Commit | Line | Data |
---|---|---|
c113de23 | 1 | ;;; qp.el --- Quoted-Printable functions |
657b2c65 | 2 | |
73b0cd50 | 3 | ;; Copyright (C) 1998-2011 Free Software Foundation, Inc. |
c113de23 GM |
4 | |
5 | ;; Author: Lars Magne Ingebrigtsen <larsi@gnus.org> | |
657b2c65 DL |
6 | ;; Keywords: mail, extensions |
7 | ||
c113de23 GM |
8 | ;; This file is part of GNU Emacs. |
9 | ||
5e809f55 | 10 | ;; GNU Emacs is free software: you can redistribute it and/or modify |
c113de23 | 11 | ;; it under the terms of the GNU General Public License as published by |
5e809f55 GM |
12 | ;; the Free Software Foundation, either version 3 of the License, or |
13 | ;; (at your option) any later version. | |
c113de23 GM |
14 | |
15 | ;; GNU Emacs is distributed in the hope that it will be useful, | |
16 | ;; but WITHOUT ANY WARRANTY; without even the implied warranty of | |
5e809f55 | 17 | ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
c113de23 GM |
18 | ;; GNU General Public License for more details. |
19 | ||
20 | ;; You should have received a copy of the GNU General Public License | |
5e809f55 | 21 | ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>. |
c113de23 GM |
22 | |
23 | ;;; Commentary: | |
24 | ||
657b2c65 DL |
25 | ;; Functions for encoding and decoding quoted-printable text as |
26 | ;; defined in RFC 2045. | |
c113de23 | 27 | |
657b2c65 | 28 | ;;; Code: |
c113de23 | 29 | |
b756f5b4 | 30 | (require 'mm-util) |
9efa445f | 31 | (defvar mm-use-ultra-safe-encoding) |
d21a76da | 32 | |
23f87bed | 33 | ;;;###autoload |
d21a76da | 34 | (defun quoted-printable-decode-region (from to &optional coding-system) |
657b2c65 | 35 | "Decode quoted-printable in the region between FROM and TO, per RFC 2045. |
d21a76da | 36 | If CODING-SYSTEM is non-nil, decode bytes into characters with that |
5c543a33 SZ |
37 | coding-system. |
38 | ||
39 | Interactively, you can supply the CODING-SYSTEM argument | |
23f87bed MB |
40 | with \\[universal-coding-system-argument]. |
41 | ||
42 | The CODING-SYSTEM argument is a historical hangover and is deprecated. | |
43 | QP encodes raw bytes and should be decoded into raw bytes. Decoding | |
44 | them into characters should be done separately." | |
753b4645 EZ |
45 | (interactive |
46 | ;; Let the user determine the coding system with "C-x RET c". | |
47 | (list (region-beginning) (region-end) coding-system-for-read)) | |
b5287163 DL |
48 | (unless (mm-coding-system-p coding-system) ; e.g. `ascii' from Gnus |
49 | (setq coding-system nil)) | |
c113de23 GM |
50 | (save-excursion |
51 | (save-restriction | |
b5287163 DL |
52 | ;; RFC 2045: ``An "=" followed by two hexadecimal digits, one |
53 | ;; or both of which are lowercase letters in "abcdef", is | |
54 | ;; formally illegal. A robust implementation might choose to | |
55 | ;; recognize them as the corresponding uppercase letters.'' | |
d21a76da | 56 | (let ((case-fold-search t)) |
c113de23 | 57 | (narrow-to-region from to) |
d21a76da DL |
58 | ;; Do this in case we're called from Gnus, say, in a buffer |
59 | ;; which already contains non-ASCII characters which would | |
60 | ;; then get doubly-decoded below. | |
61 | (if coding-system | |
62 | (mm-encode-coding-region (point-min) (point-max) coding-system)) | |
63 | (goto-char (point-min)) | |
e4f99da7 | 64 | (while (and (skip-chars-forward "^=") |
657b2c65 DL |
65 | (not (eobp))) |
66 | (cond ((eq (char-after (1+ (point))) ?\n) | |
67 | (delete-char 2)) | |
68 | ((looking-at "=[0-9A-F][0-9A-F]") | |
e9bd5782 | 69 | (let ((byte (string-to-number (buffer-substring (1+ (point)) |
01c52d31 MB |
70 | (+ 3 (point))) |
71 | 16))) | |
5f8dd322 | 72 | (mm-insert-byte byte 1) |
23f87bed | 73 | (delete-char 3))) |
657b2c65 | 74 | (t |
23f87bed | 75 | (message "Malformed quoted-printable text") |
d21a76da DL |
76 | (forward-char))))) |
77 | (if coding-system | |
78 | (mm-decode-coding-region (point-min) (point-max) coding-system))))) | |
c113de23 | 79 | |
d21a76da | 80 | (defun quoted-printable-decode-string (string &optional coding-system) |
657b2c65 | 81 | "Decode the quoted-printable encoded STRING and return the result. |
c590ed3a | 82 | If CODING-SYSTEM is non-nil, decode the string with coding-system. |
23f87bed MB |
83 | Use of CODING-SYSTEM is deprecated; this function should deal with |
84 | raw bytes, and coding conversion should be done separately." | |
85 | (mm-with-unibyte-buffer | |
c113de23 | 86 | (insert string) |
d21a76da | 87 | (quoted-printable-decode-region (point-min) (point-max) coding-system) |
c113de23 GM |
88 | (buffer-string))) |
89 | ||
90 | (defun quoted-printable-encode-region (from to &optional fold class) | |
657b2c65 | 91 | "Quoted-printable encode the region between FROM and TO per RFC 2045. |
c113de23 | 92 | |
657b2c65 | 93 | If FOLD, fold long lines at 76 characters (as required by the RFC). |
e4f99da7 DL |
94 | If CLASS is non-nil, translate the characters not matched by that |
95 | regexp class, which is in the form expected by `skip-chars-forward'. | |
96 | You should probably avoid non-ASCII characters in this arg. | |
c113de23 | 97 | |
657b2c65 | 98 | If `mm-use-ultra-safe-encoding' is set, fold lines unconditionally and |
c113de23 GM |
99 | encode lines starting with \"From\"." |
100 | (interactive "r") | |
657b2c65 | 101 | (unless class |
158d6e07 SZ |
102 | ;; Avoid using 8bit characters. = is \075. |
103 | ;; Equivalent to "^\000-\007\013\015-\037\200-\377=" | |
104 | (setq class "\010-\012\014\040-\074\076-\177")) | |
c113de23 | 105 | (save-excursion |
cd01ded9 MB |
106 | (goto-char from) |
107 | (if (re-search-forward (mm-string-to-multibyte "[^\x0-\x7f\x80-\xff]") | |
108 | to t) | |
109 | (error "Multibyte character in QP encoding region")) | |
c113de23 GM |
110 | (save-restriction |
111 | (narrow-to-region from to) | |
e4f99da7 DL |
112 | ;; Encode all the non-ascii and control characters. |
113 | (goto-char (point-min)) | |
114 | (while (and (skip-chars-forward class) | |
115 | (not (eobp))) | |
116 | (insert | |
117 | (prog1 | |
5f4264e5 | 118 | ;; To unibyte in case of Emacs 23 (unicode) eight-bit. |
5f8dd322 | 119 | (format "=%02X" (mm-multibyte-char-to-unibyte (char-after))) |
e4f99da7 DL |
120 | (delete-char 1)))) |
121 | ;; Encode white space at the end of lines. | |
122 | (goto-char (point-min)) | |
123 | (while (re-search-forward "[ \t]+$" nil t) | |
124 | (goto-char (match-beginning 0)) | |
125 | (while (not (eolp)) | |
c113de23 GM |
126 | (insert |
127 | (prog1 | |
158d6e07 | 128 | (format "=%02X" (char-after)) |
b03b1ad2 DL |
129 | (delete-char 1))))) |
130 | (let ((mm-use-ultra-safe-encoding | |
131 | (and (boundp 'mm-use-ultra-safe-encoding) | |
132 | mm-use-ultra-safe-encoding))) | |
133 | (when (or fold mm-use-ultra-safe-encoding) | |
134 | (let ((tab-width 1)) ; HTAB is one character. | |
135 | (goto-char (point-min)) | |
136 | (while (not (eobp)) | |
137 | ;; In ultra-safe mode, encode "From " at the beginning | |
138 | ;; of a line. | |
139 | (when mm-use-ultra-safe-encoding | |
140 | (if (looking-at "From ") | |
141 | (replace-match "From=20" nil t) | |
142 | (if (looking-at "-") | |
143 | (replace-match "=2D" nil t)))) | |
144 | (end-of-line) | |
145 | ;; Fold long lines. | |
146 | (while (> (current-column) 76) ; tab-width must be 1. | |
147 | (beginning-of-line) | |
148 | (forward-char 75) ; 75 chars plus an "=" | |
149 | (search-backward "=" (- (point) 2) t) | |
150 | (insert "=\n") | |
151 | (end-of-line)) | |
152 | (forward-line)))))))) | |
c113de23 GM |
153 | |
154 | (defun quoted-printable-encode-string (string) | |
657b2c65 | 155 | "Encode the STRING as quoted-printable and return the result." |
9c1cf631 GM |
156 | (with-temp-buffer |
157 | (if (mm-multibyte-string-p string) | |
158 | (mm-enable-multibyte) | |
159 | (mm-disable-multibyte)) | |
160 | (insert string) | |
161 | (quoted-printable-encode-region (point-min) (point-max)) | |
162 | (buffer-string))) | |
c113de23 GM |
163 | |
164 | (provide 'qp) | |
165 | ||
657b2c65 | 166 | ;;; qp.el ends here |