| blob | 3cf2940e031f7eed007548b86a2949896558d5dd |
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 2, 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:
37 "Evaluate BODY with syntax table of current buffer set to TABLE.
38 The syntax table of the current buffer is saved, BODY is evaluated, and the
39 saved table is restored, even in case of an abnormal exit.
40 Value is what BODY returns."
56 ;; Fixme: Avoid this (used for mail-parse-charset) mm dependence on gnus.
62 ;; Avoid gnus-util for mm- code.
65 'point-at-bol
70 'point-at-eol
80 "*Header/encoding method alist.
81 The list is traversed sequentially. The keys can either be
82 header regexps or t.
84 The values can be:
86 1) nil, in which case no encoding is done;
87 2) `mime', in which case the header will be encoded according to RFC2047;
88 3) `address-mime', like `mime', but takes account of the rules for address
89 fields (where quoted strings and comments must be treated separately);
90 4) a charset, in which case it will be encoded as that charset;
91 5) `default', in which case the field will be encoded as the rest
118 "Alist of MIME charsets to RFC2047 encodings.
119 Valid encodings are nil, `Q' and `B'. These indicate binary (no) encoding,
131 ;;;
132 ;;; Functions for encoding RFC2047 messages
133 ;;;
136 "Return the type with which to encode the buffer.
137 This is either `base64' or `quoted-printable'."
148 ;; Don't base64, say, a short line with a single
149 ;; non-ASCII char when splitting parts by charset.
151 'quoted-printable
155 "Narrow the buffer to the header on the current line."
167 "Return the value of the field at point."
175 encodable-regexp)
177 Quoting will not be done in a quoted string if it contains characters
178 matching ENCODABLE-REGEXP."
181 beg end)
211 "The type of encoding done by `rfc2047-encode-region'.
212 This should be dynamically bound around calls to
213 `rfc2047-encode-region' to either `mime' or `address-mime'. See
217 "Encode the message header according to `rfc2047-header-encoding-alist'.
218 Should be called narrowed to the head of the message."
227 alist rfc2047-header-encoding-alist)
242 ;; 8 bit must be decoded.
247 ;; No encoding necessary, but folding is nice
257 ;; We found something that may perhaps be encoded.
268 default-enable-multibyte-characters)
269 mail-parse-charset)
271 mail-parse-charset)))
272 ;; We get this when CC'ing messsages to newsgroups with
273 ;; 8-bit names. The group name mail copy just got
274 ;; unconditionally encoded. Previously, it would ask
275 ;; whether to encode, which was quite confusing for the
276 ;; user. If the new behaviour is wrong, tell me. I have
277 ;; left the old code commented out below.
278 ;; -- Per Abrahamsen <abraham@dina.kvl.dk> Date: 2001-10-07.
279 ;; Modified by Dave Love, with the commented-out code changed
280 ;; in accordance with changes elsewhere.
283 ;;; ((null method)
284 ;;; (if (or (message-options-get
285 ;;; 'rfc2047-encode-message-header-encode-any)
286 ;;; (message-options-set
287 ;;; 'rfc2047-encode-message-header-encode-any
288 ;;; (y-or-n-p
289 ;;; "Some texts are not encoded. Encode anyway?")))
290 ;;; (rfc2047-encode-region (point-min) (point-max))
291 ;;; (error "Cannot send unencoded text")))
295 default-enable-multibyte-characters))
297 ;; Hm.
301 ;; Fixme: This, and the require below may not be the Right Thing, but
302 ;; should be safe just before release. -- fx 2001-02-08
306 "Return non-nil if any characters in current buffer need encoding in headers.
307 The buffer may be narrowed."
319 ;; Use this syntax table when parsing into regions that may need
320 ;; encoding. Double quotes are string delimiters, backslash is
321 ;; character quoting, and all other RFC 2822 special characters are
322 ;; treated as punctuation so we can use forward-sexp/forward-word to
323 ;; skip to the end of regions appropriately. Nb. ietf-drums does
324 ;; things differently.
326 ;; (make-char-table 'syntax-table '(2)) only works in Emacs.
328 ;; The following is done to work for setting all elements of the table
329 ;; in Emacs 21 and 22 and XEmacs; it appears to be the cleanest way.
330 ;; Play safe and don't assume the form of the word syntax entry --
331 ;; copy it from ?a.
338 table))))
351 table))
354 "Encode words in region B to E that need encoding.
355 By default, the region is treated as containing RFC2822 addresses.
356 Dynamically bind `rfc2047-encoding-type' to change that."
362 start ; start of current token
363 end begin csyntax
364 ;; Whether there's an encoded word before the current token,
365 ;; either immediately or separated by space.
366 last-encoded
369 ;; Simple case. Continuous words in which all those contain
370 ;; non-ASCII characters are encoded collectively. Encoding
371 ;; ASCII words, including `Re:' used in Subject headers, is
372 ;; avoided for interoperability with non-MIME clients and
373 ;; for making it easy to find keywords.
387 ;; `address-mime' case -- take care of quoted words, comments.
392 ;; Look for rfc2822-style: sequences of atoms, quoted
393 ;; strings, specials, whitespace. (Specials mustn't be
394 ;; encoded.)
396 ;; Skip whitespace.
401 ;; else token start
403 ;; Quoted word.
406 ;; Does it need encoding?
409 ;; It needs encoding. Strip the quotes first,
410 ;; since encoded words can't occur in quotes.
417 ;; There was a preceding quoted word. We need
418 ;; to include any separating whitespace in this
419 ;; word to avoid it getting lost.
421 ;; A space is needed between the encoded words.
425 ;; Adjust the end position for the deleted quotes.
430 ;; Skip other delimiters, but record that they've
431 ;; potentially separated quoted words.
437 ;; Look for the end of parentheses.
439 ;; Encode text as an unstructured field.
444 ;; Find the end.
445 ;; Skip one ASCII word, or encode continuous words
446 ;; in which all those contain non-ASCII characters.
457 ;; Skip backslash-quoted characters.
459 end)
468 ;; Where the value nil of `end' means there may be
469 ;; text to have to be encoded following the point.
470 ;; Otherwise, the point reached to the end of ASCII
471 ;; words separated by whitespace or a special char.
479 encodable-regexp end t))
485 nil)))
490 ;; There are special characters better
491 ;; to be encoded so that MTAs may parse
492 ;; them safely.
495 encodable-regexp
501 ;; Separate encodable text and
502 ;; delimiter.
522 ;; Also encode bogus delimiters.
524 ;; Separate encodable text and delimiter.
541 "Encode words in STRING.
542 By default, the string is treated as containing addresses (see
543 `rfc2047-encoding-type')."
550 "Maximum characters of each header line that contain encoded-words.
551 If it is nil, encoded-words will not be folded. Too small value may
556 "Subroutine used by `rfc2047-encode'."
563 string))
568 ;; Reomove a superfluous empty line.
571 ;; Break the line before the open parenthesis.
580 next len)
588 cs)
594 prev nil)
598 rfc2047-encode-max-chars))
602 ;; Break the line after the close parenthesis.
606 prev next
609 prev nil)))))
621 ;; Reomove a superfluous empty line.
624 cs encoder start " " tail
628 "Encode the word(s) in the region B to E.
629 Point moves to the end of the region."
631 cs encoding tail crest eword)
645 rfc2047-charset-encoding-alist))
646 ;; For the charsets that don't have a preferred
647 ;; encoding, choose the one that's shorter.
649 'B
654 ;; `tail' may contain a close parenthesis.
663 ;; `crest' may contain whitespace and an open parenthesis.
670 cs
672 rfc2047-encode-function-alist))
677 tail))
680 ;; The line was folded before encoding.
685 ;; `eword' contains `crest' and `tail'.
697 "Fold the current header field."
704 "Fold long lines in region B to E."
719 qword-break nil)
725 ;; Don't break before the first non-LWSP characters.
733 break nil
734 qword-break nil)
749 ;; Don't break at the start of the field.
760 qword-break nil)
766 ;; Don't break before the first non-LWSP characters.
772 "Fold the current line."
779 "Unfold lines in region B to E."
800 "Base64-encode the header contained in STRING."
804 "Quoted-printable-encode the header in STRING."
809 ;; = (\075), _ (\137), ? (\077) are used in the encoded word.
810 ;; Avoid using 8bit characters.
811 ;; This list excludes `especials' (see the RFC2047 syntax),
812 ;; meaning that some characters in non-structured fields will
813 ;; get encoded when they con't need to be. The following is
814 ;; what it used to be.
815 ;;; ;; Equivalent to "^\000-\007\011\013\015-\037\200-\377=_?"
816 ;;; "\010\012\014\040-\074\076\100-\136\140-\177")
822 "Return and PARAM=VALUE string encoded in the RFC2047-like style.
823 This is a replacement for the `rfc2231-encode-string' function.
825 When attaching files as MIME parts, we should use the RFC2231 encoding
826 to specify the file names containing non-ASCII characters. However,
827 many mail softwares don't support it in practice and recipients won't
828 be able to extract files with correct names. Instead, the RFC2047-like
829 encoding is acceptable generally. This function provides the very
830 RFC2047-like encoding, resigning to such a regrettable trend. To use
831 it, put the following line in your ~/.gnus.el file:
833 \(defalias 'mail-header-encode-parameter 'rfc2047-encode-parameter)
834 "
839 ;;;
840 ;;; Functions for decoding RFC2047 messages
841 ;;;
852 "*Non-nil means allow incomplete encoded-text in successive encoded-words.
853 Dividing of encoded-text in the place other than character boundaries
854 violates RFC2047 section 5, while we have a capability to decode it.
855 If it is non-nil, the decoder will decode B- or Q-encoding in each
856 encoded-word, concatenate them, and decode it by charset. Otherwise,
857 the decoder will fully decode each encoded-word before concatenating
884 "Return coding-system corresponding to MIME CHARSET.
885 If your Emacs implementation can't decode CHARSET, return nil."
903 'raw-text
904 cs)))
907 "Decode successive encoded-words in WORDS and return a decoded string.
908 Each element of WORDS looks like (CHARSET ENCODING ENCODED-TEXT
909 ENCODED-WORD)."
925 nil)))
928 ;; Concatenate text of which the charset is the same.
931 ;; Don't decode encoded-word.
940 nil)))
946 words)
948 words))
950 ;; Fixme: This should decode in place, not cons intermediate strings.
951 ;; Also check whether it needs to worry about delimiting fields like
952 ;; encoding.
954 ;; In fact it's reported that (invalid) encoding of mailboxes in
955 ;; addr-specs is in use, so delimiting fields might help. Probably
956 ;; not decoding a word which isn't properly delimited is good enough
957 ;; and worthwhile (is it more correct or not?), e.g. something like
958 ;; `=?iso-8859-1?q?foo?=@'.
961 "Decode MIME-encoded words in region between START and END.
962 If ADDRESS-MIME is non-nil, strip backslashes which precede characters
967 ;; Ignore whitespace between encoded-words.
970 b e match words)
977 ;; Look for the encoded-words.
981 words nil)
987 words)
988 ;; Look for the subsequent encoded-words.
991 ;; Replace the encoded-words with the decoded one.
997 ;; Remove newlines between decoded words, though such
998 ;; things essentially must not be there.
1001 ;; Quote decoded words if there are special characters
1002 ;; which might violate RFC2822.
1005 'address-mime
1006 rfc2047-header-encoding-alist))))
1042 end t)))
1050 mail-parse-charset
1056 mail-parse-charset
1062 "Decode MIME-encoded words in region between START and END.
1064 strings are stripped."
1068 "Decode MIME-encoded STRING and return the result.
1069 If ADDRESS-MIME is non-nil, strip backslashes which precede characters
1074 ;; Fixme: This logic is wrong, but seems to be required by
1075 ;; Gnus summary buffer generation. The value of `m' depends
1076 ;; on the current buffer, not global multibyteness or that
1077 ;; of the string. Also the string returned should always be
1078 ;; multibyte in a multibyte session, i.e. the buffer should
1079 ;; be multibyte before `buffer-string' is called.
1094 ;; Fixme: As above, `m' here is inappropriate.
1096 mail-parse-charset
1099 ;; `decode-coding-string' in Emacs offers a third optional
1100 ;; arg NOCOPY to avoid consing a new string if the decoding
1101 ;; is "trivial". Unfortunately it currently doesn't
1102 ;; consider anything else than a `nil' coding system
1103 ;; trivial.
1104 ;; `rfc2047-decode-string' is called multiple times for each
1105 ;; article during summary buffer generation, and we really
1106 ;; want to avoid unnecessary consing. So we bypass
1107 ;; `decode-coding-string' if the string is purely ASCII.
1109 ;; string is purely ASCII
1111 string
1116 "Decode MIME-encoded STRING and return the result.
1118 strings are stripped."
1122 "Pad STRING to quartets."
1123 ;; Be more liberal to accept buggy base64 strings. If
1124 ;; base64-decode-string accepts buggy strings, this function could
1125 ;; be aliased to identity.
1127 string
1138 ;;; arch-tag: a07fe3d4-22b5-4c4a-bd89-b1f82d5d36f6
1139 ;;; rfc2047.el ends here