| blob | 4f63cae9eec6e9d098d4ed034930ed81c038b4ab |
1 ;;; rfc2047.el --- functions for encoding and decoding rfc2047 messages
3 ;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
4 ;; 2005, 2006, 2007 Free Software Foundation, Inc.
6 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
7 ;; MORIOKA Tomohiko <morioka@jaist.ac.jp>
8 ;; This file is part of GNU Emacs.
10 ;; GNU Emacs is free software; you can redistribute it and/or modify
11 ;; it under the terms of the GNU General Public License as published by
12 ;; the Free Software Foundation; either version 3, or (at your option)
13 ;; any later version.
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.
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.
25 ;;; Commentary:
27 ;; RFC 2047 is "MIME (Multipurpose Internet Mail Extensions) Part
28 ;; Three: Message Header Extensions for Non-ASCII Text".
30 ;;; Code:
39 ;; Fixme: Avoid this (used for mail-parse-charset) mm dependence on gnus.
51 "*Header/encoding method alist.
52 The list is traversed sequentially. The keys can either be
53 header regexps or t.
55 The values can be:
57 1) nil, in which case no encoding is done;
58 2) `mime', in which case the header will be encoded according to RFC2047;
59 3) `address-mime', like `mime', but takes account of the rules for address
60 fields (where quoted strings and comments must be treated separately);
61 4) a charset, in which case it will be encoded as that charset;
62 5) `default', in which case the field will be encoded as the rest
89 "Alist of MIME charsets to RFC2047 encodings.
90 Valid encodings are nil, `Q' and `B'. These indicate binary (no) encoding,
102 ;;;
103 ;;; Functions for encoding RFC2047 messages
104 ;;;
107 "Return the type with which to encode the buffer.
108 This is either `base64' or `quoted-printable'."
119 ;; Don't base64, say, a short line with a single
120 ;; non-ASCII char when splitting parts by charset.
122 'quoted-printable
126 "Narrow the buffer to the header on the current line."
138 "Return the value of the field at point."
146 encodable-regexp)
148 Quoting will not be done in a quoted string if it contains characters
149 matching ENCODABLE-REGEXP or it is within parentheses."
153 beg end)
195 "The type of encoding done by `rfc2047-encode-region'.
196 This should be dynamically bound around calls to
197 `rfc2047-encode-region' to either `mime' or `address-mime'. See
201 "Encode the message header according to `rfc2047-header-encoding-alist'.
202 Should be called narrowed to the head of the message."
211 alist rfc2047-header-encoding-alist)
226 ;; 8 bit must be decoded.
231 ;; No encoding necessary, but folding is nice
241 ;; We found something that may perhaps be encoded.
252 default-enable-multibyte-characters)
253 mail-parse-charset)
255 mail-parse-charset)))
256 ;; We get this when CC'ing messsages to newsgroups with
257 ;; 8-bit names. The group name mail copy just got
258 ;; unconditionally encoded. Previously, it would ask
259 ;; whether to encode, which was quite confusing for the
260 ;; user. If the new behaviour is wrong, tell me. I have
261 ;; left the old code commented out below.
262 ;; -- Per Abrahamsen <abraham@dina.kvl.dk> Date: 2001-10-07.
263 ;; Modified by Dave Love, with the commented-out code changed
264 ;; in accordance with changes elsewhere.
267 ;;; ((null method)
268 ;;; (if (or (message-options-get
269 ;;; 'rfc2047-encode-message-header-encode-any)
270 ;;; (message-options-set
271 ;;; 'rfc2047-encode-message-header-encode-any
272 ;;; (y-or-n-p
273 ;;; "Some texts are not encoded. Encode anyway?")))
274 ;;; (rfc2047-encode-region (point-min) (point-max))
275 ;;; (error "Cannot send unencoded text")))
279 default-enable-multibyte-characters))
282 ;; Hm.
286 ;; Fixme: This, and the require below may not be the Right Thing, but
287 ;; should be safe just before release. -- fx 2001-02-08
291 "Return non-nil if any characters in current buffer need encoding in headers.
292 The buffer may be narrowed."
304 ;; Use this syntax table when parsing into regions that may need
305 ;; encoding. Double quotes are string delimiters, backslash is
306 ;; character quoting, and all other RFC 2822 special characters are
307 ;; treated as punctuation so we can use forward-sexp/forward-word to
308 ;; skip to the end of regions appropriately. Nb. ietf-drums does
309 ;; things differently.
311 ;; (make-char-table 'syntax-table '(2)) only works in Emacs.
313 ;; The following is done to work for setting all elements of the table
314 ;; in Emacs 21 and 22 and XEmacs; it appears to be the cleanest way.
315 ;; Play safe and don't assume the form of the word syntax entry --
316 ;; copy it from ?a.
323 table))))
336 table))
339 "Encode words in region B to E that need encoding.
340 By default, the region is treated as containing RFC2822 addresses.
341 Dynamically bind `rfc2047-encoding-type' to change that."
347 start ; start of current token
348 end begin csyntax
349 ;; Whether there's an encoded word before the current token,
350 ;; either immediately or separated by space.
351 last-encoded
354 ;; Simple case. Continuous words in which all those contain
355 ;; non-ASCII characters are encoded collectively. Encoding
356 ;; ASCII words, including `Re:' used in Subject headers, is
357 ;; avoided for interoperability with non-MIME clients and
358 ;; for making it easy to find keywords.
372 ;; `address-mime' case -- take care of quoted words, comments.
377 ;; Look for rfc2822-style: sequences of atoms, quoted
378 ;; strings, specials, whitespace. (Specials mustn't be
379 ;; encoded.)
381 ;; Skip whitespace.
386 ;; else token start
388 ;; Quoted word.
391 ;; Does it need encoding?
394 ;; It needs encoding. Strip the quotes first,
395 ;; since encoded words can't occur in quotes.
402 ;; There was a preceding quoted word. We need
403 ;; to include any separating whitespace in this
404 ;; word to avoid it getting lost.
406 ;; A space is needed between the encoded words.
410 ;; Adjust the end position for the deleted quotes.
415 ;; Skip other delimiters, but record that they've
416 ;; potentially separated quoted words.
422 ;; Look for the end of parentheses.
424 ;; Encode text as an unstructured field.
429 ;; Find the end.
430 ;; Skip one ASCII word, or encode continuous words
431 ;; in which all those contain non-ASCII characters.
442 ;; Skip backslash-quoted characters.
444 end)
453 ;; Where the value nil of `end' means there may be
454 ;; text to have to be encoded following the point.
455 ;; Otherwise, the point reached to the end of ASCII
456 ;; words separated by whitespace or a special char.
464 encodable-regexp end t))
470 nil)))
475 ;; There are special characters better
476 ;; to be encoded so that MTAs may parse
477 ;; them safely.
480 encodable-regexp
486 ;; Separate encodable text and
487 ;; delimiter.
507 ;; Also encode bogus delimiters.
509 ;; Separate encodable text and delimiter.
526 "Encode words in STRING.
527 By default, the string is treated as containing addresses (see
528 `rfc2047-encoding-type')."
535 "Maximum characters of each header line that contain encoded-words.
536 If it is nil, encoded-words will not be folded. Too small value may
541 "Subroutine used by `rfc2047-encode'."
548 string))
553 ;; Reomove a superfluous empty line.
556 ;; Break the line before the open parenthesis.
565 next len)
573 cs)
579 prev nil)
583 rfc2047-encode-max-chars))
587 ;; Break the line after the close parenthesis.
591 prev next
594 prev nil)))))
606 ;; Reomove a superfluous empty line.
609 cs encoder start " " tail
613 "Encode the word(s) in the region B to E.
614 Point moves to the end of the region."
616 cs encoding tail crest eword)
630 rfc2047-charset-encoding-alist))
631 ;; For the charsets that don't have a preferred
632 ;; encoding, choose the one that's shorter.
634 'B
639 ;; `tail' may contain a close parenthesis.
648 ;; `crest' may contain whitespace and an open parenthesis.
655 cs
657 rfc2047-encode-function-alist))
662 tail))
665 ;; The line was folded before encoding.
670 ;; `eword' contains `crest' and `tail'.
682 "Fold the current header field."
689 "Fold long lines in region B to E."
704 qword-break nil)
710 ;; Don't break before the first non-LWSP characters.
718 break nil
719 qword-break nil)
734 ;; Don't break at the start of the field.
745 qword-break nil)
751 ;; Don't break before the first non-LWSP characters.
757 "Fold the current line."
764 "Unfold lines in region B to E."
785 "Base64-encode the header contained in STRING."
789 "Quoted-printable-encode the header in STRING."
794 ;; = (\075), _ (\137), ? (\077) are used in the encoded word.
795 ;; Avoid using 8bit characters.
796 ;; This list excludes `especials' (see the RFC2047 syntax),
797 ;; meaning that some characters in non-structured fields will
798 ;; get encoded when they con't need to be. The following is
799 ;; what it used to be.
800 ;;; ;; Equivalent to "^\000-\007\011\013\015-\037\200-\377=_?"
801 ;;; "\010\012\014\040-\074\076\100-\136\140-\177")
807 "Return and PARAM=VALUE string encoded in the RFC2047-like style.
808 This is a replacement for the `rfc2231-encode-string' function.
810 When attaching files as MIME parts, we should use the RFC2231 encoding
811 to specify the file names containing non-ASCII characters. However,
812 many mail softwares don't support it in practice and recipients won't
813 be able to extract files with correct names. Instead, the RFC2047-like
814 encoding is acceptable generally. This function provides the very
815 RFC2047-like encoding, resigning to such a regrettable trend. To use
816 it, put the following line in your ~/.gnus.el file:
818 \(defalias 'mail-header-encode-parameter 'rfc2047-encode-parameter)
819 "
824 ;;;
825 ;;; Functions for decoding RFC2047 messages
826 ;;;
837 "*Non-nil means allow incomplete encoded-text in successive encoded-words.
838 Dividing of encoded-text in the place other than character boundaries
839 violates RFC2047 section 5, while we have a capability to decode it.
840 If it is non-nil, the decoder will decode B- or Q-encoding in each
841 encoded-word, concatenate them, and decode it by charset. Otherwise,
842 the decoder will fully decode each encoded-word before concatenating
869 "Return coding-system corresponding to MIME CHARSET.
870 If your Emacs implementation can't decode CHARSET, return nil."
888 'raw-text
889 cs)))
892 "Decode successive encoded-words in WORDS and return a decoded string.
893 Each element of WORDS looks like (CHARSET ENCODING ENCODED-TEXT
894 ENCODED-WORD)."
910 nil)))
913 ;; Concatenate text of which the charset is the same.
916 ;; Don't decode encoded-word.
925 nil)))
931 words)
933 words))
935 ;; Fixme: This should decode in place, not cons intermediate strings.
936 ;; Also check whether it needs to worry about delimiting fields like
937 ;; encoding.
939 ;; In fact it's reported that (invalid) encoding of mailboxes in
940 ;; addr-specs is in use, so delimiting fields might help. Probably
941 ;; not decoding a word which isn't properly delimited is good enough
942 ;; and worthwhile (is it more correct or not?), e.g. something like
943 ;; `=?iso-8859-1?q?foo?=@'.
946 "Decode MIME-encoded words in region between START and END.
947 If ADDRESS-MIME is non-nil, strip backslashes which precede characters
952 ;; Ignore whitespace between encoded-words.
955 b e match words)
962 ;; Look for the encoded-words.
966 words nil)
972 words)
973 ;; Look for the subsequent encoded-words.
976 ;; Replace the encoded-words with the decoded one.
982 ;; Remove newlines between decoded words, though such
983 ;; things essentially must not be there.
986 ;; Quote decoded words if there are special characters
987 ;; which might violate RFC2822.
990 'address-mime
991 rfc2047-header-encoding-alist))))
1027 end t)))
1035 mail-parse-charset
1041 mail-parse-charset
1047 "Decode MIME-encoded words in region between START and END.
1049 strings are stripped."
1053 "Decode MIME-encoded STRING and return the result.
1054 If ADDRESS-MIME is non-nil, strip backslashes which precede characters
1059 ;; Fixme: This logic is wrong, but seems to be required by
1060 ;; Gnus summary buffer generation. The value of `m' depends
1061 ;; on the current buffer, not global multibyteness or that
1062 ;; of the string. Also the string returned should always be
1063 ;; multibyte in a multibyte session, i.e. the buffer should
1064 ;; be multibyte before `buffer-string' is called.
1079 ;; Fixme: As above, `m' here is inappropriate.
1081 mail-parse-charset
1084 ;; `decode-coding-string' in Emacs offers a third optional
1085 ;; arg NOCOPY to avoid consing a new string if the decoding
1086 ;; is "trivial". Unfortunately it currently doesn't
1087 ;; consider anything else than a `nil' coding system
1088 ;; trivial.
1089 ;; `rfc2047-decode-string' is called multiple times for each
1090 ;; article during summary buffer generation, and we really
1091 ;; want to avoid unnecessary consing. So we bypass
1092 ;; `decode-coding-string' if the string is purely ASCII.
1094 ;; string is purely ASCII
1096 string
1101 "Decode MIME-encoded STRING and return the result.
1103 strings are stripped."
1107 "Pad STRING to quartets."
1108 ;; Be more liberal to accept buggy base64 strings. If
1109 ;; base64-decode-string accepts buggy strings, this function could
1110 ;; be aliased to identity.
1112 string
1123 ;;; arch-tag: a07fe3d4-22b5-4c4a-bd89-b1f82d5d36f6
1124 ;;; rfc2047.el ends here