 899 But this is very much a corner case, so don't worry about it."
 900   (let ((systems mm-coding-system-priorities) csets psets curset)
 901
 902     ;; Load the Latin Unity library, if available.
 903     (when (and (not (featurep 'latin-unity)) (locate-library "latin-unity"))
 904       (ignore-errors (require 'latin-unity)))
 905
 906     ;; Now, can we use it?
 907     (if (featurep 'latin-unity)
 908         (progn
 909           (setq csets (latin-unity-representations-feasible-region begin end)
 910                 psets (latin-unity-representations-present-region begin end))
 911
 912           (catch 'done
 913
 914             ;; Pass back the first coding system in the preferred list
 915             ;; that can encode the whole region.
 916             (dolist (curset systems)
 917               (setq curset (latin-unity-massage-name 'buffer-default curset))
 918
 919               ;; If the coding system is a universal coding system, then
 920               ;; it can certainly encode all the characters in the region.
 921               (if (memq curset latin-unity-ucs-list)
 922                   (throw 'done (list curset)))
 923
 924               ;; If a coding system isn't universal, and isn't in
 925               ;; the list that latin unity knows about, we can't
 926               ;; decide whether to use it here. Leave that until later
 927               ;; in `mm-find-mime-charset-region' function, whence we
 928               ;; have been called.
 929               (unless (memq curset latin-unity-coding-systems)
 930                 (throw 'done nil))
 931
 932               ;; Right, we know about this coding system, and it may
 933               ;; conceivably be able to encode all the characters in
 934               ;; the region.
 935               (if (latin-unity-maybe-remap begin end curset csets psets t)
 936                   (throw 'done (list curset))))
 937
 938             ;; Can't encode using anything from the
 939             ;; `mm-coding-system-priorities' list.
 940             ;; Leave `mm-find-mime-charset' to do most of the work.
 941             nil))
 942
 943       ;; Right, latin unity isn't available; let `mm-find-charset-region'
 944       ;; take its default action, which equally applies to GNU Emacs.
 945       nil)))
 946
 947 (defmacro mm-xemacs-find-mime-charset (begin end)
 948   (when (featurep 'xemacs)
 949     `(and (featurep 'mule) (mm-xemacs-find-mime-charset-1 ,begin ,end))))
 950
 951 (defun mm-find-mime-charset-region (b e &optional hack-charsets)
 952   "Return the MIME charsets needed to encode the region between B and E.
 953 nil means ASCII, a single-element list represents an appropriate MIME
 954 charset, and a longer list means no appropriate charset."
 955   (let (charsets)
 956     ;; The return possibilities of this function are a mess...
 957     (or (and (mm-multibyte-p)
 958              mm-use-find-coding-systems-region
 959              ;; Find the mime-charset of the most preferred coding
 960              ;; system that has one.
 961              (let ((systems (find-coding-systems-region b e)))
 962                (when mm-coding-system-priorities
 963                  (setq systems
 964                        (sort systems 'mm-sort-coding-systems-predicate)))
 965                (setq systems (delq 'compound-text systems))
 966                (unless (equal systems '(undecided))
 967                  (while systems
 968                    (let* ((head (pop systems))
 969                           (cs (or (coding-system-get head :mime-charset)
 970                                   (coding-system-get head 'mime-charset))))
 971                      ;; The mime-charset (`x-ctext') of
 972                      ;; `compound-text' is not in the IANA list.  We
 973                      ;; shouldn't normally use anything here with a
 974                      ;; mime-charset having an `x-' prefix.
 975                      ;; Fixme:  Allow this to be overridden, since
 976                      ;; there is existing use of x-ctext.
 977                      ;; Also people apparently need the coding system
 978                      ;; `iso-2022-jp-3' (which Mule-UCS defines with
 979                      ;; mime-charset, though it's not valid).
 980                      (if (and cs
 981                               (not (string-match "^[Xx]-" (symbol-name cs)))
 982                               ;; UTF-16 of any variety is invalid for
 983                               ;; text parts and, unfortunately, has
 984                               ;; mime-charset defined both in Mule-UCS
 985                               ;; and versions of Emacs.  (The name
 986                               ;; might be `mule-utf-16...'  or
 987                               ;; `utf-16...'.)
 988                               (not (string-match "utf-16" (symbol-name cs))))
 989                          (setq systems nil
 990                                charsets (list cs))))))
 991                charsets))
 992         ;; If we're XEmacs, and some coding system is appropriate,
 993         ;; mm-xemacs-find-mime-charset will return an appropriate list.
 994         ;; Otherwise, we'll get nil, and the next setq will get invoked.
 995         (setq charsets (mm-xemacs-find-mime-charset b e))
 996
 997         ;; We're not multibyte, or a single coding system won't cover it.
 998         (setq charsets
 999               (mm-delete-duplicates
1000                (mapcar 'mm-mime-charset
1001                        (delq 'ascii
1002                              (mm-find-charset-region b e))))))
1003     (if (and (> (length charsets) 1)
1004              (memq 'iso-8859-15 charsets)
1005              (memq 'iso-8859-15 hack-charsets)
1006              (save-excursion (mm-iso-8859-x-to-15-region b e)))
1007         (mapcar (lambda (x) (setq charsets (delq (car x) charsets)))
1008                 mm-iso-8859-15-compatible))
1009     (if (and (memq 'iso-2022-jp-2 charsets)
1010              (memq 'iso-2022-jp-2 hack-charsets))
1011         (setq charsets (delq 'iso-2022-jp charsets)))
1012     ;; Attempt to reduce the number of charsets if utf-8 is available.
1013     (if (and (featurep 'xemacs)
1014              (> (length charsets) 1)
1015              (mm-coding-system-p 'utf-8))
1016         (let ((mm-coding-system-priorities
1017                (cons 'utf-8 mm-coding-system-priorities)))
1018           (setq charsets
1019                 (mm-delete-duplicates
1020                  (mapcar 'mm-mime-charset
1021                          (delq 'ascii
1022                                (mm-find-charset-region b e)))))))
1023     charsets))
1024
1025 (defmacro mm-with-unibyte-buffer (&rest forms)
1026   "Create a temporary buffer, and evaluate FORMS there like `progn'.
1027 Use unibyte mode for this."
1028   `(let (default-enable-multibyte-characters)
1029      (with-temp-buffer ,@forms)))
1030 (put 'mm-with-unibyte-buffer 'lisp-indent-function 0)
1031 (put 'mm-with-unibyte-buffer 'edebug-form-spec '(body))
1032
1033 (defmacro mm-with-multibyte-buffer (&rest forms)
1034   "Create a temporary buffer, and evaluate FORMS there like `progn'.
1035 Use multibyte mode for this."
1036   `(let ((default-enable-multibyte-characters t))
1037      (with-temp-buffer ,@forms)))
1038 (put 'mm-with-multibyte-buffer 'lisp-indent-function 0)
1039 (put 'mm-with-multibyte-buffer 'edebug-form-spec '(body))
1040
1041 (defmacro mm-with-unibyte-current-buffer (&rest forms)
1042   "Evaluate FORMS with current buffer temporarily made unibyte.
1043 Also bind `default-enable-multibyte-characters' to nil.
1044 Equivalent to `progn' in XEmacs
1045
1046 NOTE: Use this macro with caution in multibyte buffers (it is not
1047 worth using this macro in unibyte buffers of course).  Use of
1048 `(set-buffer-multibyte t)', which is run finally, is generally
1049 harmful since it is likely to modify existing data in the buffer.
1050 For instance, it converts \"\\300\\255\" into \"\\255\" in
1051 Emacs 23 (unicode)."
1052   (let ((multibyte (make-symbol "multibyte"))
1053         (buffer (make-symbol "buffer")))
1054     `(if mm-emacs-mule
1055          (let ((,multibyte enable-multibyte-characters)
1056                (,buffer (current-buffer)))
1057            (unwind-protect
1058                (let (default-enable-multibyte-characters)
1059                  (set-buffer-multibyte nil)
1060                  ,@forms)
1061              (set-buffer ,buffer)
1062              (set-buffer-multibyte ,multibyte)))
1063        (let (default-enable-multibyte-characters)
1064          ,@forms))))
1065 (put 'mm-with-unibyte-current-buffer 'lisp-indent-function 0)
1066 (put 'mm-with-unibyte-current-buffer 'edebug-form-spec '(body))
1067
1068 (defmacro mm-with-unibyte (&rest forms)
1069   "Eval the FORMS with the default value of `enable-multibyte-characters' nil."
1070   `(let (default-enable-multibyte-characters)
1071      ,@forms))
1072 (put 'mm-with-unibyte 'lisp-indent-function 0)
1073 (put 'mm-with-unibyte 'edebug-form-spec '(body))
1074
1075 (defmacro mm-with-multibyte (&rest forms)
1076   "Eval the FORMS with the default value of `enable-multibyte-characters' t."
1077   `(let ((default-enable-multibyte-characters t))
1078      ,@forms))
1079 (put 'mm-with-multibyte 'lisp-indent-function 0)
1080 (put 'mm-with-multibyte 'edebug-form-spec '(body))
1081
1082 (defun mm-find-charset-region (b e)
1083   "Return a list of Emacs charsets in the region B to E."
1084   (cond
1085    ((and (mm-multibyte-p)
1086          (fboundp 'find-charset-region))
1087     ;; Remove composition since the base charsets have been included.
1088     ;; Remove eight-bit-*, treat them as ascii.
1089     (let ((css (find-charset-region b e)))
1090       (mapcar (lambda (cs) (setq css (delq cs css)))
1091               '(composition eight-bit-control eight-bit-graphic
1092                             control-1))
1093       css))
1094    (t
1095     ;; We are in a unibyte buffer or XEmacs non-mule, so we futz around a bit.
1096     (save-excursion
1097       (save-restriction
1098         (narrow-to-region b e)
1099         (goto-char (point-min))
1100         (skip-chars-forward "\0-\177")
1101         (if (eobp)
1102             '(ascii)
1103           (let (charset)
1104             (setq charset
1105                   (and (boundp 'current-language-environment)
1106                        (car (last (assq 'charset
1107                                         (assoc current-language-environment
1108                                                language-info-alist))))))
1109             (if (eq charset 'ascii) (setq charset nil))
1110             (or charset
1111                 (setq charset
1112                       (car (last (assq mail-parse-charset
1113                                        mm-mime-mule-charset-alist)))))
1114             (list 'ascii (or charset 'latin-iso8859-1)))))))))
1115
1116 (if (fboundp 'shell-quote-argument)
1117     (defalias 'mm-quote-arg 'shell-quote-argument)
1118   (defun mm-quote-arg (arg)
1119     "Return a version of ARG that is safe to evaluate in a shell."
1120     (let ((pos 0) new-pos accum)
1121       ;; *** bug: we don't handle newline characters properly
1122       (while (setq new-pos (string-match "[]*[;!'`\"$\\& \t{} |()<>]" arg pos))
1123         (push (substring arg pos new-pos) accum)
1124         (push "\\" accum)
1125         (push (list (aref arg new-pos)) accum)
1126         (setq pos (1+ new-pos)))
1127       (if (= pos 0)
1128           arg
1129         (apply 'concat (nconc (nreverse accum) (list (substring arg pos))))))))
1130
1131 (defun mm-auto-mode-alist ()
1132   "Return an `auto-mode-alist' with only the .gz (etc) thingies."
1133   (let ((alist auto-mode-alist)
1134         out)
1135     (while alist
1136       (when (listp (cdar alist))
1137         (push (car alist) out))
1138       (pop alist))
1139     (nreverse out)))
1140
1141 (defvar mm-inhibit-file-name-handlers
1142   '(jka-compr-handler image-file-handler)
1143   "A list of handlers doing (un)compression (etc) thingies.")
1144
1145 (defun mm-insert-file-contents (filename &optional visit beg end replace
1146                                          inhibit)
1147   "Like `insert-file-contents', but only reads in the file.
1148 A buffer may be modified in several ways after reading into the buffer due
1149 to advanced Emacs features, such as file-name-handlers, format decoding,
1150 `find-file-hooks', etc.
1151 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'.
1152   This function ensures that none of these modifications will take place."
1153   (let* ((format-alist nil)
1154          (auto-mode-alist (if inhibit nil (mm-auto-mode-alist)))
1155          (default-major-mode 'fundamental-mode)
1156          (enable-local-variables nil)
1157          (after-insert-file-functions nil)
1158          (enable-local-eval nil)
1159          (inhibit-file-name-operation (if inhibit
1160                                           'insert-file-contents
1161                                         inhibit-file-name-operation))
1162          (inhibit-file-name-handlers
1163           (if inhibit
1164               (append mm-inhibit-file-name-handlers
1165                       inhibit-file-name-handlers)
1166             inhibit-file-name-handlers))
1167          (ffh (if (boundp 'find-file-hook)
1168                   'find-file-hook
1169                 'find-file-hooks))
1170          (val (symbol-value ffh)))
1171     (set ffh nil)
1172     (unwind-protect
1173         (insert-file-contents filename visit beg end replace)
1174       (set ffh val))))
1175
1176 (defun mm-append-to-file (start end filename &optional codesys inhibit)
1177   "Append the contents of the region to the end of file FILENAME.
1178 When called from a function, expects three arguments,
1179 START, END and FILENAME.  START and END are buffer positions
1180 saying what text to write.
1181 Optional fourth argument specifies the coding system to use when
1182 encoding the file.
1183 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1184   (let ((coding-system-for-write
1185          (or codesys mm-text-coding-system-for-write
1186              mm-text-coding-system))
1187         (inhibit-file-name-operation (if inhibit
1188                                          'append-to-file
1189                                        inhibit-file-name-operation))
1190         (inhibit-file-name-handlers
1191          (if inhibit
1192              (append mm-inhibit-file-name-handlers
1193                      inhibit-file-name-handlers)
1194            inhibit-file-name-handlers)))
1195     (write-region start end filename t 'no-message)
1196     (message "Appended to %s" filename)))
1197
1198 (defun mm-write-region (start end filename &optional append visit lockname
1199                               coding-system inhibit)
1200
1201   "Like `write-region'.
1202 If INHIBIT is non-nil, inhibit `mm-inhibit-file-name-handlers'."
1203   (let ((coding-system-for-write
1204          (or coding-system mm-text-coding-system-for-write
1205              mm-text-coding-system))
1206         (inhibit-file-name-operation (if inhibit
1207                                          'write-region
1208                                        inhibit-file-name-operation))
1209         (inhibit-file-name-handlers
1210          (if inhibit
1211              (append mm-inhibit-file-name-handlers
1212                      inhibit-file-name-handlers)
1213            inhibit-file-name-handlers)))
1214     (write-region start end filename append visit lockname)))
1215
1216 ;; It is not a MIME function, but some MIME functions use it.
1217 (if (and (fboundp 'make-temp-file)
1218          (ignore-errors
1219            (let ((def (symbol-function 'make-temp-file)))
1220              (and (byte-code-function-p def)
1221                   (setq def (if (fboundp 'compiled-function-arglist)
1222                                 ;; XEmacs
1223                                 (eval (list 'compiled-function-arglist def))
1224                               (aref def 0)))
1225                   (>= (length def) 4)
1226                   (eq (nth 3 def) 'suffix)))))
1227     (defalias 'mm-make-temp-file 'make-temp-file)
1228   ;; Stolen (and modified for Emacs 20 and XEmacs) from Emacs 22.
1229   (defun mm-make-temp-file (prefix &optional dir-flag suffix)
1230     "Create a temporary file.
1231 The returned file name (created by appending some random characters at the end
1232 of PREFIX, and expanding against `temporary-file-directory' if necessary),
1233 is guaranteed to point to a newly created empty file.
1234 You can then use `write-region' to write new data into the file.
1235
1236 If DIR-FLAG is non-nil, create a new empty directory instead of a file.
1237
1238 If SUFFIX is non-nil, add that at the end of the file name."
1239     (let ((umask (default-file-modes))
1240           file)
1241       (unwind-protect
1242           (progn
1243             ;; Create temp files with strict access rights.  It's easy to
1244             ;; loosen them later, whereas it's impossible to close the
1245             ;; time-window of loose permissions otherwise.
1246             (set-default-file-modes 448)
1247             (while (condition-case err
1248                        (progn
1249                          (setq file
1250                                (make-temp-name
1251                                 (expand-file-name
1252                                  prefix
1253                                  (if (fboundp 'temp-directory)
1254                                      ;; XEmacs
1255                                      (temp-directory)
1256                                    temporary-file-directory))))
1257                          (if suffix
1258                              (setq file (concat file suffix)))
1259                          (if dir-flag
1260                              (make-directory file)
1261                            ;; NOTE: This is unsafe if Emacs 20
1262                            ;; users and XEmacs users don't use
1263                            ;; a secure temp directory.
1264                            (gmm-write-region "" nil file nil 'silent
1265                                              nil 'excl))
1266                          nil)
1267                      (file-already-exists t)
1268                      ;; The Emacs 20 and XEmacs versions of
1269                      ;; `make-directory' issue `file-error'.
1270                      (file-error (or (and (or (featurep 'xemacs)
1271                                               (= emacs-major-version 20))
1272                                           (file-exists-p file))
1273                                      (signal (car err) (cdr err)))))
1274               ;; the file was somehow created by someone else between
1275               ;; `make-temp-name' and `write-region', let's try again.
1276               nil)
1277             file)
1278         ;; Reset the umask.
1279         (set-default-file-modes umask)))))
1280
1281 (defun mm-image-load-path (&optional package)
1282   (let (dir result)
1283     (dolist (path load-path (nreverse result))
1284       (when (and path
1285                  (file-directory-p
1286                   (setq dir (concat (file-name-directory
1287                                      (directory-file-name path))
1288                                     "etc/images/" (or package "gnus/")))))
1289         (push dir result))
1290       (push path result))))
1291
1292 ;; Fixme: This doesn't look useful where it's used.
1293 (if (fboundp 'detect-coding-region)
1294     (defun mm-detect-coding-region (start end)
1295       "Like `detect-coding-region' except returning the best one."
1296       (let ((coding-systems
1297              (detect-coding-region start end)))
1298         (or (car-safe coding-systems)
1299             coding-systems)))
1300   (defun mm-detect-coding-region (start end)
1301     (let ((point (point)))
1302       (goto-char start)
1303       (skip-chars-forward "\0-\177" end)
1304       (prog1
1305           (if (eq (point) end) 'ascii (mm-guess-charset))
1306         (goto-char point)))))
1307
1308 (if (fboundp 'coding-system-get)
1309     (defun mm-detect-mime-charset-region (start end)
1310       "Detect MIME charset of the text in the region between START and END."
1311       (let ((cs (mm-detect-coding-region start end)))
1312         (or (coding-system-get cs :mime-charset)
1313             (coding-system-get cs 'mime-charset))))
1314   (defun mm-detect-mime-charset-region (start end)
1315     "Detect MIME charset of the text in the region between START and END."
1316     (let ((cs (mm-detect-coding-region start end)))
1317       cs)))
1318
1319
1320 (provide 'mm-util)
1321
1322 ;; arch-tag: 94dc5388-825d-4fd1-bfa5-2100aa351238
1323 ;;; mm-util.el ends here