1 ;;; eword-decode.el --- RFC 2047 based encoded-word decoder for GNU Emacs
3 ;; Copyright (C) 1995,1996,1997,1998 Free Software Foundation, Inc.
5 ;; Author: ENAMI Tsugutomo <enami@sys.ptg.sony.co.jp>
6 ;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
7 ;; Maintainer: MORIOKA Tomohiko <morioka@jaist.ac.jp>
9 ;; Original: 1992/07/20 ENAMI Tsugutomo's `mime.el'.
10 ;; Renamed: 1993/06/03 to tiny-mime.el
11 ;; Renamed: 1995/10/03 from tiny-mime.el (split off encoder)
12 ;; Renamed: 1997/02/22 from tm-ew-d.el
13 ;; Keywords: encoded-word, MIME, multilingual, header, mail, news
15 ;; This file is part of SEMI (Spadework for Emacs MIME Interfaces).
17 ;; This program is free software; you can redistribute it and/or
18 ;; modify it under the terms of the GNU General Public License as
19 ;; published by the Free Software Foundation; either version 2, or (at
20 ;; your option) any later version.
22 ;; This program is distributed in the hope that it will be useful, but
23 ;; WITHOUT ANY WARRANTY; without even the implied warranty of
24 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
25 ;; General Public License for more details.
27 ;; You should have received a copy of the GNU General Public License
28 ;; along with GNU Emacs; see the file COPYING. If not, write to the
29 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330,
30 ;; Boston, MA 02111-1307, USA.
38 (defgroup eword-decode nil
39 "Encoded-word decoding"
42 (defcustom eword-max-size-to-decode 1000
43 "*Max size to decode header field."
45 :type '(choice (integer :tag "Limit (bytes)")
46 (const :tag "Don't limit" nil)))
49 ;;; @ MIME encoded-word definition
53 (defconst eword-encoded-text-regexp "[!->@-~]+")
55 (defconst eword-encoded-word-regexp
57 (concat (regexp-quote "=?")
65 eword-encoded-text-regexp
67 (regexp-quote "?="))))
73 (defun eword-decode-string (string &optional must-unfold)
74 "Decode MIME encoded-words in STRING.
76 STRING is unfolded before decoding.
78 If an encoded-word is broken or your emacs implementation can not
79 decode the charset included in it, it is not decoded.
81 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
82 if there are in decoded encoded-words (generated by bad manner MUA
83 such as a version of Net$cape)."
84 (setq string (std11-unfold-string string))
85 (let ((dest "")(ew nil)
87 (while (and (string-match eword-encoded-word-regexp string)
88 (setq beg (match-beginning 0)
94 (string-match "^[ \t]+$" (substring string 0 beg))
96 (setq dest (concat dest (substring string 0 beg)))
101 (eword-decode-encoded-word
102 (substring string beg end) must-unfold)
104 (setq string (substring string end))
110 (defun eword-decode-and-fold-structured-field
111 (string start-column &optional max-column must-unfold)
112 "Decode and fold (fill) STRING as structured field body.
113 It decodes non us-ascii characters in FULL-NAME encoded as
114 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
115 characters are regarded as variable `default-mime-charset'.
117 If an encoded-word is broken or your emacs implementation can not
118 decode the charset included in it, it is not decoded.
120 If MAX-COLUMN is omitted, `fill-column' is used.
122 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
123 if there are in decoded encoded-words (generated by bad manner MUA
124 such as a version of Net$cape)."
125 (if (and eword-max-size-to-decode
126 (> (length string) eword-max-size-to-decode))
129 (setq max-column fill-column))
130 (let ((c start-column)
131 (tokens (eword-lexical-analyze string must-unfold))
134 (while (and (setq token (car tokens))
135 (setq tokens (cdr tokens)))
136 (let* ((type (car token)))
137 (if (eq type 'spaces)
138 (let* ((next-token (car tokens))
139 (next-str (eword-decode-token next-token))
140 (next-len (string-width next-str))
141 (next-c (+ c next-len 1)))
142 (if (< next-c max-column)
143 (setq result (concat result " " next-str)
145 (setq result (concat result "\n " next-str)
147 (setq tokens (cdr tokens))
149 (let* ((str (eword-decode-token token)))
150 (setq result (concat result str)
151 c (+ c (string-width str)))
154 (concat result (eword-decode-token token))
157 (defun eword-decode-and-unfold-structured-field (string)
158 "Decode and unfold STRING as structured field body.
159 It decodes non us-ascii characters in FULL-NAME encoded as
160 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
161 characters are regarded as variable `default-mime-charset'.
163 If an encoded-word is broken or your emacs implementation can not
164 decode the charset included in it, it is not decoded."
165 (let ((tokens (eword-lexical-analyze string 'must-unfold))
168 (let* ((token (car tokens))
170 (setq tokens (cdr tokens))
172 (if (eq type 'spaces)
174 (concat result (eword-decode-token token))
178 (defun eword-decode-structured-field-body (string &optional must-unfold
179 start-column max-column)
180 "Decode non us-ascii characters in STRING as structured field body.
181 STRING is unfolded before decoding.
183 It decodes non us-ascii characters in FULL-NAME encoded as
184 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
185 characters are regarded as variable `default-mime-charset'.
187 If an encoded-word is broken or your emacs implementation can not
188 decode the charset included in it, it is not decoded.
190 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
191 if there are in decoded encoded-words (generated by bad manner MUA
192 such as a version of Net$cape)."
194 ;; fold with max-column
195 (eword-decode-and-fold-structured-field
196 string start-column max-column must-unfold)
198 (mapconcat (function eword-decode-token)
199 (eword-lexical-analyze string must-unfold)
203 (defun eword-decode-unstructured-field-body (string &optional must-unfold)
204 "Decode non us-ascii characters in STRING as unstructured field body.
205 STRING is unfolded before decoding.
207 It decodes non us-ascii characters in FULL-NAME encoded as
208 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
209 characters are regarded as variable `default-mime-charset'.
211 If an encoded-word is broken or your emacs implementation can not
212 decode the charset included in it, it is not decoded.
214 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
215 if there are in decoded encoded-words (generated by bad manner MUA
216 such as a version of Net$cape)."
218 (decode-mime-charset-string string default-mime-charset)
225 (defun eword-decode-region (start end &optional unfolding must-unfold)
226 "Decode MIME encoded-words in region between START and END.
228 If UNFOLDING is not nil, it unfolds before decoding.
230 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
231 if there are in decoded encoded-words (generated by bad manner MUA
232 such as a version of Net$cape)."
236 (narrow-to-region start end)
238 (eword-decode-unfold)
240 (goto-char (point-min))
241 (while (re-search-forward (concat "\\(" eword-encoded-word-regexp "\\)"
243 "\\(" eword-encoded-word-regexp "\\)")
245 (replace-match "\\1\\6")
246 (goto-char (point-min))
248 (while (re-search-forward eword-encoded-word-regexp nil t)
249 (insert (eword-decode-encoded-word
251 (buffer-substring (match-beginning 0) (match-end 0))
252 (delete-region (match-beginning 0) (match-end 0))
258 ;;; @ for message header
261 (defcustom eword-decode-ignored-field-list
262 '(Newsgroups Path Lines Nntp-Posting-Host Received Message-Id Date)
263 "*List of field-names to be ignored when decoding.
264 Each field name must be symbol."
266 :type '(repeat symbol))
268 (defcustom eword-decode-structured-field-list
269 '(Reply-To Resent-Reply-To From Resent-From Sender Resent-Sender
270 To Resent-To Cc Resent-Cc Bcc Resent-Bcc Dcc
272 Mime-Version Content-Type Content-Transfer-Encoding
273 Content-Disposition User-Agent)
274 "*List of field-names to decode as structured field.
275 Each field name must be symbol."
277 :type '(repeat symbol))
279 (defun eword-decode-header (&optional code-conversion separator)
280 "Decode MIME encoded-words in header fields.
281 If CODE-CONVERSION is nil, it decodes only encoded-words. If it is
282 mime-charset, it decodes non-ASCII bit patterns as the mime-charset.
283 Otherwise it decodes non-ASCII bit patterns as the
284 default-mime-charset.
285 If SEPARATOR is not nil, it is used as header separator."
289 (std11-narrow-to-header separator)
290 (let ((default-charset
292 (if (mime-charset-to-coding-system code-conversion)
294 default-mime-charset))))
296 (let (beg p end field-name len)
297 (goto-char (point-min))
298 (while (re-search-forward std11-field-head-regexp nil t)
299 (setq beg (match-beginning 0)
301 field-name (buffer-substring beg (1- p))
302 len (string-width field-name)
303 field-name (intern (capitalize field-name))
304 end (std11-field-end))
305 (cond ((memq field-name eword-decode-ignored-field-list)
308 ((memq field-name eword-decode-structured-field-list)
309 ;; Decode as structured field
310 (let ((body (buffer-substring p end))
311 (default-mime-charset default-charset))
312 (delete-region p end)
313 (insert (eword-decode-and-fold-structured-field
317 ;; Decode as unstructured field
319 (narrow-to-region beg (1+ end))
320 (decode-mime-charset-region p end default-charset)
322 (if (re-search-forward eword-encoded-word-regexp
324 (eword-decode-region beg (point-max) 'unfold))
326 (eword-decode-region (point-min) (point-max) t)
329 (defun eword-decode-unfold ()
330 (goto-char (point-min))
332 (while (re-search-forward std11-field-head-regexp nil t)
333 (setq beg (match-beginning 0)
334 end (std11-field-end))
335 (setq field (buffer-substring beg end))
336 (if (string-match eword-encoded-word-regexp field)
338 (narrow-to-region (goto-char beg) end)
339 (while (re-search-forward "\n\\([ \t]\\)" nil t)
340 (replace-match (match-string 1))
342 (goto-char (point-max))
347 ;;; @ encoded-word decoder
350 (defvar eword-decode-encoded-word-error-handler
351 'eword-decode-encoded-word-default-error-handler)
353 (defvar eword-warning-face nil
354 "Face used for invalid encoded-word.")
356 (defun eword-decode-encoded-word-default-error-handler (word signal)
357 (and (add-text-properties 0 (length word)
358 (and eword-warning-face
359 (list 'face eword-warning-face))
363 (defun eword-decode-encoded-word (word &optional must-unfold)
364 "Decode WORD if it is an encoded-word.
366 If your emacs implementation can not decode the charset of WORD, it
367 returns WORD. Similarly the encoded-word is broken, it returns WORD.
369 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
370 if there are in decoded encoded-word (generated by bad manner MUA such
371 as a version of Net$cape)."
372 (or (if (string-match eword-encoded-word-regexp word)
374 (substring word (match-beginning 1) (match-end 1))
378 (substring word (match-beginning 2) (match-end 2))
381 (substring word (match-beginning 3) (match-end 3))
384 (eword-decode-encoded-text charset encoding text must-unfold)
386 (funcall eword-decode-encoded-word-error-handler word err)
392 ;;; @ encoded-text decoder
395 (defun eword-decode-encoded-text (charset encoding string
396 &optional must-unfold)
397 "Decode STRING as an encoded-text.
399 If your emacs implementation can not decode CHARSET, it returns nil.
401 If ENCODING is not \"B\" or \"Q\", it occurs error.
402 So you should write error-handling code if you don't want break by errors.
404 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
405 if there are in decoded encoded-text (generated by bad manner MUA such
406 as a version of Net$cape)."
407 (let ((cs (mime-charset-to-coding-system charset)))
409 (let ((dest (encoded-text-decode-string string encoding)))
411 (setq dest (decode-mime-charset-string dest charset))
415 (cond ((eq chr ?\n) "")
417 (t (char-to-string chr)))
419 (std11-unfold-string dest)
424 ;;; @ lexical analyze
427 (defvar eword-lexical-analyze-cache nil)
428 (defvar eword-lexical-analyze-cache-max 299
429 "*Max position of eword-lexical-analyze-cache.
430 It is max size of eword-lexical-analyze-cache - 1.")
432 (defcustom eword-lexical-analyzers
433 '(eword-analyze-quoted-string
434 eword-analyze-domain-literal
435 eword-analyze-comment
437 eword-analyze-special
438 eword-analyze-encoded-word
440 "*List of functions to return result of lexical analyze.
441 Each function must have two arguments: STRING and MUST-UNFOLD.
442 STRING is the target string to be analyzed.
443 If MUST-UNFOLD is not nil, each function must unfold and eliminate
444 bare-CR and bare-LF from the result even if they are included in
445 content of the encoded-word.
446 Each function must return nil if it can not analyze STRING as its
449 Previous function is preferred to next function. If a function
450 returns nil, next function is used. Otherwise the return value will
453 :type '(repeat function))
455 (defun eword-analyze-quoted-string (string &optional must-unfold)
456 (let ((p (std11-check-enclosure string ?\" ?\")))
458 (cons (cons 'quoted-string
459 (decode-mime-charset-string
460 (std11-strip-quoted-pair (substring string 1 (1- p)))
461 default-mime-charset))
462 (substring string p))
465 (defun eword-analyze-domain-literal (string &optional must-unfold)
466 (std11-analyze-domain-literal string))
468 (defun eword-analyze-comment (string &optional must-unfold)
469 (let ((p (std11-check-enclosure string ?\( ?\) t)))
473 (decode-mime-charset-string
474 (std11-strip-quoted-pair (substring string 1 (1- p)))
475 default-mime-charset)
477 (substring string p))
480 (defun eword-analyze-spaces (string &optional must-unfold)
481 (std11-analyze-spaces string))
483 (defun eword-analyze-special (string &optional must-unfold)
484 (std11-analyze-special string))
486 (defun eword-analyze-encoded-word (string &optional must-unfold)
487 (if (eq (string-match eword-encoded-word-regexp string) 0)
488 (let ((end (match-end 0))
489 (dest (eword-decode-encoded-word (match-string 0 string)
492 (setq string (substring string end))
493 (while (eq (string-match `,(concat "[ \t\n]*\\("
494 eword-encoded-word-regexp
498 (setq end (match-end 0))
501 (eword-decode-encoded-word (match-string 1 string)
503 string (substring string end))
505 (cons (cons 'atom dest) string)
508 (defun eword-analyze-atom (string &optional must-unfold)
509 (if (string-match std11-atom-regexp string)
510 (let ((end (match-end 0)))
511 (cons (cons 'atom (decode-mime-charset-string
512 (substring string 0 end)
513 default-mime-charset))
514 (substring string end)
517 (defun eword-lexical-analyze-internal (string must-unfold)
519 (while (not (string-equal string ""))
521 (let ((rest eword-lexical-analyzers)
523 (while (and (setq func (car rest))
524 (null (setq r (funcall func string must-unfold)))
526 (setq rest (cdr rest)))
527 (or r `((error . ,string) . ""))
529 (setq dest (cons (car ret) dest))
530 (setq string (cdr ret))
535 (defun eword-lexical-analyze (string &optional must-unfold)
536 "Return lexical analyzed list corresponding STRING.
537 It is like std11-lexical-analyze, but it decodes non us-ascii
538 characters encoded as encoded-words or invalid \"raw\" format.
539 \"Raw\" non us-ascii characters are regarded as variable
540 `default-mime-charset'."
541 (let ((key (copy-sequence string))
543 (set-text-properties 0 (length key) nil key)
544 (if (setq ret (assoc key eword-lexical-analyze-cache))
546 (setq ret (eword-lexical-analyze-internal key must-unfold))
547 (setq eword-lexical-analyze-cache
549 (last eword-lexical-analyze-cache
550 eword-lexical-analyze-cache-max)))
553 (defun eword-decode-token (token)
554 (let ((type (car token))
556 (cond ((eq type 'quoted-string)
557 (std11-wrap-as-quoted-string value))
559 (concat "(" (std11-wrap-as-quoted-pairs value '(?( ?))) ")"))
562 (defun eword-extract-address-components (string)
563 "Extract full name and canonical address from STRING.
564 Returns a list of the form (FULL-NAME CANONICAL-ADDRESS).
565 If no name can be extracted, FULL-NAME will be nil.
566 It decodes non us-ascii characters in FULL-NAME encoded as
567 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
568 characters are regarded as variable `default-mime-charset'."
569 (let* ((structure (car (std11-parse-address
570 (eword-lexical-analyze
571 (std11-unfold-string string) 'must-unfold))))
572 (phrase (std11-full-name-string structure))
573 (address (std11-address-string structure))
575 (list phrase address)
582 (provide 'eword-decode)
584 ;;; eword-decode.el ends here