1 ;;; rfc2231.el --- Functions for decoding rfc2231 headers
3 ;; Copyright (C) 1998, 1999, 2000, 2002, 2003, 2004, 2005,
4 ;; 2006 Free Software Foundation, Inc.
6 ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org>
7 ;; This file is part of GNU Emacs.
9 ;; GNU Emacs is free software; you can redistribute it and/or modify
10 ;; it under the terms of the GNU General Public License as published by
11 ;; the Free Software Foundation; either version 2, or (at your option)
14 ;; GNU Emacs is distributed in the hope that it will be useful,
15 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
16 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17 ;; GNU General Public License for more details.
19 ;; You should have received a copy of the GNU General Public License
20 ;; along with GNU Emacs; see the file COPYING. If not, write to the
21 ;; Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
22 ;; Boston, MA 02110-1301, USA.
28 (eval-when-compile (require 'cl))
31 (autoload 'mm-encode-body "mm-bodies")
32 (autoload 'mail-header-remove-whitespace "mail-parse")
33 (autoload 'mail-header-remove-comments "mail-parse")
35 (defun rfc2231-get-value (ct attribute)
36 "Return the value of ATTRIBUTE from CT."
37 (cdr (assq attribute (cdr ct))))
39 (defun rfc2231-parse-qp-string (string)
40 "Parse QP-encoded string using `rfc2231-parse-string'.
41 N.B. This is in violation with RFC2047, but it seem to be in common use."
42 (rfc2231-parse-string (rfc2047-decode-string string)))
44 (defun rfc2231-parse-string (string &optional signal-error)
45 "Parse STRING and return a list.
46 The list will be on the form
47 `(name (attribute . value) (attribute . value)...)'.
49 If the optional SIGNAL-ERROR is non-nil, signal an error when this
50 function fails in parsing of parameters. Otherwise, this function
51 must never cause a Lisp error."
53 (let ((ttoken (ietf-drums-token-to-list ietf-drums-text-token))
54 (stoken (ietf-drums-token-to-list ietf-drums-tspecials))
55 (ntoken (ietf-drums-token-to-list "0-9"))
56 c type attribute encoded number prev-attribute vals
57 prev-encoded parameters value)
60 (mail-header-remove-whitespace
61 (mail-header-remove-comments string))
62 ;; The most likely cause of an error is unbalanced parentheses
63 ;; or double-quotes. If all parentheses and double-quotes are
64 ;; quoted meaninglessly with backslashes, removing them might
65 ;; make it parseable. Let's try...
68 (when (and (string-match "\\\\\"" string)
69 (not (string-match "\\`\"\\|[^\\]\"" string)))
70 (setq string (mm-replace-in-string string "\\\\\"" "\"")
72 (when (and (string-match "\\\\(" string)
73 (string-match "\\\\)" string)
74 (not (string-match "\\`(\\|[^\\][()]" string)))
75 (setq string (mm-replace-in-string string "\\\\\\([()]\\)" "\\1")
79 (mail-header-remove-whitespace
80 (mail-header-remove-comments string))))
81 ;; Finally, attempt to extract only type.
83 (concat "\\`[\t\n ]*\\([^" ietf-drums-tspecials "\t\n ]+"
84 "\\(?:/[^" ietf-drums-tspecials
85 "\t\n ]+\\)?\\)\\(?:[\t\n ;]\\|\\'\\)")
87 (match-string 1 string)
89 (let ((table (copy-syntax-table ietf-drums-syntax-table)))
90 (modify-syntax-entry ?\' "w" table)
91 (modify-syntax-entry ?* " " table)
92 (modify-syntax-entry ?\; " " table)
93 (modify-syntax-entry ?= " " table)
94 ;; The following isn't valid, but one should be liberal
95 ;; in what one receives.
96 (modify-syntax-entry ?\: "w" table)
97 (set-syntax-table table))
99 (when (and (memq c ttoken)
100 (not (memq c stoken))
101 (setq type (ignore-errors
103 (buffer-substring (point) (progn
110 (setq c (char-after))
112 (error "Invalid header: %s" string))
114 ;; If c in nil, then this is an invalid header, but
115 ;; since elm generates invalid headers on this form,
117 (when (setq c (char-after))
118 (if (and (memq c ttoken)
119 (not (memq c stoken)))
124 (point) (progn (forward-sexp 1) (point))))))
125 (error "Invalid header: %s" string))
126 (setq c (char-after))
130 (setq c (char-after))
131 (if (not (memq c ntoken))
137 (point) (progn (forward-sexp 1) (point)))))
138 (setq c (char-after))
142 (setq c (char-after)))))
145 ;; See if we have any previous continuations.
146 (when (and prev-attribute
147 (not (eq prev-attribute attribute)))
149 (mapconcat 'cdr (sort vals 'car-less-than-car) ""))
150 (push (cons prev-attribute
152 (rfc2231-decode-encoded-string vals)
155 (setq prev-attribute nil
159 (error "Invalid header: %s" string))
161 (setq c (char-after))
164 (setq value (buffer-substring (1+ (point))
169 (setq value (mapconcat (lambda (c) (format "%%%02x" c))
171 ((and (or (memq c ttoken)
172 ;; EXTENSION: Support non-ascii chars.
174 (not (memq c stoken)))
180 ;; We might not have reached at the end of
181 ;; the value because of non-ascii chars,
182 ;; so we should jump over them if any.
183 (while (and (not (eobp))
184 (> (char-after) ?\177))
189 (error "Invalid header: %s" string)))
192 (push (cons number value) vals)
193 (setq prev-attribute attribute
194 prev-encoded encoded))
195 (push (cons attribute
197 (rfc2231-decode-encoded-string value)
201 ;; Take care of any final continuations.
203 (setq vals (mapconcat 'cdr (sort vals 'car-less-than-car) ""))
204 (push (cons prev-attribute
206 (rfc2231-decode-encoded-string vals)
210 (setq parameters nil)
212 (signal (car err) (cdr err))
213 ;;(message "%s" (error-message-string err))
216 (cons type (nreverse parameters))))))
218 (defun rfc2231-decode-encoded-string (string)
219 "Decode an RFC2231-encoded string.
221 \"us-ascii'en-us'This%20is%20%2A%2A%2Afun%2A%2A%2A\",
222 \"us-ascii''This%20is%20%2A%2A%2Afun%2A%2A%2A\",
223 \"'en-us'This%20is%20%2A%2A%2Afun%2A%2A%2A\",
224 \"''This%20is%20%2A%2A%2Afun%2A%2A%2A\", or
225 \"This is ***fun***\"."
226 (string-match "\\`\\(?:\\([^']+\\)?'\\([^']+\\)?'\\)?\\(.+\\)" string)
227 (let ((coding-system (mm-charset-to-coding-system (match-string 1 string)))
228 ;;(language (match-string 2 string))
229 (value (match-string 3 string)))
230 (mm-with-multibyte-buffer
232 (goto-char (point-min))
233 (while (search-forward "%" nil t)
236 (string-to-number (buffer-substring (point) (+ (point) 2)) 16)
237 (delete-region (1- (point)) (+ (point) 2)))))
238 ;; Decode using the charset, if any.
239 (unless (memq coding-system '(nil ascii))
240 (mm-decode-coding-region (point-min) (point-max) coding-system))
243 (defun rfc2231-encode-string (param value)
244 "Return and PARAM=VALUE string encoded according to RFC2231.
245 Use `mml-insert-parameter' or `mml-insert-parameter-string' to insert
246 the result of this function."
247 (let ((control (ietf-drums-token-to-list ietf-drums-no-ws-ctl-token))
248 (tspecial (ietf-drums-token-to-list ietf-drums-tspecials))
249 (special (ietf-drums-token-to-list "*'%\n\t"))
250 (ascii (ietf-drums-token-to-list ietf-drums-text-token))
252 ;; Don't make lines exceeding 76 column.
253 (limit (- 74 (length param)))
254 spacep encodep charsetp charset broken)
257 (goto-char (point-min))
260 ((or (memq (following-char) control)
261 (memq (following-char) tspecial)
262 (memq (following-char) special))
264 ((eq (following-char) ? )
266 ((not (memq (following-char) ascii))
270 (setq charset (mm-encode-body)))
272 ((or encodep charsetp
275 (> (current-column) (if spacep (- limit 2) limit))))
276 (setq limit (- limit 6))
277 (goto-char (point-min))
278 (insert (symbol-name (or charset 'us-ascii)) "''")
280 (if (or (not (memq (following-char) ascii))
281 (memq (following-char) control)
282 (memq (following-char) tspecial)
283 (memq (following-char) special)
284 (eq (following-char) ? ))
286 (when (>= (current-column) (1- limit))
289 (insert "%" (format "%02x" (following-char)))
291 (when (> (current-column) limit)
295 (goto-char (point-min))
299 (insert (if (>= num 0) " " "")
300 param "*" (format "%d" (incf num)) "*=")
303 (goto-char (point-min))
305 (goto-char (point-max))
308 (goto-char (point-min))
314 ;;; arch-tag: c3ab751d-d108-406a-b301-68882ad8cd63
315 ;;; rfc2231.el ends here