| blob | fc184666ac5cb309e1831ed44d2eea50d8416798 |
1 ;;; rfc2047.el --- functions for encoding and decoding rfc2047 messages
3 ;; Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004,
4 ;; 2005, 2006, 2007, 2008 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
91 "Alist of MIME charsets to RFC2047 encodings.
92 Valid encodings are nil, `Q' and `B'. These indicate binary (no) encoding,
113 "Regexp that matches encoded word."
114 ;; The patterns for the B encoding and the Q encoding, i.e. the ones
115 ;; beginning with "B" and "Q" respectively, are restricted into only
116 ;; the characters that those encodings may generally use.
117 )
123 "Regexp that matches encoded word allowing loose Q encoding."
124 ;; The pattern for the Q encoding, i.e. the one beginning with "Q",
125 ;; is similar to:
126 ;; "Q\\?\\(\\?+[^\n=?]\\)?\\([^\n?]+\\?+[^\n=?]\\)*[^\n?]*\\?*"
127 ;; <--------1-------><----------2,3----------><--4--><-5->
128 ;; They mean:
129 ;; 1. After "Q?", allow "?"s that follow a character other than "=".
130 ;; 2. Allow "=" after "Q?"; it isn't regarded as the terminator.
131 ;; 3. In the middle of an encoded word, allow "?"s that follow a
132 ;; character other than "=".
133 ;; 4. Allow any characters other than "?" in the middle of an
134 ;; encoded word.
135 ;; 5. At the end, allow "?"s.
136 ))
138 ;;;
139 ;;; Functions for encoding RFC2047 messages
140 ;;;
143 "Return the type with which to encode the buffer.
144 This is either `base64' or `quoted-printable'."
155 ;; Don't base64, say, a short line with a single
156 ;; non-ASCII char when splitting parts by charset.
158 'quoted-printable
162 "Narrow the buffer to the header on the current line."
174 "Return the value of the field at point."
182 encodable-regexp)
184 Quoting will not be done in a quoted string if it contains characters
185 matching ENCODABLE-REGEXP or it is within parentheses."
189 beg end)
231 "The type of encoding done by `rfc2047-encode-region'.
232 This should be dynamically bound around calls to
233 `rfc2047-encode-region' to either `mime' or `address-mime'. See
237 "Encode the message header according to `rfc2047-header-encoding-alist'.
238 Should be called narrowed to the head of the message."
247 alist rfc2047-header-encoding-alist)
262 ;; 8 bit must be decoded.
267 ;; No encoding necessary, but folding is nice
277 ;; We found something that may perhaps be encoded.
288 default-enable-multibyte-characters)
289 mail-parse-charset)
291 mail-parse-charset)))
292 ;; We get this when CC'ing messsages to newsgroups with
293 ;; 8-bit names. The group name mail copy just got
294 ;; unconditionally encoded. Previously, it would ask
295 ;; whether to encode, which was quite confusing for the
296 ;; user. If the new behaviour is wrong, tell me. I have
297 ;; left the old code commented out below.
298 ;; -- Per Abrahamsen <abraham@dina.kvl.dk> Date: 2001-10-07.
299 ;; Modified by Dave Love, with the commented-out code changed
300 ;; in accordance with changes elsewhere.
303 ;;; ((null method)
304 ;;; (if (or (message-options-get
305 ;;; 'rfc2047-encode-message-header-encode-any)
306 ;;; (message-options-set
307 ;;; 'rfc2047-encode-message-header-encode-any
308 ;;; (y-or-n-p
309 ;;; "Some texts are not encoded. Encode anyway?")))
310 ;;; (rfc2047-encode-region (point-min) (point-max))
311 ;;; (error "Cannot send unencoded text")))
315 default-enable-multibyte-characters))
318 ;; Hm.
322 ;; Fixme: This, and the require below may not be the Right Thing, but
323 ;; should be safe just before release. -- fx 2001-02-08
326 "Return non-nil if any characters in current buffer need encoding in headers.
327 The buffer may be narrowed."
339 ;; Use this syntax table when parsing into regions that may need
340 ;; encoding. Double quotes are string delimiters, backslash is
341 ;; character quoting, and all other RFC 2822 special characters are
342 ;; treated as punctuation so we can use forward-sexp/forward-word to
343 ;; skip to the end of regions appropriately. Nb. ietf-drums does
344 ;; things differently.
346 ;; (make-char-table 'syntax-table '(2)) only works in Emacs.
348 ;; The following is done to work for setting all elements of the table
349 ;; in Emacs 21-23 and XEmacs; it appears to be the cleanest way.
350 ;; Play safe and don't assume the form of the word syntax entry --
351 ;; copy it from ?a.
358 table))))
371 table))
374 "Encode words in region B to E that need encoding.
375 By default, the region is treated as containing RFC2822 addresses.
376 Dynamically bind `rfc2047-encoding-type' to change that."
382 start ; start of current token
383 end begin csyntax
384 ;; Whether there's an encoded word before the current token,
385 ;; either immediately or separated by space.
386 last-encoded
389 ;; Simple case. Continuous words in which all those contain
390 ;; non-ASCII characters are encoded collectively. Encoding
391 ;; ASCII words, including `Re:' used in Subject headers, is
392 ;; avoided for interoperability with non-MIME clients and
393 ;; for making it easy to find keywords.
407 ;; `address-mime' case -- take care of quoted words, comments.
412 ;; Look for rfc2822-style: sequences of atoms, quoted
413 ;; strings, specials, whitespace. (Specials mustn't be
414 ;; encoded.)
416 ;; Skip whitespace.
421 ;; else token start
423 ;; Quoted word.
426 ;; Does it need encoding?
429 ;; It needs encoding. Strip the quotes first,
430 ;; since encoded words can't occur in quotes.
437 ;; There was a preceding quoted word. We need
438 ;; to include any separating whitespace in this
439 ;; word to avoid it getting lost.
441 ;; A space is needed between the encoded words.
445 ;; Adjust the end position for the deleted quotes.
450 ;; Skip other delimiters, but record that they've
451 ;; potentially separated quoted words.
457 ;; Look for the end of parentheses.
459 ;; Encode text as an unstructured field.
464 ;; Find the end.
465 ;; Skip one ASCII word, or encode continuous words
466 ;; in which all those contain non-ASCII characters.
477 ;; Skip backslash-quoted characters.
479 end)
488 ;; Where the value nil of `end' means there may be
489 ;; text to have to be encoded following the point.
490 ;; Otherwise, the point reached to the end of ASCII
491 ;; words separated by whitespace or a special char.
499 encodable-regexp end t))
505 nil)))
510 ;; There are special characters better
511 ;; to be encoded so that MTAs may parse
512 ;; them safely.
515 encodable-regexp
521 ;; Separate encodable text and
522 ;; delimiter.
542 ;; Also encode bogus delimiters.
544 ;; Separate encodable text and delimiter.
561 "Encode words in STRING.
562 By default, the string is treated as containing addresses (see
563 `rfc2047-encoding-type')."
569 ;; From RFC 2047:
570 ;; 2. Syntax of encoded-words
571 ;; [...]
572 ;; While there is no limit to the length of a multiple-line header
573 ;; field, each line of a header field that contains one or more
574 ;; 'encoded-word's is limited to 76 characters.
575 ;;
576 ;; In `rfc2047-encode-parameter' it is bound to nil, so don't defconst it.
578 "Maximum characters of each header line that contain encoded-words.
579 According to RFC 2047, it is 76. If it is nil, encoded-words
580 will not be folded. Too small value may cause an error. You
585 "Subroutine used by `rfc2047-encode'."
592 string))
597 ;; Reomove a superfluous empty line.
600 ;; Break the line before the open parenthesis.
609 next len)
617 cs)
623 prev nil)
627 rfc2047-encode-max-chars))
631 ;; Break the line after the close parenthesis.
635 prev next
638 prev nil)))))
650 ;; Reomove a superfluous empty line.
653 cs encoder start " " tail
657 "Encode the word(s) in the region B to E.
658 Point moves to the end of the region."
660 cs encoding tail crest eword)
674 rfc2047-charset-encoding-alist))
675 ;; For the charsets that don't have a preferred
676 ;; encoding, choose the one that's shorter.
678 'B
683 ;; `tail' may contain a close parenthesis.
692 ;; `crest' may contain whitespace and an open parenthesis.
699 cs
701 rfc2047-encode-function-alist))
706 tail))
709 ;; The line was folded before encoding.
714 ;; `eword' contains `crest' and `tail'.
726 "Fold the current header field."
733 "Fold long lines in region B to E."
748 qword-break nil)
754 ;; Don't break before the first non-LWSP characters.
762 break nil
763 qword-break nil)
778 ;; Don't break at the start of the field.
789 qword-break nil)
795 ;; Don't break before the first non-LWSP characters.
801 "Fold the current line."
808 "Unfold lines in region B to E."
829 "Base64-encode the header contained in STRING."
833 "Quoted-printable-encode the header in STRING."
838 ;; = (\075), _ (\137), ? (\077) are used in the encoded word.
839 ;; Avoid using 8bit characters.
840 ;; This list excludes `especials' (see the RFC2047 syntax),
841 ;; meaning that some characters in non-structured fields will
842 ;; get encoded when they con't need to be. The following is
843 ;; what it used to be.
844 ;;; ;; Equivalent to "^\000-\007\011\013\015-\037\200-\377=_?"
845 ;;; "\010\012\014\040-\074\076\100-\136\140-\177")
851 "Return and PARAM=VALUE string encoded in the RFC2047-like style.
852 This is a replacement for the `rfc2231-encode-string' function.
854 When attaching files as MIME parts, we should use the RFC2231 encoding
855 to specify the file names containing non-ASCII characters. However,
856 many mail softwares don't support it in practice and recipients won't
857 be able to extract files with correct names. Instead, the RFC2047-like
858 encoding is acceptable generally. This function provides the very
859 RFC2047-like encoding, resigning to such a regrettable trend. To use
860 it, put the following line in your ~/.gnus.el file:
862 \(defalias 'mail-header-encode-parameter 'rfc2047-encode-parameter)
863 "
868 ;;;
869 ;;; Functions for decoding RFC2047 messages
870 ;;;
876 "*Non-nil means allow incomplete encoded-text in successive encoded-words.
877 Dividing of encoded-text in the place other than character boundaries
878 violates RFC2047 section 5, while we have a capability to decode it.
879 If it is non-nil, the decoder will decode B- or Q-encoding in each
880 encoded-word, concatenate them, and decode it by charset. Otherwise,
881 the decoder will fully decode each encoded-word before concatenating
908 "Return coding-system corresponding to MIME CHARSET.
909 If your Emacs implementation can't decode CHARSET, return nil."
927 'raw-text
928 cs)))
931 "Decode successive encoded-words in WORDS and return a decoded string.
932 Each element of WORDS looks like (CHARSET ENCODING ENCODED-TEXT
933 ENCODED-WORD)."
949 nil)))
952 ;; Concatenate text of which the charset is the same.
955 ;; Don't decode encoded-word.
964 nil)))
970 words)
972 words))
974 ;; Fixme: This should decode in place, not cons intermediate strings.
975 ;; Also check whether it needs to worry about delimiting fields like
976 ;; encoding.
978 ;; In fact it's reported that (invalid) encoding of mailboxes in
979 ;; addr-specs is in use, so delimiting fields might help. Probably
980 ;; not decoding a word which isn't properly delimited is good enough
981 ;; and worthwhile (is it more correct or not?), e.g. something like
982 ;; `=?iso-8859-1?q?foo?=@'.
985 "Decode MIME-encoded words in region between START and END.
986 If ADDRESS-MIME is non-nil, strip backslashes which precede characters
996 b e match words)
1003 ;; Look for the encoded-words.
1007 words nil)
1013 words)
1014 ;; Look for the subsequent encoded-words.
1017 ;; Replace the encoded-words with the decoded one.
1023 ;; Remove newlines between decoded words, though such
1024 ;; things essentially must not be there.
1027 ;; Quote decoded words if there are special characters
1028 ;; which might violate RFC2822.
1031 'address-mime
1032 rfc2047-header-encoding-alist))))
1068 end t)))
1076 mail-parse-charset
1082 mail-parse-charset
1088 "Decode MIME-encoded words in region between START and END.
1090 strings are stripped."
1094 "Decode MIME-encoded STRING and return the result.
1095 If ADDRESS-MIME is non-nil, strip backslashes which precede characters
1097 ;; (let ((m (mm-multibyte-p)))
1100 ;; We used to only call mm-enable-multibyte if `m' is non-nil,
1101 ;; but this can't be the right criterion. Don't just revert this
1102 ;; change if it encounters a bug. Please help me fix it
1103 ;; right instead. --Stef
1104 ;; The string returned should always be multibyte in a multibyte
1105 ;; session, i.e. the buffer should be multibyte before
1106 ;; `buffer-string' is called.
1120 ;; Fixme: As above, `m' here is inappropriate.
1122 mail-parse-charset
1125 ;; `decode-coding-string' in Emacs offers a third optional
1126 ;; arg NOCOPY to avoid consing a new string if the decoding
1127 ;; is "trivial". Unfortunately it currently doesn't
1128 ;; consider anything else than a `nil' coding system
1129 ;; trivial.
1130 ;; `rfc2047-decode-string' is called multiple times for each
1131 ;; article during summary buffer generation, and we really
1132 ;; want to avoid unnecessary consing. So we bypass
1133 ;; `decode-coding-string' if the string is purely ASCII.
1135 ;; string is purely ASCII
1137 string
1142 "Decode MIME-encoded STRING and return the result.
1144 strings are stripped."
1148 "Pad STRING to quartets."
1149 ;; Be more liberal to accept buggy base64 strings. If
1150 ;; base64-decode-string accepts buggy strings, this function could
1151 ;; be aliased to identity.
1153 string
1164 ;; arch-tag: a07fe3d4-22b5-4c4a-bd89-b1f82d5d36f6
1165 ;;; rfc2047.el ends here