rfc2047.el 39.5 KB
Newer Older
1
;;; rfc2047.el --- functions for encoding and decoding rfc2047 messages
2

3
;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006,
Glenn Morris's avatar
Glenn Morris committed
4
;;   2007, 2008, 2009, 2010  Free Software Foundation, Inc.
Gerd Moellmann's avatar
Gerd Moellmann committed
5 6 7 8 9

;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
;;	MORIOKA Tomohiko <morioka@jaist.ac.jp>
;; This file is part of GNU Emacs.

10
;; GNU Emacs is free software: you can redistribute it and/or modify
Gerd Moellmann's avatar
Gerd Moellmann committed
11
;; it under the terms of the GNU General Public License as published by
12 13
;; the Free Software Foundation, either version 3 of the License, or
;; (at your option) any later version.
Gerd Moellmann's avatar
Gerd Moellmann committed
14 15 16

;; GNU Emacs is distributed in the hope that it will be useful,
;; but WITHOUT ANY WARRANTY; without even the implied warranty of
17
;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
Gerd Moellmann's avatar
Gerd Moellmann committed
18 19 20
;; GNU General Public License for more details.

;; You should have received a copy of the GNU General Public License
21
;; along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.
Gerd Moellmann's avatar
Gerd Moellmann committed
22 23 24

;;; Commentary:

Dave Love's avatar
Dave Love committed
25 26 27
;; RFC 2047 is "MIME (Multipurpose Internet Mail Extensions) Part
;; Three:  Message Header Extensions for Non-ASCII Text".

Gerd Moellmann's avatar
Gerd Moellmann committed
28 29
;;; Code:

30
(eval-when-compile
Dan Nicolaescu's avatar
Dan Nicolaescu committed
31 32
  (require 'cl))
(defvar message-posting-charset)
Gerd Moellmann's avatar
Gerd Moellmann committed
33 34

(require 'mm-util)
35
(require 'ietf-drums)
36
;; Fixme: Avoid this (used for mail-parse-charset) mm dependence on gnus.
Gerd Moellmann's avatar
Gerd Moellmann committed
37
(require 'mail-prsvr)
Miles Bader's avatar
Miles Bader committed
38
(require 'rfc2045) ;; rfc2045-encode-string
Dave Love's avatar
Dave Love committed
39
(autoload 'mm-body-7-or-8 "mm-bodies")
Dave Love's avatar
Dave Love committed
40

Gerd Moellmann's avatar
Gerd Moellmann committed
41
(defvar rfc2047-header-encoding-alist
42 43
  '(("Newsgroups" . nil)
    ("Followup-To" . nil)
Gerd Moellmann's avatar
Gerd Moellmann committed
44
    ("Message-ID" . nil)
45
    ("\\(Resent-\\)?\\(From\\|Cc\\|To\\|Bcc\\|\\(In-\\)?Reply-To\\|Sender\
46
\\|Mail-Followup-To\\|Mail-Copies-To\\|Approved\\)" . address-mime)
Gerd Moellmann's avatar
Gerd Moellmann committed
47 48 49
    (t . mime))
  "*Header/encoding method alist.
The list is traversed sequentially.  The keys can either be
Dave Love's avatar
Dave Love committed
50
header regexps or t.
Gerd Moellmann's avatar
Gerd Moellmann committed
51 52 53 54 55

The values can be:

1) nil, in which case no encoding is done;
2) `mime', in which case the header will be encoded according to RFC2047;
56 57 58 59
3) `address-mime', like `mime', but takes account of the rules for address
   fields (where quoted strings and comments must be treated separately);
4) a charset, in which case it will be encoded as that charset;
5) `default', in which case the field will be encoded as the rest
Gerd Moellmann's avatar
Gerd Moellmann committed
60 61 62 63 64 65 66 67 68 69
   of the article.")

(defvar rfc2047-charset-encoding-alist
  '((us-ascii . nil)
    (iso-8859-1 . Q)
    (iso-8859-2 . Q)
    (iso-8859-3 . Q)
    (iso-8859-4 . Q)
    (iso-8859-5 . B)
    (koi8-r . B)
70 71
    (iso-8859-7 . B)
    (iso-8859-8 . B)
Gerd Moellmann's avatar
Gerd Moellmann committed
72
    (iso-8859-9 . Q)
Dave Love's avatar
Dave Love committed
73 74
    (iso-8859-14 . Q)
    (iso-8859-15 . Q)
Gerd Moellmann's avatar
Gerd Moellmann committed
75 76 77
    (iso-2022-jp . B)
    (iso-2022-kr . B)
    (gb2312 . B)
78 79
    (gbk . B)
    (gb18030 . B)
80 81
    (big5 . B)
    (cn-big5 . B)
Gerd Moellmann's avatar
Gerd Moellmann committed
82 83 84 85
    (cn-gb . B)
    (cn-gb-2312 . B)
    (euc-kr . B)
    (iso-2022-jp-2 . B)
86 87
    (iso-2022-int-1 . B)
    (viscii . Q))
Gerd Moellmann's avatar
Gerd Moellmann committed
88
  "Alist of MIME charsets to RFC2047 encodings.
89 90
Valid encodings are nil, `Q' and `B'.  These indicate binary (no) encoding,
quoted-printable and base64 respectively.")
Gerd Moellmann's avatar
Gerd Moellmann committed
91

92 93 94 95
(defvar rfc2047-encode-function-alist
  '((Q . rfc2047-q-encode-string)
    (B . rfc2047-b-encode-string)
    (nil . identity))
Gerd Moellmann's avatar
Gerd Moellmann committed
96 97
  "Alist of RFC2047 encodings to encoding functions.")

98 99 100
(defvar rfc2047-encode-encoded-words t
  "Whether encoded words should be encoded again.")

Miles Bader's avatar
Miles Bader committed
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134
(defvar rfc2047-allow-irregular-q-encoded-words t
  "*Whether to decode irregular Q-encoded words.")

(eval-and-compile ;; Necessary to hard code them in `rfc2047-decode-region'.
  (defconst rfc2047-encoded-word-regexp
    "=\\?\\([^][\000-\040()<>@,\;:*\\\"/?.=]+\\)\\(?:\\*[^?]+\\)?\\?\
\\(B\\?[+/0-9A-Za-z]*=*\
\\|Q\\?[ ->@-~]*\
\\)\\?="
    "Regexp that matches encoded word."
    ;; The patterns for the B encoding and the Q encoding, i.e. the ones
    ;; beginning with "B" and "Q" respectively, are restricted into only
    ;; the characters that those encodings may generally use.
    )
  (defconst rfc2047-encoded-word-regexp-loose
    "=\\?\\([^][\000-\040()<>@,\;:*\\\"/?.=]+\\)\\(?:\\*[^?]+\\)?\\?\
\\(B\\?[+/0-9A-Za-z]*=*\
\\|Q\\?\\(?:\\?+[ -<>@-~]\\)?\\(?:[ ->@-~]+\\?+[ -<>@-~]\\)*[ ->@-~]*\\?*\
\\)\\?="
    "Regexp that matches encoded word allowing loose Q encoding."
    ;; The pattern for the Q encoding, i.e. the one beginning with "Q",
    ;; is similar to:
    ;; "Q\\?\\(\\?+[^\n=?]\\)?\\([^\n?]+\\?+[^\n=?]\\)*[^\n?]*\\?*"
    ;;      <--------1-------><----------2,3----------><--4--><-5->
    ;; They mean:
    ;; 1. After "Q?", allow "?"s that follow a character other than "=".
    ;; 2. Allow "=" after "Q?"; it isn't regarded as the terminator.
    ;; 3. In the middle of an encoded word, allow "?"s that follow a
    ;;    character other than "=".
    ;; 4. Allow any characters other than "?" in the middle of an
    ;;    encoded word.
    ;; 5. At the end, allow "?"s.
    ))

Gerd Moellmann's avatar
Gerd Moellmann committed
135 136 137 138
;;;
;;; Functions for encoding RFC2047 messages
;;;

139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157
(defun rfc2047-qp-or-base64 ()
  "Return the type with which to encode the buffer.
This is either `base64' or `quoted-printable'."
  (save-excursion
    (let ((limit (min (point-max) (+ 2000 (point-min))))
	  (n8bit 0))
      (goto-char (point-min))
      (skip-chars-forward "\x20-\x7f\r\n\t" limit)
      (while (< (point) limit)
	(incf n8bit)
	(forward-char 1)
	(skip-chars-forward "\x20-\x7f\r\n\t" limit))
      (if (or (< (* 6 n8bit) (- limit (point-min)))
	      ;; Don't base64, say, a short line with a single
	      ;; non-ASCII char when splitting parts by charset.
	      (= n8bit 1))
	  'quoted-printable
	'base64))))

Gerd Moellmann's avatar
Gerd Moellmann committed
158 159 160 161 162 163 164 165
(defun rfc2047-narrow-to-field ()
  "Narrow the buffer to the header on the current line."
  (beginning-of-line)
  (narrow-to-region
   (point)
   (progn
     (forward-line 1)
     (if (re-search-forward "^[^ \n\t]" nil t)
166
	 (point-at-bol)
Gerd Moellmann's avatar
Gerd Moellmann committed
167 168 169
       (point-max))))
  (goto-char (point-min)))

170 171 172 173 174 175
(defun rfc2047-field-value ()
  "Return the value of the field at point."
  (save-excursion
    (save-restriction
      (rfc2047-narrow-to-field)
      (re-search-forward ":[ \t\n]*" nil t)
176
      (buffer-substring-no-properties (point) (point-max)))))
177

Miles Bader's avatar
Miles Bader committed
178 179 180 181
(defun rfc2047-quote-special-characters-in-quoted-strings (&optional
							   encodable-regexp)
  "Quote special characters with `\\'s in quoted strings.
Quoting will not be done in a quoted string if it contains characters
182
matching ENCODABLE-REGEXP or it is within parentheses."
Miles Bader's avatar
Miles Bader committed
183 184
  (goto-char (point-min))
  (let ((tspecials (concat "[" ietf-drums-tspecials "]"))
185
	(start (point))
Miles Bader's avatar
Miles Bader committed
186
	beg end)
Miles Bader's avatar
Miles Bader committed
187
    (with-syntax-table (standard-syntax-table)
188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
      (while (not (eobp))
	(if (ignore-errors
	      (forward-list 1)
	      (eq (char-before) ?\)))
	    (forward-list -1)
	  (goto-char (point-max)))
	(save-restriction
	  (narrow-to-region start (point))
	  (goto-char start)
	  (while (search-forward "\"" nil t)
	    (setq beg (match-beginning 0))
	    (unless (eq (char-before beg) ?\\)
	      (goto-char beg)
	      (setq beg (1+ beg))
	      (condition-case nil
		  (progn
		    (forward-sexp)
		    (setq end (1- (point)))
		    (goto-char beg)
		    (if (and encodable-regexp
			     (re-search-forward encodable-regexp end t))
			(goto-char (1+ end))
		      (save-restriction
			(narrow-to-region beg end)
			(while (re-search-forward tspecials nil 'move)
			  (if (eq (char-before) ?\\)
			      (if (looking-at tspecials) ;; Already quoted.
				  (forward-char)
				(insert "\\"))
			    (goto-char (match-beginning 0))
			    (insert "\\")
			    (forward-char))))
		      (forward-char)))
		(error
		 (goto-char beg)))))
	  (goto-char (point-max)))
	(forward-list 1)
	(setq start (point))))))
Miles Bader's avatar
Miles Bader committed
226

227 228 229 230 231 232
(defvar rfc2047-encoding-type 'address-mime
  "The type of encoding done by `rfc2047-encode-region'.
This should be dynamically bound around calls to
`rfc2047-encode-region' to either `mime' or `address-mime'.  See
`rfc2047-header-encoding-alist', for definitions.")

Gerd Moellmann's avatar
Gerd Moellmann committed
233 234 235 236 237 238 239 240 241 242
(defun rfc2047-encode-message-header ()
  "Encode the message header according to `rfc2047-header-encoding-alist'.
Should be called narrowed to the head of the message."
  (interactive "*")
  (save-excursion
    (goto-char (point-min))
    (let (alist elem method)
      (while (not (eobp))
	(save-restriction
	  (rfc2047-narrow-to-field)
Miles Bader's avatar
Miles Bader committed
243 244 245 246 247 248 249 250
	  (setq method nil
		alist rfc2047-header-encoding-alist)
	  (while (setq elem (pop alist))
	    (when (or (and (stringp (car elem))
			   (looking-at (car elem)))
		      (eq (car elem) t))
	      (setq alist nil
		    method (cdr elem))))
Gerd Moellmann's avatar
Gerd Moellmann committed
251
	  (if (not (rfc2047-encodable-p))
Miles Bader's avatar
Miles Bader committed
252 253 254
	      (prog2
		  (when (eq method 'address-mime)
		    (rfc2047-quote-special-characters-in-quoted-strings))
255 256 257 258 259 260 261 262 263
		  (if (and (eq (mm-body-7-or-8) '8bit)
			   (mm-multibyte-p)
			   (mm-coding-system-p
			    (car message-posting-charset)))
		      ;; 8 bit must be decoded.
		      (mm-encode-coding-region
		       (point-min) (point-max)
		       (mm-charset-to-coding-system
			(car message-posting-charset))))
264
		;; No encoding necessary, but folding is nice
265 266 267 268 269 270 271 272 273
		(when nil
		  (rfc2047-fold-region
		   (save-excursion
		     (goto-char (point-min))
		     (skip-chars-forward "^:")
		     (when (looking-at ": ")
		       (forward-char 2))
		     (point))
		   (point-max))))
Gerd Moellmann's avatar
Gerd Moellmann committed
274
	    ;; We found something that may perhaps be encoded.
275
	    (re-search-forward "^[^:]+: *" nil t)
Gerd Moellmann's avatar
Gerd Moellmann committed
276
	    (cond
277 278
	     ((eq method 'address-mime)
	      (rfc2047-encode-region (point) (point-max)))
Gerd Moellmann's avatar
Gerd Moellmann committed
279
	     ((eq method 'mime)
280
	      (let ((rfc2047-encoding-type 'mime))
281
		(rfc2047-encode-region (point) (point-max))))
Gerd Moellmann's avatar
Gerd Moellmann committed
282 283
	     ((eq method 'default)
	      (if (and (featurep 'mule)
284 285
		       (if (boundp 'enable-multibyte-characters)
			   (default-value 'enable-multibyte-characters))
Gerd Moellmann's avatar
Gerd Moellmann committed
286
		       mail-parse-charset)
287
		  (mm-encode-coding-region (point) (point-max)
Gerd Moellmann's avatar
Gerd Moellmann committed
288
					   mail-parse-charset)))
289 290 291 292
	     ;; We get this when CC'ing messsages to newsgroups with
	     ;; 8-bit names.  The group name mail copy just got
	     ;; unconditionally encoded.  Previously, it would ask
	     ;; whether to encode, which was quite confusing for the
Glenn Morris's avatar
Glenn Morris committed
293
	     ;; user.  If the new behavior is wrong, tell me. I have
294 295 296 297 298 299 300 301 302 303 304 305 306 307 308
	     ;; left the old code commented out below.
	     ;; -- Per Abrahamsen <abraham@dina.kvl.dk> Date: 2001-10-07.
	     ;; Modified by Dave Love, with the commented-out code changed
	     ;; in accordance with changes elsewhere.
	     ((null method)
	      (rfc2047-encode-region (point) (point-max)))
;;; 	     ((null method)
;;; 	      (if (or (message-options-get
;;; 		       'rfc2047-encode-message-header-encode-any)
;;; 		      (message-options-set
;;; 		       'rfc2047-encode-message-header-encode-any
;;; 		       (y-or-n-p
;;; 			"Some texts are not encoded. Encode anyway?")))
;;; 		  (rfc2047-encode-region (point-min) (point-max))
;;; 		(error "Cannot send unencoded text")))
Gerd Moellmann's avatar
Gerd Moellmann committed
309
	     ((mm-coding-system-p method)
310
	      (if (or (and (featurep 'mule)
311 312
			   (if (boundp 'enable-multibyte-characters)
			       (default-value 'enable-multibyte-characters)))
313
		      (featurep 'file-coding))
314
		  (mm-encode-coding-region (point) (point-max) method)))
Gerd Moellmann's avatar
Gerd Moellmann committed
315 316 317 318
	     ;; Hm.
	     (t)))
	  (goto-char (point-max)))))))

319 320 321
;; Fixme: This, and the require below may not be the Right Thing, but
;; should be safe just before release.  -- fx 2001-02-08

Dave Love's avatar
Dave Love committed
322 323 324
(defun rfc2047-encodable-p ()
  "Return non-nil if any characters in current buffer need encoding in headers.
The buffer may be narrowed."
325
  (require 'message)			; for message-posting-charset
Gerd Moellmann's avatar
Gerd Moellmann committed
326
  (let ((charsets
327
	 (mm-find-mime-charset-region (point-min) (point-max))))
328 329 330
    (goto-char (point-min))
    (or (and rfc2047-encode-encoded-words
	     (prog1
Miles Bader's avatar
Miles Bader committed
331
		 (re-search-forward rfc2047-encoded-word-regexp nil t)
332 333 334
	       (goto-char (point-min))))
	(and charsets
	     (not (equal charsets (list (car message-posting-charset))))))))
335 336 337 338 339 340 341 342

;; Use this syntax table when parsing into regions that may need
;; encoding.  Double quotes are string delimiters, backslash is
;; character quoting, and all other RFC 2822 special characters are
;; treated as punctuation so we can use forward-sexp/forward-word to
;; skip to the end of regions appropriately.  Nb. ietf-drums does
;; things differently.
(defconst rfc2047-syntax-table
343 344 345
  ;; (make-char-table 'syntax-table '(2)) only works in Emacs.
  (let ((table (make-syntax-table)))
    ;; The following is done to work for setting all elements of the table
346
    ;; in Emacs 21-23 and XEmacs; it appears to be the cleanest way.
347 348 349 350 351 352 353 354 355
    ;; Play safe and don't assume the form of the word syntax entry --
    ;; copy it from ?a.
    (if (fboundp 'set-char-table-range)	; Emacs
	(funcall (intern "set-char-table-range")
		 table t (aref (standard-syntax-table) ?a))
      (if (fboundp 'put-char-table)
	  (if (fboundp 'get-char-table)	; warning avoidance
	      (put-char-table t (get-char-table ?a (standard-syntax-table))
			      table))))
356 357
    (modify-syntax-entry ?\\ "\\" table)
    (modify-syntax-entry ?\" "\"" table)
358 359
    (modify-syntax-entry ?\( "(" table)
    (modify-syntax-entry ?\) ")" table)
360 361 362 363 364 365 366 367 368
    (modify-syntax-entry ?\< "." table)
    (modify-syntax-entry ?\> "." table)
    (modify-syntax-entry ?\[ "." table)
    (modify-syntax-entry ?\] "." table)
    (modify-syntax-entry ?: "." table)
    (modify-syntax-entry ?\; "." table)
    (modify-syntax-entry ?, "." table)
    (modify-syntax-entry ?@ "." table)
    table))
Gerd Moellmann's avatar
Gerd Moellmann committed
369 370

(defun rfc2047-encode-region (b e)
371 372 373 374 375
  "Encode words in region B to E that need encoding.
By default, the region is treated as containing RFC2822 addresses.
Dynamically bind `rfc2047-encoding-type' to change that."
  (save-restriction
    (narrow-to-region b e)
376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404
    (let ((encodable-regexp (if rfc2047-encode-encoded-words
				"[^\000-\177]+\\|=\\?"
			      "[^\000-\177]+"))
	  start				; start of current token
	  end begin csyntax
	  ;; Whether there's an encoded word before the current token,
	  ;; either immediately or separated by space.
	  last-encoded
	  (orig-text (buffer-substring-no-properties b e)))
      (if (eq 'mime rfc2047-encoding-type)
	  ;; Simple case.  Continuous words in which all those contain
	  ;; non-ASCII characters are encoded collectively.  Encoding
	  ;; ASCII words, including `Re:' used in Subject headers, is
	  ;; avoided for interoperability with non-MIME clients and
	  ;; for making it easy to find keywords.
	  (progn
	    (goto-char (point-min))
	    (while (progn (skip-chars-forward " \t\n")
			  (not (eobp)))
	      (setq start (point))
	      (while (and (looking-at "[ \t\n]*\\([^ \t\n]+\\)")
			  (progn
			    (setq end (match-end 0))
			    (re-search-forward encodable-regexp end t)))
		(goto-char end))
	      (if (> (point) start)
		  (rfc2047-encode start (point))
		(goto-char end))))
	;; `address-mime' case -- take care of quoted words, comments.
Miles Bader's avatar
Miles Bader committed
405
	(rfc2047-quote-special-characters-in-quoted-strings encodable-regexp)
406
	(with-syntax-table rfc2047-syntax-table
407
	  (goto-char (point-min))
408
	  (condition-case err		; in case of unbalanced quotes
409 410 411 412 413
	      ;; Look for rfc2822-style: sequences of atoms, quoted
	      ;; strings, specials, whitespace.  (Specials mustn't be
	      ;; encoded.)
	      (while (not (eobp))
		;; Skip whitespace.
414 415
		(skip-chars-forward " \t\n")
		(setq start (point))
416 417 418
		(cond
		 ((not (char-after)))	; eob
		 ;; else token start
419
		 ((eq ?\" (setq csyntax (char-syntax (char-after))))
420 421 422 423 424
		  ;; Quoted word.
		  (forward-sexp)
		  (setq end (point))
		  ;; Does it need encoding?
		  (goto-char start)
425 426 427 428 429
		  (if (re-search-forward encodable-regexp end 'move)
		      ;; It needs encoding.  Strip the quotes first,
		      ;; since encoded words can't occur in quotes.
		      (progn
			(goto-char end)
430
			(delete-char -1)
431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
			(goto-char start)
			(delete-char 1)
			(when last-encoded
			  ;; There was a preceding quoted word.  We need
			  ;; to include any separating whitespace in this
			  ;; word to avoid it getting lost.
			  (skip-chars-backward " \t")
			  ;; A space is needed between the encoded words.
			  (insert ? )
			  (setq start (point)
				end (1+ end)))
			;; Adjust the end position for the deleted quotes.
			(rfc2047-encode start (- end 2))
			(setq last-encoded t)) ; record that it was encoded
		    (setq last-encoded  nil)))
		 ((eq ?. csyntax)
447 448 449 450
		  ;; Skip other delimiters, but record that they've
		  ;; potentially separated quoted words.
		  (forward-char)
		  (setq last-encoded nil))
451 452 453 454 455 456 457 458 459
		 ((eq ?\) csyntax)
		  (error "Unbalanced parentheses"))
		 ((eq ?\( csyntax)
		  ;; Look for the end of parentheses.
		  (forward-list)
		  ;; Encode text as an unstructured field.
		  (let ((rfc2047-encoding-type 'mime))
		    (rfc2047-encode-region (1+ start) (1- (point))))
		  (skip-chars-forward ")"))
460 461
		 (t		    ; normal token/whitespace sequence
		  ;; Find the end.
462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528
		  ;; Skip one ASCII word, or encode continuous words
		  ;; in which all those contain non-ASCII characters.
		  (setq end nil)
		  (while (not (or end (eobp)))
		    (when (looking-at "[\000-\177]+")
		      (setq begin (point)
			    end (match-end 0))
		      (when (progn
			      (while (and (or (re-search-forward
					       "[ \t\n]\\|\\Sw" end 'move)
					      (setq end nil))
					  (eq ?\\ (char-syntax (char-before))))
				;; Skip backslash-quoted characters.
				(forward-char))
			      end)
			(setq end (match-beginning 0))
			(if rfc2047-encode-encoded-words
			    (progn
			      (goto-char begin)
			      (when (search-forward "=?" end 'move)
				(goto-char (match-beginning 0))
				(setq end nil)))
			  (goto-char end))))
		    ;; Where the value nil of `end' means there may be
		    ;; text to have to be encoded following the point.
		    ;; Otherwise, the point reached to the end of ASCII
		    ;; words separated by whitespace or a special char.
		    (unless end
		      (when (looking-at encodable-regexp)
			(goto-char (setq begin (match-end 0)))
			(while (and (looking-at "[ \t\n]+\\([^ \t\n]+\\)")
				    (setq end (match-end 0))
				    (progn
				      (while (re-search-forward
					      encodable-regexp end t))
				      (< begin (point)))
				    (goto-char begin)
				    (or (not (re-search-forward "\\Sw" end t))
					(progn
					  (goto-char (match-beginning 0))
					  nil)))
			  (goto-char end))
			(when (looking-at "[^ \t\n]+")
			  (setq end (match-end 0))
			  (if (re-search-forward "\\Sw+" end t)
			      ;; There are special characters better
			      ;; to be encoded so that MTAs may parse
			      ;; them safely.
			      (cond ((= end (point)))
				    ((looking-at (concat "\\sw*\\("
							 encodable-regexp
							 "\\)"))
				     (setq end nil))
				    (t
				     (goto-char (1- (match-end 0)))
				     (unless (= (point) (match-beginning 0))
				       ;; Separate encodable text and
				       ;; delimiter.
				       (insert " "))))
			    (goto-char end)
			    (skip-chars-forward " \t\n")
			    (if (and (looking-at "[^ \t\n]+")
				     (string-match encodable-regexp
						   (match-string 0)))
				(setq end nil)
			      (goto-char end)))))))
		  (skip-chars-backward " \t\n")
529 530
		  (setq end (point))
		  (goto-char start)
531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548
		  (if (re-search-forward encodable-regexp end 'move)
		      (progn
			(unless (memq (char-before start) '(nil ?\t ? ))
			  (if (progn
				(goto-char start)
				(skip-chars-backward "^ \t\n")
				(and (looking-at "\\Sw+")
				     (= (match-end 0) start)))
			      ;; Also encode bogus delimiters.
			      (setq start (point))
			    ;; Separate encodable text and delimiter.
			    (goto-char start)
			    (insert " ")
			    (setq start (1+ start)
				  end (1+ end))))
			(rfc2047-encode start end)
			(setq last-encoded t))
		    (setq last-encoded nil)))))
549
	    (error
550 551 552 553 554 555
	     (if (or debug-on-quit debug-on-error)
		 (signal (car err) (cdr err))
	       (error "Invalid data for rfc2047 encoding: %s"
		      (mm-replace-in-string orig-text "[ \t\n]+" " "))))))))
    (rfc2047-fold-region b (point))
    (goto-char (point-max))))
Gerd Moellmann's avatar
Gerd Moellmann committed
556 557

(defun rfc2047-encode-string (string)
558 559
  "Encode words in STRING.
By default, the string is treated as containing addresses (see
560
`rfc2047-encoding-type')."
561
  (mm-with-multibyte-buffer
Gerd Moellmann's avatar
Gerd Moellmann committed
562 563 564 565
    (insert string)
    (rfc2047-encode-region (point-min) (point-max))
    (buffer-string)))

Miles Bader's avatar
Miles Bader committed
566 567 568 569 570 571 572 573
;; From RFC 2047:
;; 2. Syntax of encoded-words
;;    [...]
;;    While there is no limit to the length of a multiple-line header
;;    field, each line of a header field that contains one or more
;;    'encoded-word's is limited to 76 characters.
;;
;; In `rfc2047-encode-parameter' it is bound to nil, so don't defconst it.
574 575
(defvar rfc2047-encode-max-chars 76
  "Maximum characters of each header line that contain encoded-words.
Miles Bader's avatar
Miles Bader committed
576 577 578
According to RFC 2047, it is 76.  If it is nil, encoded-words
will not be folded.  Too small value may cause an error.  You
should not change this value.")
579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652

(defun rfc2047-encode-1 (column string cs encoder start crest tail
				&optional eword)
  "Subroutine used by `rfc2047-encode'."
  (cond ((string-equal string "")
	 (or eword ""))
	((not rfc2047-encode-max-chars)
	 (concat start
		 (funcall encoder (if cs
				      (mm-encode-coding-string string cs)
				    string))
		 "?="))
	((>= column rfc2047-encode-max-chars)
	 (when eword
	   (cond ((string-match "\n[ \t]+\\'" eword)
		  ;; Reomove a superfluous empty line.
		  (setq eword (substring eword 0 (match-beginning 0))))
		 ((string-match "(+\\'" eword)
		  ;; Break the line before the open parenthesis.
		  (setq crest (concat crest (match-string 0 eword))
			eword (substring eword 0 (match-beginning 0))))))
	 (rfc2047-encode-1 (length crest) string cs encoder start " " tail
			   (concat eword "\n" crest)))
	(t
	 (let ((index 0)
	       (limit (1- (length string)))
	       (prev "")
	       next len)
	   (while (and prev
		       (<= index limit))
	     (setq next (concat start
				(funcall encoder
					 (if cs
					     (mm-encode-coding-string
					      (substring string 0 (1+ index))
					      cs)
					   (substring string 0 (1+ index))))
				"?=")
		   len (+ column (length next)))
	     (if (> len rfc2047-encode-max-chars)
		 (setq next prev
		       prev nil)
	       (if (or (< index limit)
		       (<= (+ len (or (string-match "\n" tail)
				      (length tail)))
			   rfc2047-encode-max-chars))
		   (setq prev next
			 index (1+ index))
		 (if (string-match "\\`)+" tail)
		     ;; Break the line after the close parenthesis.
		     (setq tail (concat (substring tail 0 (match-end 0))
					"\n "
					(substring tail (match-end 0)))
			   prev next
			   index (1+ index))
		   (setq next prev
			 prev nil)))))
	   (if (> index limit)
	       (concat eword next tail)
	     (if (= 0 index)
		 (if (and eword
			  (string-match "(+\\'" eword))
		     (setq crest (concat crest (match-string 0 eword))
			   eword (substring eword 0 (match-beginning 0)))
		   (setq eword (concat eword next)))
	       (setq crest " "
		     eword (concat eword next)))
	     (when (string-match "\n[ \t]+\\'" eword)
	       ;; Reomove a superfluous empty line.
	       (setq eword (substring eword 0 (match-beginning 0))))
	     (rfc2047-encode-1 (length crest) (substring string index)
			       cs encoder start " " tail
			       (concat eword "\n" crest)))))))

653 654
(defun rfc2047-encode (b e)
  "Encode the word(s) in the region B to E.
655 656 657
Point moves to the end of the region."
  (let ((mime-charset (or (mm-find-mime-charset-region b e) (list 'us-ascii)))
	cs encoding tail crest eword)
658 659 660
    ;; Use utf-8 as a last resort if determining charset of text fails.
    (if (memq nil mime-charset)
	(setq mime-charset (list 'utf-8)))
661 662 663 664 665 666 667 668 669 670 671 672 673
    (cond ((> (length mime-charset) 1)
	   (error "Can't rfc2047-encode `%s'"
		  (buffer-substring-no-properties b e)))
	  ((= (length mime-charset) 1)
	   (setq mime-charset (car mime-charset)
		 cs (mm-charset-to-coding-system mime-charset))
	   (unless (and (mm-multibyte-p)
			(mm-coding-system-p cs))
	     (setq cs nil))
	   (save-restriction
	     (narrow-to-region b e)
	     (setq encoding
		   (or (cdr (assq mime-charset
Gerd Moellmann's avatar
Gerd Moellmann committed
674
				  rfc2047-charset-encoding-alist))
675 676
		       ;; For the charsets that don't have a preferred
		       ;; encoding, choose the one that's shorter.
677 678 679 680 681 682 683 684 685 686 687
		       (if (eq (rfc2047-qp-or-base64) 'base64)
			   'B
			 'Q)))
	     (widen)
	     (goto-char e)
	     (skip-chars-forward "^ \t\n")
	     ;; `tail' may contain a close parenthesis.
	     (setq tail (buffer-substring-no-properties e (point)))
	     (goto-char b)
	     (setq b (point-marker)
		   e (set-marker (make-marker) e))
688
	     (rfc2047-fold-region (point-at-bol) b)
689 690 691 692 693 694
	     (goto-char b)
	     (skip-chars-backward "^ \t\n")
	     (unless (= 0 (skip-chars-backward " \t"))
	       ;; `crest' may contain whitespace and an open parenthesis.
	       (setq crest (buffer-substring-no-properties (point) b)))
	     (setq eword (rfc2047-encode-1
695
			  (- b (point-at-bol))
696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723
			  (mm-replace-in-string
			   (buffer-substring-no-properties b e)
			   "\n\\([ \t]?\\)" "\\1")
			  cs
			  (or (cdr (assq encoding
					 rfc2047-encode-function-alist))
			      'identity)
			  (concat "=?" (downcase (symbol-name mime-charset))
				  "?" (upcase (symbol-name encoding)) "?")
			  (or crest " ")
			  tail))
	     (delete-region (if (eq (aref eword 0) ?\n)
				(if (bolp)
				    ;; The line was folded before encoding.
				    (1- (point))
				  (point))
			      (goto-char b))
			    (+ e (length tail)))
	     ;; `eword' contains `crest' and `tail'.
	     (insert eword)
	     (set-marker b nil)
	     (set-marker e nil)
	     (unless (or (/= 0 (length tail))
			 (eobp)
			 (looking-at "[ \t\n)]"))
	       (insert " "))))
	  (t
	   (goto-char e)))))
Gerd Moellmann's avatar
Gerd Moellmann committed
724

725 726 727 728 729 730 731
(defun rfc2047-fold-field ()
  "Fold the current header field."
  (save-excursion
    (save-restriction
      (rfc2047-narrow-to-field)
      (rfc2047-fold-region (point-min) (point-max)))))

Gerd Moellmann's avatar
Gerd Moellmann committed
732
(defun rfc2047-fold-region (b e)
733
  "Fold long lines in region B to E."
Gerd Moellmann's avatar
Gerd Moellmann committed
734 735 736
  (save-restriction
    (narrow-to-region b e)
    (goto-char (point-min))
Dave Love's avatar
Dave Love committed
737 738
    (let ((break nil)
	  (qword-break nil)
739
	  (first t)
Dave Love's avatar
Dave Love committed
740 741
	  (bol (save-restriction
		 (widen)
742
		 (point-at-bol))))
Gerd Moellmann's avatar
Gerd Moellmann committed
743
      (while (not (eobp))
744 745
	(when (and (or break qword-break)
		   (> (- (point) bol) 76))
Dave Love's avatar
Dave Love committed
746 747 748
	  (goto-char (or break qword-break))
	  (setq break nil
		qword-break nil)
749
	  (skip-chars-backward " \t")
750
	  (if (looking-at "[ \t]")
751
	      (insert ?\n)
752
	    (insert "\n "))
Dave Love's avatar
Dave Love committed
753 754 755
	  (setq bol (1- (point)))
	  ;; Don't break before the first non-LWSP characters.
	  (skip-chars-forward " \t")
756 757
	  (unless (eobp)
	    (forward-char 1)))
Gerd Moellmann's avatar
Gerd Moellmann committed
758
	(cond
Dave Love's avatar
Dave Love committed
759 760 761 762 763 764 765 766 767 768
	 ((eq (char-after) ?\n)
	  (forward-char 1)
	  (setq bol (point)
		break nil
		qword-break nil)
	  (skip-chars-forward " \t")
	  (unless (or (eobp) (eq (char-after) ?\n))
	    (forward-char 1)))
	 ((eq (char-after) ?\r)
	  (forward-char 1))
Gerd Moellmann's avatar
Gerd Moellmann committed
769
	 ((memq (char-after) '(?  ?\t))
Dave Love's avatar
Dave Love committed
770
	  (skip-chars-forward " \t")
771 772
	  (unless first ;; Don't break just after the header name.
	    (setq break (point))))
Dave Love's avatar
Dave Love committed
773 774 775 776 777
	 ((not break)
	  (if (not (looking-at "=\\?[^=]"))
	      (if (eq (char-after) ?=)
		  (forward-char 1)
		(skip-chars-forward "^ \t\n\r="))
778 779 780
	    ;; Don't break at the start of the field.
	    (unless (= (point) b)
	      (setq qword-break (point)))
Dave Love's avatar
Dave Love committed
781 782
	    (skip-chars-forward "^ \t\n\r")))
	 (t
783 784
	  (skip-chars-forward "^ \t\n\r")))
	(setq first nil))
785 786
      (when (and (or break qword-break)
		 (> (- (point) bol) 76))
Dave Love's avatar
Dave Love committed
787 788 789
	(goto-char (or break qword-break))
	(setq break nil
	      qword-break nil)
790 791 792 793
	(if (or (> 0 (skip-chars-backward " \t"))
		(looking-at "[ \t]"))
	    (insert ?\n)
	  (insert "\n "))
Dave Love's avatar
Dave Love committed
794 795 796
	(setq bol (1- (point)))
	;; Don't break before the first non-LWSP characters.
	(skip-chars-forward " \t")
797 798 799 800 801 802 803 804 805
	(unless (eobp)
	  (forward-char 1))))))

(defun rfc2047-unfold-field ()
  "Fold the current line."
  (save-excursion
    (save-restriction
      (rfc2047-narrow-to-field)
      (rfc2047-unfold-region (point-min) (point-max)))))
Dave Love's avatar
Dave Love committed
806 807

(defun rfc2047-unfold-region (b e)
808
  "Unfold lines in region B to E."
Dave Love's avatar
Dave Love committed
809 810 811 812 813
  (save-restriction
    (narrow-to-region b e)
    (goto-char (point-min))
    (let ((bol (save-restriction
		 (widen)
814 815
		 (point-at-bol)))
	  (eol (point-at-eol)))
Dave Love's avatar
Dave Love committed
816 817
      (forward-line 1)
      (while (not (eobp))
818
	(if (and (looking-at "[ \t]")
819
		 (< (- (point-at-eol) bol) 76))
820 821 822 823
	    (delete-region eol (progn
				 (goto-char eol)
				 (skip-chars-forward "\r\n")
				 (point)))
824 825
	  (setq bol (point-at-bol)))
	(setq eol (point-at-eol))
Dave Love's avatar
Dave Love committed
826
	(forward-line 1)))))
Gerd Moellmann's avatar
Gerd Moellmann committed
827

828 829 830 831
(defun rfc2047-b-encode-string (string)
  "Base64-encode the header contained in STRING."
  (base64-encode-string string t))

832 833
(autoload 'quoted-printable-encode-region "qp")

834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853
(defun rfc2047-q-encode-string (string)
  "Quoted-printable-encode the header in STRING."
  (mm-with-unibyte-buffer
    (insert string)
    (quoted-printable-encode-region
     (point-min) (point-max) nil
     ;; = (\075), _ (\137), ? (\077) are used in the encoded word.
     ;; Avoid using 8bit characters.
     ;; This list excludes `especials' (see the RFC2047 syntax),
     ;; meaning that some characters in non-structured fields will
     ;; get encoded when they con't need to be.  The following is
     ;; what it used to be.
     ;;;  ;; Equivalent to "^\000-\007\011\013\015-\037\200-\377=_?"
     ;;;  "\010\012\014\040-\074\076\100-\136\140-\177")
     "-\b\n\f !#-'*+0-9A-Z\\^`-~\d")
    (subst-char-in-region (point-min) (point-max) ?  ?_)
    (buffer-string)))

(defun rfc2047-encode-parameter (param value)
  "Return and PARAM=VALUE string encoded in the RFC2047-like style.
854 855
This is a substitution for the `rfc2231-encode-string' function, that
is the standard but many mailers don't support it."
Miles Bader's avatar
Miles Bader committed
856 857 858
  (let ((rfc2047-encoding-type 'mime)
	(rfc2047-encode-max-chars nil))
    (rfc2045-encode-string param (rfc2047-encode-string value))))
Gerd Moellmann's avatar
Gerd Moellmann committed
859 860 861 862 863

;;;
;;; Functions for decoding RFC2047 messages
;;;

864 865 866
(defvar rfc2047-quote-decoded-words-containing-tspecials nil
  "If non-nil, quote decoded words containing special characters.")

867 868 869 870 871 872 873 874 875
(defvar rfc2047-allow-incomplete-encoded-text t
  "*Non-nil means allow incomplete encoded-text in successive encoded-words.
Dividing of encoded-text in the place other than character boundaries
violates RFC2047 section 5, while we have a capability to decode it.
If it is non-nil, the decoder will decode B- or Q-encoding in each
encoded-word, concatenate them, and decode it by charset.  Otherwise,
the decoder will fully decode each encoded-word before concatenating
them.")

Miles Bader's avatar
Miles Bader committed
876
(defun rfc2047-strip-backslashes-in-quoted-strings ()
Miles Bader's avatar
Miles Bader committed
877
  "Strip backslashes in quoted strings.  `\\\"' remains."
Miles Bader's avatar
Miles Bader committed
878 879 880 881 882 883 884 885 886 887 888 889 890 891
  (goto-char (point-min))
  (let (beg)
    (with-syntax-table (standard-syntax-table)
      (while (search-forward "\"" nil t)
	(unless (eq (char-before) ?\\)
	  (setq beg (match-end 0))
	  (goto-char (match-beginning 0))
	  (condition-case nil
	      (progn
		(forward-sexp)
		(save-restriction
		  (narrow-to-region beg (1- (point)))
		  (goto-char beg)
		  (while (search-forward "\\" nil 'move)
Miles Bader's avatar
Miles Bader committed
892
		    (unless (memq (char-after) '(?\"))
893
		      (delete-char -1))
Miles Bader's avatar
Miles Bader committed
894 895 896 897 898
		    (forward-char)))
		(forward-char))
	    (error
	     (goto-char beg))))))))

899
(defun rfc2047-charset-to-coding-system (charset &optional allow-override)
900
  "Return coding-system corresponding to MIME CHARSET.
901 902 903 904 905
If your Emacs implementation can't decode CHARSET, return nil.

If allow-override is given, use `mm-charset-override-alist' to
map undesired charset names to their replacement.  This should
only be used for decoding, not for encoding."
906 907 908 909 910 911 912
  (when (stringp charset)
    (setq charset (intern (downcase charset))))
  (when (or (not charset)
	    (eq 'gnus-all mail-parse-ignored-charsets)
	    (memq 'gnus-all mail-parse-ignored-charsets)
	    (memq charset mail-parse-ignored-charsets))
    (setq charset mail-parse-charset))
913
  (let ((cs (mm-charset-to-coding-system charset nil allow-override)))
914 915 916
    (cond ((eq cs 'ascii)
	   (setq cs (or (mm-charset-to-coding-system mail-parse-charset)
			'raw-text)))
917
	  ((mm-coding-system-p cs))
918 919 920 921 922 923 924 925
	  ((and charset
		(listp mail-parse-ignored-charsets)
		(memq 'gnus-unknown mail-parse-ignored-charsets))
	   (setq cs (mm-charset-to-coding-system mail-parse-charset))))
    (if (eq cs 'ascii)
	'raw-text
      cs)))

926 927
(autoload 'quoted-printable-decode-string "qp")

928 929 930 931 932 933 934
(defun rfc2047-decode-encoded-words (words)
  "Decode successive encoded-words in WORDS and return a decoded string.
Each element of WORDS looks like (CHARSET ENCODING ENCODED-TEXT
ENCODED-WORD)."
  (let (word charset cs encoding text rest)
    (while words
      (setq word (pop words))
935
      (if (and (setq cs (rfc2047-charset-to-coding-system
936
			 (setq charset (car word)) t))
937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971
	       (condition-case code
		   (cond ((char-equal ?B (nth 1 word))
			  (setq text (base64-decode-string
				      (rfc2047-pad-base64 (nth 2 word)))))
			 ((char-equal ?Q (nth 1 word))
			  (setq text (quoted-printable-decode-string
				      (mm-subst-char-in-string
				       ?_ ?  (nth 2 word) t)))))
		 (error
		  (message "%s" (error-message-string code))
		  nil)))
	  (if (and rfc2047-allow-incomplete-encoded-text
		   (eq cs (caar rest)))
	      ;; Concatenate text of which the charset is the same.
	      (setcdr (car rest) (concat (cdar rest) text))
	    (push (cons cs text) rest))
	;; Don't decode encoded-word.
	(push (cons nil (nth 3 word)) rest)))
    (while rest
      (setq words (concat
		   (or (and (setq cs (caar rest))
			    (condition-case code
				(mm-decode-coding-string (cdar rest) cs)
			      (error
			       (message "%s" (error-message-string code))
			       nil)))
		       (concat (when (cdr rest) " ")
			       (cdar rest)
			       (when (and words
					  (not (eq (string-to-char words) ? )))
				 " ")))
		   words)
	    rest (cdr rest)))
    words))

972 973 974 975 976 977 978 979 980
;; Fixme: This should decode in place, not cons intermediate strings.
;; Also check whether it needs to worry about delimiting fields like
;; encoding.

;; In fact it's reported that (invalid) encoding of mailboxes in
;; addr-specs is in use, so delimiting fields might help.  Probably
;; not decoding a word which isn't properly delimited is good enough
;; and worthwhile (is it more correct or not?), e.g. something like
;; `=?iso-8859-1?q?foo?=@'.
Gerd Moellmann's avatar
Gerd Moellmann committed
981

Miles Bader's avatar
Miles Bader committed
982 983 984 985
(defun rfc2047-decode-region (start end &optional address-mime)
  "Decode MIME-encoded words in region between START and END.
If ADDRESS-MIME is non-nil, strip backslashes which precede characters
other than `\"' and `\\' in quoted strings."
Gerd Moellmann's avatar
Gerd Moellmann committed
986 987
  (interactive "r")
  (let ((case-fold-search t)
Miles Bader's avatar
Miles Bader committed
988 989 990 991 992 993
	(eword-regexp
	 (if rfc2047-allow-irregular-q-encoded-words
	     (eval-when-compile
	       (concat "[\n\t ]*\\(" rfc2047-encoded-word-regexp-loose "\\)"))
	   (eval-when-compile
	     (concat "[\n\t ]*\\(" rfc2047-encoded-word-regexp "\\)"))))
994
	b e match words)
995 996 997
    (save-excursion
      (save-restriction
	(narrow-to-region start end)
Miles Bader's avatar
Miles Bader committed
998 999
	(when address-mime
	  (rfc2047-strip-backslashes-in-quoted-strings))
1000 1001 1002 1003 1004 1005 1006 1007
	(goto-char (setq b start))
	;; Look for the encoded-words.
	(while (setq match (re-search-forward eword-regexp nil t))
	  (setq e (match-beginning 1)
		end (match-end 0)
		words nil)
	  (while match
	    (push (list (match-string 2) ;; charset
1008
			(char-after (match-beginning 3)) ;; encoding
Miles Bader's avatar
Miles Bader committed
1009
			(substring (match-string 3) 2) ;; encoded-text
1010 1011 1012 1013 1014 1015 1016 1017
			(match-string 1)) ;; encoded-word
		  words)
	    ;; Look for the subsequent encoded-words.
	    (when (setq match (looking-at eword-regexp))
	      (goto-char (setq end (match-end 0)))))
	  ;; Replace the encoded-words with the decoded one.
	  (delete-region e end)
	  (insert (rfc2047-decode-encoded-words (nreverse words)))
1018 1019 1020
	  (save-restriction
	    (narrow-to-region e (point))
	    (goto-char e)
1021 1022
	    ;; Remove newlines between decoded words, though such
	    ;; things essentially must not be there.
1023 1024
	    (while (re-search-forward "[\n\r]+" nil t)
	      (replace-match " "))
1025
	    (setq end (point-max))
1026 1027 1028 1029 1030 1031 1032 1033 1034
	    ;; Quote decoded words if there are special characters
	    ;; which might violate RFC2822.
	    (when (and rfc2047-quote-decoded-words-containing-tspecials
		       (let ((regexp (car (rassq
					   'address-mime
					   rfc2047-header-encoding-alist))))
			 (when regexp
			   (save-restriction
			     (widen)
1035 1036 1037 1038 1039 1040 1041 1042 1043
			     (and
			      ;; Don't quote words if already quoted.
			      (not (and (eq (char-before e) ?\")
					(eq (char-after end) ?\")))
			      (progn
				(beginning-of-line)
				(while (and (memq (char-after) '(?  ?\t))
					    (zerop (forward-line -1))))
				(looking-at regexp)))))))
1044 1045 1046 1047 1048 1049
	      (let (quoted)
		(goto-char e)
		(skip-chars-forward " \t")
		(setq start (point))
		(setq quoted (eq (char-after) ?\"))
		(goto-char (point-max))
1050
		(skip-chars-backward " \t" start)
1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077
		(if (setq quoted (and quoted
				      (> (point) (1+ start))
				      (eq (char-before) ?\")))
		    (progn
		      (backward-char)
		      (setq start (1+ start)
			    end (point-marker)))
		  (setq end (point-marker)))
		(goto-char start)
		(while (search-forward "\"" end t)
		  (when (prog2
			    (backward-char)
			    (zerop (% (skip-chars-backward "\\\\") 2))
			  (goto-char (match-beginning 0)))
		    (insert "\\"))
		  (forward-char))
		(when (and (not quoted)
			   (progn
			     (goto-char start)
			     (re-search-forward
			      (concat "[" ietf-drums-tspecials "]")
			      end t)))
		  (goto-char start)
		  (insert "\"")
		  (goto-char end)
		  (insert "\""))
		(set-marker end nil)))
1078
	    (goto-char (point-max)))
1079 1080
	  (when (and (mm-multibyte-p)
		     mail-parse-charset
1081
		     (not (eq mail-parse-charset 'us-ascii))
1082 1083 1084 1085 1086 1087 1088
		     (not (eq mail-parse-charset 'gnus-decoded)))
	    (mm-decode-coding-region b e mail-parse-charset))
	  (setq b (point)))
	(when (and (mm-multibyte-p)
		   mail-parse-charset
		   (not (eq mail-parse-charset 'us-ascii))
		   (not (eq mail-parse-charset 'gnus-decoded)))
1089
	  (mm-decode-coding-region b (point-max) mail-parse-charset))))))
Gerd Moellmann's avatar
Gerd Moellmann committed
1090

Miles Bader's avatar
Miles Bader committed
1091 1092 1093 1094 1095 1096 1097 1098 1099 1100
(defun rfc2047-decode-address-region (start end)
  "Decode MIME-encoded words in region between START and END.
Backslashes which precede characters other than `\"' and `\\' in quoted
strings are stripped."
  (rfc2047-decode-region start end t))

(defun rfc2047-decode-string (string &optional address-mime)
  "Decode MIME-encoded STRING and return the result.
If ADDRESS-MIME is non-nil, strip backslashes which precede characters
other than `\"' and `\\' in quoted strings."
1101
  ;; (let ((m (mm-multibyte-p)))
1102 1103
    (if (string-match "=\\?" string)
	(with-temp-buffer
1104 1105 1106 1107 1108 1109 1110 1111
          ;; We used to only call mm-enable-multibyte if `m' is non-nil,
          ;; but this can't be the right criterion.  Don't just revert this
          ;; change if it encounters a bug.  Please help me fix it
          ;; right instead.  --Stef
          ;; The string returned should always be multibyte in a multibyte
	  ;; session, i.e. the buffer should be multibyte before
	  ;; `buffer-string' is called.
          (mm-enable-multibyte)
1112 1113
	  (insert string)
	  (inline
Miles Bader's avatar
Miles Bader committed
1114
	    (rfc2047-decode-region (point-min) (point-max) address-mime))
1115
	  (buffer-string))
Miles Bader's avatar
Miles Bader committed
1116 1117 1118 1119 1120 1121 1122 1123
      (when address-mime
	(setq string
	      (with-temp-buffer
		(when (mm-multibyte-string-p string)
		  (mm-enable-multibyte))
		(insert string)
		(rfc2047-strip-backslashes-in-quoted-strings)
		(buffer-string))))
1124
      ;; Fixme: As above, `m' here is inappropriate.
1125
      (if (and ;; m
1126 1127 1128
	       mail-parse-charset
	       (not (eq mail-parse-charset 'us-ascii))
	       (not (eq mail-parse-charset 'gnus-decoded)))
1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140
	  ;; `decode-coding-string' in Emacs offers a third optional
	  ;; arg NOCOPY to avoid consing a new string if the decoding
	  ;; is "trivial".  Unfortunately it currently doesn't
	  ;; consider anything else than a `nil' coding system
	  ;; trivial.
	  ;; `rfc2047-decode-string' is called multiple times for each
	  ;; article during summary buffer generation, and we really
	  ;; want to avoid unnecessary consing.  So we bypass
	  ;; `decode-coding-string' if the string is purely ASCII.
	  (if (and (fboundp 'detect-coding-string)
		   ;; string is purely ASCII
		   (eq (detect-coding-string string t) 'undecided))
1141 1142 1143
              string
            (mm-decode-coding-string string mail-parse-charset))
        (mm-string-to-multibyte string)))) ;; )
Gerd Moellmann's avatar
Gerd Moellmann committed
1144

Miles Bader's avatar
Miles Bader committed
1145 1146 1147 1148 1149 1150
(defun rfc2047-decode-address-string (string)
  "Decode MIME-encoded STRING and return the result.
Backslashes which precede characters other than `\"' and `\\' in quoted
strings are stripped."
  (rfc2047-decode-string string t))

1151 1152 1153 1154 1155
(defun rfc2047-pad-base64 (string)
  "Pad STRING to quartets."
  ;; Be more liberal to accept buggy base64 strings. If
  ;; base64-decode-string accepts buggy strings, this function could
  ;; be aliased to identity.
1156 1157 1158 1159 1160 1161 1162 1163 1164
  (if (= 0 (mod (length string) 4))
      string
    (when (string-match "=+$" string)
      (setq string (substring string 0 (match-beginning 0))))
    (case (mod (length string) 4)
      (0 string)
      (1 string) ;; Error, don't pad it.
      (2 (concat string "=="))
      (3 (concat string "=")))))
1165

Gerd Moellmann's avatar
Gerd Moellmann committed
1166 1167 1168
(provide 'rfc2047)

;;; rfc2047.el ends here