89427
|
1 ;;; utf-7.el --- utf-7 coding system
|
|
2
|
|
3 ;; Copyright (C) 2003 Free Software Foundation, Inc.
|
|
4
|
|
5 ;; Author: Dave Love <fx@gnu.org>
|
|
6 ;; Keywords: i18n, mail
|
|
7
|
|
8 ;; This file is free software; you can redistribute it and/or modify
|
|
9 ;; it under the terms of the GNU General Public License as published by
|
|
10 ;; the Free Software Foundation; either version 2, or (at your option)
|
|
11 ;; any later version.
|
|
12
|
|
13 ;; This file is distributed in the hope that it will be useful,
|
|
14 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
15 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
16 ;; GNU General Public License for more details.
|
|
17
|
|
18 ;; You should have received a copy of the GNU General Public License
|
|
19 ;; along with GNU Emacs; see the file COPYING. If not, write to
|
|
20 ;; the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
|
|
21 ;; Boston, MA 02111-1307, USA.
|
|
22
|
|
23 ;;; Commentary:
|
|
24
|
|
25 ;; Defines a coding system for UTF-7, defined in RFC 2152. Non-ASCII
|
|
26 ;; segments are encoded as base64-encoded big endian UTF-16. Also
|
|
27 ;; defines a variation required for IMAP (RFC 2060).
|
|
28
|
|
29 ;; The encoding and decoding was originally taken from Jon K Hellan's
|
|
30 ;; implementation in Gnus, but has been substantially re-done.
|
|
31
|
|
32 ;; This probably needs more attention. In particular, it's not
|
|
33 ;; completely consistent with iconv's behaviour. It's arguable
|
|
34 ;; whether the IMAP version should be a coding system since it's
|
|
35 ;; apparently only used for IMAP mailbox names.
|
|
36
|
|
37 ;;; Code:
|
|
38
|
|
39 ;; See mule-conf.
|
|
40 ;; (define-coding-system 'utf-7
|
|
41 ;; "UTF-7 encoding of Unicode (RFC 2152)"
|
|
42 ;; :coding-type 'utf-8
|
|
43 ;; :mnemonic ?U
|
|
44 ;; :mime-charset 'utf-7
|
|
45 ;; :charset-list '(unicode)
|
|
46 ;; :pre-write-conversion 'utf-7-pre-write-conversion
|
|
47 ;; :post-read-conversion 'utf-7-post-read-conversion)
|
|
48
|
|
49 ;; (define-coding-system 'utf-7-imap
|
|
50 ;; "UTF-7 encoding of Unicode, IMAP version (RFC 2060)"
|
|
51 ;; :coding-type 'utf-8
|
|
52 ;; :mnemonic ?U
|
|
53 ;; :mime-charset 'utf-7
|
|
54 ;; :charset-list '(unicode)
|
|
55 ;; :pre-write-conversion 'utf-7-imap-pre-write-conversion
|
|
56 ;; :post-read-conversion 'utf-7-imap-post-read-conversion))
|
|
57
|
|
58 ;;;###autoload
|
|
59 (defun utf-7-decode (len imap)
|
|
60 "Decode LEN bytes of UTF-7 at point.
|
|
61 IMAP non-nil means use the IMAP version."
|
|
62 (save-excursion
|
|
63 (save-restriction
|
|
64 (narrow-to-region (point) (+ (point) len))
|
|
65 (let ((not-esc (if imap "^&" "^+"))
|
|
66 (skip-chars (if imap "A-Za-z0-9+," "A-Za-z0-9+/")))
|
|
67 (while (not (eobp))
|
|
68 (skip-chars-forward not-esc)
|
|
69 (unless (eobp)
|
|
70 (forward-char)
|
|
71 (let ((p (point))
|
|
72 (run-length (skip-chars-forward skip-chars)))
|
|
73 (if (eq ?- (char-after))
|
|
74 (delete-char 1))
|
|
75 (unless (= run-length 0) ; encoded lone esc-char
|
|
76 (let ((pl (mod (- run-length) 4)))
|
|
77 (insert-char ?= pl)
|
|
78 (if imap
|
|
79 (subst-char-in-region p (point) ?, ?/))
|
|
80 (base64-decode-region p (point)))
|
|
81 (decode-coding-region p (point) 'utf-16-be)
|
|
82 (save-excursion
|
|
83 (goto-char p)
|
|
84 (delete-backward-char 1)))))))
|
|
85 (- (point-max) (point-min)))))
|
|
86
|
|
87 (defun utf-7-post-read-conversion (len)
|
|
88 (utf-7-decode len nil))
|
|
89
|
|
90 (defun utf-7-imap-post-read-conversion (len)
|
|
91 (utf-7-decode len t))
|
|
92
|
|
93 ;;;###autoload
|
|
94 (defun utf-7-encode (from to imap)
|
|
95 "Encode bytes between FROM and TO to UTF-7.
|
|
96 ESC and SKIP-CHARS are adjusted for the normal and IMAP versions."
|
|
97 (let* ((old-buf (current-buffer))
|
|
98 (esc (if imap ?& ?+))
|
|
99 ;; These are characters which can be encoded asis.
|
|
100 (skip-chars (if imap
|
|
101 "\t\n\r\x20-\x25\x27-\x7e" ; rfc2060
|
|
102 ;; This includes the rfc2152 optional set.
|
|
103 ;; Perhaps it shouldn't (like iconv).
|
|
104 "\t\n\r -*,-[]-}"))
|
|
105 (not-skip-chars (format "^%s%c" skip-chars esc)))
|
|
106 (set-buffer (generate-new-buffer " *temp*"))
|
|
107 (if (stringp from)
|
|
108 (insert from)
|
|
109 (insert-buffer-substring old-buf from to))
|
|
110 (goto-char (point-min))
|
|
111 (while (not (eobp))
|
|
112 (skip-chars-forward skip-chars)
|
|
113 (if (eq ?+ (char-after))
|
|
114 (progn (forward-char)
|
|
115 (insert ?-))
|
|
116 (unless (eobp)
|
|
117 (insert esc)
|
|
118 (let ((p (point)))
|
|
119 (skip-chars-forward not-skip-chars)
|
|
120 (save-restriction
|
|
121 ;; encode-coding-region doesn't preserve point
|
|
122 (narrow-to-region p (point))
|
|
123 (encode-coding-region p (point-max) 'utf-16-be)
|
|
124 (base64-encode-region p (point-max))
|
|
125 (if imap
|
|
126 (subst-char-in-region p (point-max) ?/ ?,))
|
|
127 (goto-char p)
|
|
128 ;; As I read the RFC, this isn't correct, but it's
|
|
129 ;; consistent with iconv, at least regarding `='.
|
|
130 (skip-chars-forward "^= \t\n")
|
|
131 (delete-region (point) (point-max))))
|
|
132 (unless (eobp)
|
|
133 (insert ?-)))))
|
|
134 nil))
|
|
135
|
|
136 (defun utf-7-pre-write-conversion (from to)
|
|
137 (utf-7-encode from to nil))
|
|
138
|
|
139 (defun utf-7-imap-pre-write-conversion (from to)
|
|
140 (utf-7-encode from to t))
|
|
141
|
|
142 (provide 'utf-7)
|
|
143 ;;; utf-7.el ends here
|