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.
34 (require 'std11-parse)
38 (defgroup eword-decode nil
39 "Encoded-word decoding"
43 ;;; @ MIME encoded-word definition
46 (defconst eword-encoded-text-regexp "[!->@-~]+")
47 (defconst eword-encoded-word-regexp
48 (concat (regexp-quote "=?")
56 eword-encoded-text-regexp
64 (defconst base64-token-regexp "[A-Za-z0-9+/]")
65 (defconst base64-token-padding-regexp "[A-Za-z0-9+/=]")
67 (defconst eword-B-encoded-text-regexp
76 base64-token-padding-regexp
77 base64-token-padding-regexp
80 ;; (defconst eword-B-encoding-and-encoded-text-regexp
81 ;; (concat "\\(B\\)\\?" eword-B-encoded-text-regexp))
84 ;;; @@ Quoted-Printable
87 (defconst eword-Q-encoded-text-regexp
88 (concat "\\([^=?]\\|" quoted-printable-octet-regexp "\\)+"))
89 ;; (defconst eword-Q-encoding-and-encoded-text-regexp
90 ;; (concat "\\(Q\\)\\?" eword-Q-encoded-text-regexp))
96 (defun eword-decode-string (string &optional must-unfold)
97 "Decode MIME encoded-words in STRING.
99 STRING is unfolded before decoding.
101 If an encoded-word is broken or your emacs implementation can not
102 decode the charset included in it, it is not decoded.
104 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
105 if there are in decoded encoded-words (generated by bad manner MUA
106 such as a version of Net$cape)."
107 (setq string (std11-unfold-string string))
108 (let ((dest "")(ew nil)
110 (while (and (string-match eword-encoded-word-regexp string)
111 (setq beg (match-beginning 0)
117 (string-match "^[ \t]+$" (substring string 0 beg))
119 (setq dest (concat dest (substring string 0 beg)))
124 (eword-decode-encoded-word
125 (substring string beg end) must-unfold)
127 (setq string (substring string end))
137 (defun eword-decode-region (start end &optional unfolding must-unfold)
138 "Decode MIME encoded-words in region between START and END.
140 If UNFOLDING is not nil, it unfolds before decoding.
142 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
143 if there are in decoded encoded-words (generated by bad manner MUA
144 such as a version of Net$cape)."
148 (narrow-to-region start end)
150 (eword-decode-unfold)
152 (goto-char (point-min))
153 (while (re-search-forward (concat "\\(" eword-encoded-word-regexp "\\)"
155 "\\(" eword-encoded-word-regexp "\\)")
157 (replace-match "\\1\\6")
158 (goto-char (point-min))
160 (while (re-search-forward eword-encoded-word-regexp nil t)
161 (insert (eword-decode-encoded-word
163 (buffer-substring (match-beginning 0) (match-end 0))
164 (delete-region (match-beginning 0) (match-end 0))
170 ;;; @ for message header
173 (defcustom eword-decode-ignored-field-list
174 '(newsgroups path lines nntp-posting-host received message-id date)
175 "*List of field-names to be ignored when decoding.
176 Each field name must be symbol."
178 :type '(repeat symbol))
180 (defcustom eword-decode-structured-field-list
181 '(reply-to resent-reply-to from resent-from sender resent-sender
182 to resent-to cc resent-cc bcc resent-bcc dcc
183 mime-version content-type content-transfer-encoding
185 "*List of field-names to decode as structured field.
186 Each field name must be symbol."
188 :type '(repeat symbol))
190 (defun eword-decode-header (&optional code-conversion separator)
191 "Decode MIME encoded-words in header fields.
192 If CODE-CONVERSION is nil, it decodes only encoded-words. If it is
193 mime-charset, it decodes non-ASCII bit patterns as the mime-charset.
194 Otherwise it decodes non-ASCII bit patterns as the
195 default-mime-charset.
196 If SEPARATOR is not nil, it is used as header separator."
200 (std11-narrow-to-header separator)
201 (let ((default-charset
203 (if (mime-charset-to-coding-system code-conversion)
205 default-mime-charset))))
207 (let (beg p end field-name len)
208 (goto-char (point-min))
209 (while (re-search-forward std11-field-head-regexp nil t)
210 (setq beg (match-beginning 0)
212 field-name (buffer-substring beg (1- p))
213 len (string-width field-name)
214 field-name (intern (downcase field-name))
215 end (std11-field-end))
216 (cond ((memq field-name eword-decode-ignored-field-list)
219 ((memq field-name eword-decode-structured-field-list)
220 ;; Decode as structured field
221 (let ((body (buffer-substring p end))
222 (default-mime-charset default-charset))
223 (delete-region p end)
224 (insert (eword-decode-and-fold-structured-field
228 ;; Decode as unstructured field
230 (narrow-to-region beg (1+ end))
231 (decode-mime-charset-region p end default-charset)
233 (if (re-search-forward eword-encoded-word-regexp
235 (eword-decode-region beg (point-max) 'unfold))
237 (eword-decode-region (point-min) (point-max) t)
240 (defun eword-decode-unfold ()
241 (goto-char (point-min))
243 (while (re-search-forward std11-field-head-regexp nil t)
244 (setq beg (match-beginning 0)
245 end (std11-field-end))
246 (setq field (buffer-substring beg end))
247 (if (string-match eword-encoded-word-regexp field)
249 (narrow-to-region (goto-char beg) end)
250 (while (re-search-forward "\n\\([ \t]\\)" nil t)
251 (replace-match (match-string 1))
253 (goto-char (point-max))
258 ;;; @ encoded-word decoder
261 (defvar eword-warning-face nil "Face used for invalid encoded-word.")
263 (defun eword-decode-encoded-word (word &optional must-unfold)
264 "Decode WORD if it is an encoded-word.
266 If your emacs implementation can not decode the charset of WORD, it
267 returns WORD. Similarly the encoded-word is broken, it returns WORD.
269 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
270 if there are in decoded encoded-word (generated by bad manner MUA such
271 as a version of Net$cape)."
272 (or (if (string-match eword-encoded-word-regexp word)
274 (substring word (match-beginning 1) (match-end 1))
278 (substring word (match-beginning 2) (match-end 2))
281 (substring word (match-beginning 3) (match-end 3))
284 (eword-decode-encoded-text charset encoding text must-unfold)
287 (add-text-properties 0 (length word)
288 (and eword-warning-face
289 (list 'face eword-warning-face))
296 ;;; @ encoded-text decoder
299 (defun eword-decode-encoded-text (charset encoding string
300 &optional must-unfold)
301 "Decode STRING as an encoded-text.
303 If your emacs implementation can not decode CHARSET, it returns nil.
305 If ENCODING is not \"B\" or \"Q\", it occurs error.
306 So you should write error-handling code if you don't want break by errors.
308 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
309 if there are in decoded encoded-text (generated by bad manner MUA such
310 as a version of Net$cape)."
311 (let ((cs (mime-charset-to-coding-system charset)))
315 ((string-equal "B" encoding)
316 (if (and (string-match eword-B-encoded-text-regexp string)
317 (string-equal string (match-string 0 string)))
318 (base64-decode-string string)
319 (error "Invalid encoded-text %s" string)))
320 ((string-equal "Q" encoding)
321 (if (and (string-match eword-Q-encoded-text-regexp string)
322 (string-equal string (match-string 0 string)))
323 (q-encoding-decode-string string)
324 (error "Invalid encoded-text %s" string)))
326 (error "Invalid encoding %s" encoding)
331 (setq dest (decode-coding-string dest cs))
338 (t (char-to-string chr)))
340 (std11-unfold-string dest)
346 ;;; @ lexical analyze
349 (defvar eword-lexical-analyze-cache nil)
350 (defvar eword-lexical-analyze-cache-max 299
351 "*Max position of eword-lexical-analyze-cache.
352 It is max size of eword-lexical-analyze-cache - 1.")
354 (defcustom eword-lexical-analyzers
355 '(eword-analyze-quoted-string
356 eword-analyze-domain-literal
357 eword-analyze-comment
359 eword-analyze-special
360 eword-analyze-encoded-word
362 "*List of functions to return result of lexical analyze.
363 Each function must have two arguments: STRING and MUST-UNFOLD.
364 STRING is the target string to be analyzed.
365 If MUST-UNFOLD is not nil, each function must unfold and eliminate
366 bare-CR and bare-LF from the result even if they are included in
367 content of the encoded-word.
368 Each function must return nil if it can not analyze STRING as its
371 Previous function is preferred to next function. If a function
372 returns nil, next function is used. Otherwise the return value will
375 :type '(repeat function))
377 (defun eword-analyze-quoted-string (string &optional must-unfold)
378 (let ((p (std11-check-enclosure string ?\" ?\")))
380 (cons (cons 'quoted-string
381 (decode-mime-charset-string
382 (std11-strip-quoted-pair (substring string 1 (1- p)))
383 default-mime-charset))
384 (substring string p))
387 (defun eword-analyze-domain-literal (string &optional must-unfold)
388 (std11-analyze-domain-literal string))
390 (defun eword-analyze-comment (string &optional must-unfold)
391 (let ((p (std11-check-enclosure string ?\( ?\) t)))
395 (decode-mime-charset-string
396 (std11-strip-quoted-pair (substring string 1 (1- p)))
397 default-mime-charset)
399 (substring string p))
402 (defun eword-analyze-spaces (string &optional must-unfold)
403 (std11-analyze-spaces string))
405 (defun eword-analyze-special (string &optional must-unfold)
406 (std11-analyze-special string))
408 (defun eword-analyze-encoded-word (string &optional must-unfold)
409 (if (eq (string-match eword-encoded-word-regexp string) 0)
410 (let ((end (match-end 0))
411 (dest (eword-decode-encoded-word (match-string 0 string)
414 (setq string (substring string end))
415 (while (eq (string-match `,(concat "[ \t\n]*\\("
416 eword-encoded-word-regexp
420 (setq end (match-end 0))
423 (eword-decode-encoded-word (match-string 1 string)
425 string (substring string end))
427 (cons (cons 'atom dest) string)
430 (defun eword-analyze-atom (string &optional must-unfold)
431 (if (string-match std11-atom-regexp string)
432 (let ((end (match-end 0)))
433 (cons (cons 'atom (decode-mime-charset-string
434 (substring string 0 end)
435 default-mime-charset))
436 (substring string end)
439 (defun eword-lexical-analyze-internal (string must-unfold)
441 (while (not (string-equal string ""))
443 (let ((rest eword-lexical-analyzers)
445 (while (and (setq func (car rest))
446 (null (setq r (funcall func string must-unfold)))
448 (setq rest (cdr rest)))
449 (or r `((error . ,string) . ""))
451 (setq dest (cons (car ret) dest))
452 (setq string (cdr ret))
457 (defun eword-lexical-analyze (string &optional must-unfold)
458 "Return lexical analyzed list corresponding STRING.
459 It is like std11-lexical-analyze, but it decodes non us-ascii
460 characters encoded as encoded-words or invalid \"raw\" format.
461 \"Raw\" non us-ascii characters are regarded as variable
462 `default-mime-charset'."
463 (let ((key (copy-sequence string))
465 (set-text-properties 0 (length key) nil key)
466 (if (setq ret (assoc key eword-lexical-analyze-cache))
468 (setq ret (eword-lexical-analyze-internal key must-unfold))
469 (setq eword-lexical-analyze-cache
471 (last eword-lexical-analyze-cache
472 eword-lexical-analyze-cache-max)))
475 (defun eword-decode-token (token)
476 (let ((type (car token))
478 (cond ((eq type 'quoted-string)
479 (std11-wrap-as-quoted-string value))
481 (concat "(" (std11-wrap-as-quoted-pairs value '(?( ?))) ")"))
484 (defun eword-decode-and-fold-structured-field
485 (string start-column &optional max-column must-unfold)
486 "Decode and fold (fill) STRING as structured field body.
487 It decodes non us-ascii characters in FULL-NAME encoded as
488 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
489 characters are regarded as variable `default-mime-charset'.
491 If an encoded-word is broken or your emacs implementation can not
492 decode the charset included in it, it is not decoded.
494 If MAX-COLUMN is omitted, `fill-column' is used.
496 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
497 if there are in decoded encoded-words (generated by bad manner MUA
498 such as a version of Net$cape)."
500 (setq max-column fill-column))
501 (let ((c start-column)
502 (tokens (eword-lexical-analyze string must-unfold))
505 (while (and (setq token (car tokens))
506 (setq tokens (cdr tokens)))
507 (let* ((type (car token)))
508 (if (eq type 'spaces)
509 (let* ((next-token (car tokens))
510 (next-str (eword-decode-token next-token))
511 (next-len (string-width next-str))
512 (next-c (+ c next-len 1)))
513 (if (< next-c max-column)
514 (setq result (concat result " " next-str)
516 (setq result (concat result "\n " next-str)
518 (setq tokens (cdr tokens))
520 (let* ((str (eword-decode-token token)))
521 (setq result (concat result str)
522 c (+ c (string-width str)))
525 (concat result (eword-decode-token token))
528 (defun eword-decode-and-unfold-structured-field (string)
529 "Decode and unfold STRING as structured field body.
530 It decodes non us-ascii characters in FULL-NAME encoded as
531 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
532 characters are regarded as variable `default-mime-charset'.
534 If an encoded-word is broken or your emacs implementation can not
535 decode the charset included in it, it is not decoded."
536 (let ((tokens (eword-lexical-analyze string 'must-unfold))
539 (let* ((token (car tokens))
541 (setq tokens (cdr tokens))
543 (if (eq type 'spaces)
545 (concat result (eword-decode-token token))
549 (defun eword-decode-structured-field-body (string &optional must-unfold
550 start-column max-column)
551 "Decode non us-ascii characters in STRING as structured field body.
552 STRING is unfolded before decoding.
554 It decodes non us-ascii characters in FULL-NAME encoded as
555 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
556 characters are regarded as variable `default-mime-charset'.
558 If an encoded-word is broken or your emacs implementation can not
559 decode the charset included in it, it is not decoded.
561 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
562 if there are in decoded encoded-words (generated by bad manner MUA
563 such as a version of Net$cape)."
565 ;; fold with max-column
566 (eword-decode-and-fold-structured-field
567 string start-column max-column must-unfold)
569 (mapconcat (function eword-decode-token)
570 (eword-lexical-analyze string must-unfold)
574 (defun eword-decode-unstructured-field-body (string &optional must-unfold)
575 "Decode non us-ascii characters in STRING as unstructured field body.
576 STRING is unfolded before decoding.
578 It decodes non us-ascii characters in FULL-NAME encoded as
579 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
580 characters are regarded as variable `default-mime-charset'.
582 If an encoded-word is broken or your emacs implementation can not
583 decode the charset included in it, it is not decoded.
585 If MUST-UNFOLD is non-nil, it unfolds and eliminates line-breaks even
586 if there are in decoded encoded-words (generated by bad manner MUA
587 such as a version of Net$cape)."
589 (decode-mime-charset-string string default-mime-charset)
592 (defun eword-extract-address-components (string)
593 "Extract full name and canonical address from STRING.
594 Returns a list of the form (FULL-NAME CANONICAL-ADDRESS).
595 If no name can be extracted, FULL-NAME will be nil.
596 It decodes non us-ascii characters in FULL-NAME encoded as
597 encoded-words or invalid \"raw\" string. \"Raw\" non us-ascii
598 characters are regarded as variable `default-mime-charset'."
599 (let* ((structure (car (std11-parse-address
600 (eword-lexical-analyze
601 (std11-unfold-string string) 'must-unfold))))
602 (phrase (std11-full-name-string structure))
603 (address (std11-address-string structure))
605 (list phrase address)
612 (provide 'eword-decode)
614 ;;; eword-decode.el ends here