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

   1 ;;; mm-bodies.el --- Functions for decoding MIME things
   2
   3 ;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   4 ;;   2005 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-and-compile
  30   (or (fboundp  'base64-decode-region)
  31       (require 'base64)))
  32
  33 (eval-when-compile
  34   (defvar mm-uu-decode-function)
  35   (defvar mm-uu-binhex-decode-function))
  36
  37 (require 'mm-util)
  38 (require 'rfc2047)
  39 (require 'mm-encode)
  40
  41 ;; 8bit treatment gets any char except: 0x32 - 0x7f, LF, TAB, BEL,
  42 ;; BS, vertical TAB, form feed, and ^_
  43 ;;
  44 ;; Note that CR is *not* included, as that would allow a non-paired CR
  45 ;; in the body contrary to RFC 2822:
  46 ;;
  47 ;;   - CR and LF MUST only occur together as CRLF; they MUST NOT
  48 ;;     appear independently in the body.
  49
  50 (defvar mm-7bit-chars "\x20-\x7f\n\t\x7\x8\xb\xc\x1f")
  51
  52 (defcustom mm-body-charset-encoding-alist
  53   '((iso-2022-jp . 7bit)
  54     (iso-2022-jp-2 . 7bit)
  55     ;; We MUST encode UTF-16 because it can contain \0's which is
  56     ;; known to break servers.
  57     ;; Note: UTF-16 variants are invalid for text parts [RFC 2781],
  58     ;; so this can't happen :-/.
  59     (utf-16 . base64)
  60     (utf-16be . base64)
  61     (utf-16le . base64))
  62   "Alist of MIME charsets to encodings.
  63 Valid encodings are `7bit', `8bit', `quoted-printable' and `base64'."
  64   :type '(repeat (cons (symbol :tag "charset")
  65                        (choice :tag "encoding"
  66                                (const 7bit)
  67                                (const 8bit)
  68                                (const quoted-printable)
  69                                (const base64))))
  70   :group 'mime)
  71
  72 (defun mm-encode-body (&optional charset)
  73   "Encode a body.
  74 Should be called narrowed to the body that is to be encoded.
  75 If there is more than one non-ASCII MULE charset in the body, then the
  76 list of MULE charsets found is returned.
  77 If CHARSET is non-nil, it is used as the MIME charset to encode the body.
  78 If successful, the MIME charset is returned.
  79 If no encoding was done, nil is returned."
  80   (if (not (mm-multibyte-p))
  81       ;; In the non-Mule case, we search for non-ASCII chars and
  82       ;; return the value of `mail-parse-charset' if any are found.
  83       (or charset
  84           (save-excursion
  85             (goto-char (point-min))
  86             (if (re-search-forward "[^\x0-\x7f]" nil t)
  87                 (or mail-parse-charset
  88                     (message-options-get 'mm-encody-body-charset)
  89                     (message-options-set
  90                      'mm-encody-body-charset
  91                      (mm-read-coding-system "Charset used in the article: ")))
  92               ;; The logic in `mml-generate-mime-1' confirms that it's OK
  93               ;; to return nil here.
  94               nil)))
  95     (save-excursion
  96       (if charset
  97           (progn
  98             (mm-encode-coding-region (point-min) (point-max)
  99                                      (mm-charset-to-coding-system charset))
 100             charset)
 101         (goto-char (point-min))
 102         (let ((charsets (mm-find-mime-charset-region (point-min) (point-max)
 103                                                      mm-hack-charsets)))
 104           (cond
 105            ;; No encoding.
 106            ((null charsets)
 107             nil)
 108            ;; Too many charsets.
 109            ((> (length charsets) 1)
 110             charsets)
 111            ;; We encode.
 112            (t
 113             (prog1
 114                 (setq charset (car charsets))
 115               (mm-encode-coding-region (point-min) (point-max)
 116                                        (mm-charset-to-coding-system charset))))
 117            ))))))
 118
 119 (defun mm-long-lines-p (length)
 120   "Say whether any of the lines in the buffer is longer than LENGTH."
 121   (save-excursion
 122     (goto-char (point-min))
 123     (end-of-line)
 124     (while (and (not (eobp))
 125                 (not (> (current-column) length)))
 126       (forward-line 1)
 127       (end-of-line))
 128     (and (> (current-column) length)
 129          (current-column))))
 130
 131 (defvar message-posting-charset)
 132
 133 (defun mm-body-encoding (charset &optional encoding)
 134   "Do Content-Transfer-Encoding and return the encoding of the current buffer."
 135   (when (stringp encoding)
 136     (setq encoding (intern (downcase encoding))))
 137   (let ((bits (mm-body-7-or-8))
 138         (longp (mm-long-lines-p 1000)))
 139     (require 'message)
 140     (cond
 141      ((and (not longp)
 142            (not (and mm-use-ultra-safe-encoding
 143                      (or (save-excursion (re-search-forward " $" nil t))
 144                          (save-excursion (re-search-forward "^From " nil t)))))
 145            (eq bits '7bit))
 146       bits)
 147      ((and (not mm-use-ultra-safe-encoding)
 148            (not longp)
 149            (not (cdr (assq charset mm-body-charset-encoding-alist)))
 150            (or (eq t (cdr message-posting-charset))
 151                (memq charset (cdr message-posting-charset))
 152                (eq charset mail-parse-charset)))
 153       bits)
 154      (t
 155       (let ((encoding (or encoding
 156                           (cdr (assq charset mm-body-charset-encoding-alist))
 157                           (mm-qp-or-base64))))
 158         (when mm-use-ultra-safe-encoding
 159           (setq encoding (mm-safer-encoding encoding)))
 160         (mm-encode-content-transfer-encoding encoding "text/plain")
 161         encoding)))))
 162
 163 (defun mm-body-7-or-8 ()
 164   "Say whether the body is 7bit or 8bit."
 165   (if (save-excursion
 166         (goto-char (point-min))
 167         (skip-chars-forward mm-7bit-chars)
 168         (eobp))
 169       '7bit
 170     '8bit))
 171
 172 ;;;
 173 ;;; Functions for decoding
 174 ;;;
 175
 176 (eval-when-compile (defvar mm-uu-yenc-decode-function))
 177
 178 (defun mm-decode-content-transfer-encoding (encoding &optional type)
 179   "Decodes buffer encoded with ENCODING, returning success status.
 180 If TYPE is `text/plain' CRLF->LF translation may occur."
 181   (prog1
 182       (condition-case error
 183           (cond
 184            ((eq encoding 'quoted-printable)
 185             (quoted-printable-decode-region (point-min) (point-max))
 186             t)
 187            ((eq encoding 'base64)
 188             (base64-decode-region
 189              (point-min)
 190              ;; Some mailers insert whitespace
 191              ;; junk at the end which
 192              ;; base64-decode-region dislikes.
 193              ;; Also remove possible junk which could
 194              ;; have been added by mailing list software.
 195              (save-excursion
 196                (goto-char (point-min))
 197                (while (re-search-forward "^[\t ]*\r?\n" nil t)
 198                  (delete-region (match-beginning 0) (match-end 0)))
 199                (goto-char (point-max))
 200                (when (re-search-backward "^[A-Za-z0-9+/]+=*[\t ]*$" nil t)
 201                  (forward-line))
 202                (point))))
 203            ((memq encoding '(7bit 8bit binary))
 204             ;; Do nothing.
 205             t)
 206            ((null encoding)
 207             ;; Do nothing.
 208             t)
 209            ((memq encoding '(x-uuencode x-uue))
 210             (require 'mm-uu)
 211             (funcall mm-uu-decode-function (point-min) (point-max))
 212             t)
 213            ((eq encoding 'x-binhex)
 214             (require 'mm-uu)
 215             (funcall mm-uu-binhex-decode-function (point-min) (point-max))
 216             t)
 217            ((eq encoding 'x-yenc)
 218             (require 'mm-uu)
 219             (funcall mm-uu-yenc-decode-function (point-min) (point-max))
 220             )
 221            ((functionp encoding)
 222             (funcall encoding (point-min) (point-max))
 223             t)
 224            (t
 225             (message "Unknown encoding %s; defaulting to 8bit" encoding)))
 226         (error
 227          (message "Error while decoding: %s" error)
 228          nil))
 229     (when (and
 230            (memq encoding '(base64 x-uuencode x-uue x-binhex x-yenc))
 231            (equal type "text/plain"))
 232       (goto-char (point-min))
 233       (while (search-forward "\r\n" nil t)
 234         (replace-match "\n" t t)))))
 235
 236 (defun mm-decode-body (charset &optional encoding type)
 237   "Decode the current article that has been encoded with ENCODING to CHARSET.
 238 ENCODING is a MIME content transfer encoding.
 239 CHARSET is the MIME charset with which to decode the data after transfer
 240 decoding.  If it is nil, default to `mail-parse-charset'."
 241   (when (stringp charset)
 242     (setq charset (intern (downcase charset))))
 243   (when (or (not charset)
 244             (eq 'gnus-all mail-parse-ignored-charsets)
 245             (memq 'gnus-all mail-parse-ignored-charsets)
 246             (memq charset mail-parse-ignored-charsets))
 247     (setq charset mail-parse-charset))
 248   (save-excursion
 249     (when encoding
 250       (mm-decode-content-transfer-encoding encoding type))
 251     (when (featurep 'mule)  ; Fixme: Wrong test for unibyte session.
 252       (let ((coding-system (mm-charset-to-coding-system charset)))
 253         (if (and (not coding-system)
 254                  (listp mail-parse-ignored-charsets)
 255                  (memq 'gnus-unknown mail-parse-ignored-charsets))
 256             (setq coding-system
 257                   (mm-charset-to-coding-system mail-parse-charset)))
 258         (when (and charset coding-system
 259                    ;; buffer-file-coding-system
 260                    ;;Article buffer is nil coding system
 261                    ;;in XEmacs
 262                    (mm-multibyte-p)
 263                    (or (not (eq coding-system 'ascii))
 264                        (setq coding-system mail-parse-charset))
 265                    (not (eq coding-system 'gnus-decoded)))
 266           (mm-decode-coding-region (point-min) (point-max)
 267                                    coding-system))
 268         (setq buffer-file-coding-system
 269               (if (boundp 'last-coding-system-used)
 270                   (symbol-value 'last-coding-system-used)
 271                 coding-system))))))
 272
 273 (defun mm-decode-string (string charset)
 274   "Decode STRING with CHARSET."
 275   (when (stringp charset)
 276     (setq charset (intern (downcase charset))))
 277   (when (or (not charset)
 278             (eq 'gnus-all mail-parse-ignored-charsets)
 279             (memq 'gnus-all mail-parse-ignored-charsets)
 280             (memq charset mail-parse-ignored-charsets))
 281     (setq charset mail-parse-charset))
 282   (or
 283    (when (featurep 'mule)
 284      (let ((coding-system (mm-charset-to-coding-system charset)))
 285        (if (and (not coding-system)
 286                 (listp mail-parse-ignored-charsets)
 287                 (memq 'gnus-unknown mail-parse-ignored-charsets))
 288            (setq coding-system
 289                  (mm-charset-to-coding-system mail-parse-charset)))
 290        (when (and charset coding-system
 291                   (mm-multibyte-p)
 292                   (or (not (eq coding-system 'ascii))
 293                       (setq coding-system mail-parse-charset)))
 294          (mm-decode-coding-string string coding-system))))
 295    string))
 296
 297 (provide 'mm-bodies)
 298
 299 ;;; arch-tag: 41104bb6-4443-4ca9-8d5c-ff87ecf27d8d
 300 ;;; mm-bodies.el ends here