Mercurial > emacs
annotate lisp/language/china-util.el @ 41015:987844bd98c9
*** empty log message ***
author | Richard M. Stallman <rms@gnu.org> |
---|---|
date | Tue, 13 Nov 2001 16:39:40 +0000 |
parents | 67b464da13ec |
children | ca6dbe4635da |
rev | line source |
---|---|
17315
a3ca5e15c82a
Fix the format of the first line.
Kenichi Handa <handa@m17n.org>
parents:
17098
diff
changeset
|
1 ;;; china-util.el --- utilities for Chinese |
17052 | 2 |
3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN. | |
18377
8b4a66c66dd6
Change copyright notice.
Richard M. Stallman <rms@gnu.org>
parents:
18309
diff
changeset
|
4 ;; Licensed to the Free Software Foundation. |
17052 | 5 |
6 ;; Keywords: mule, multilingual, Chinese | |
7 | |
8 ;; This file is part of GNU Emacs. | |
9 | |
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. | |
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 | |
17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
18 ;; GNU General Public License for more details. | |
19 | |
20 ;; You should have received a copy of the GNU General Public License | |
17071 | 21 ;; along with GNU Emacs; see the file COPYING. If not, write to the |
22 ;; Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
23 ;; Boston, MA 02111-1307, USA. | |
17052 | 24 |
38414
67b464da13ec
Some fixes to follow coding conventions.
Pavel Janík <Pavel@Janik.cz>
parents:
28906
diff
changeset
|
25 ;;; Commentary: |
67b464da13ec
Some fixes to follow coding conventions.
Pavel Janík <Pavel@Janik.cz>
parents:
28906
diff
changeset
|
26 |
17052 | 27 ;;; Code: |
28 | |
29 ;; Hz/ZW encoding stuffs | |
30 | |
31 ;; HZ is an encoding method for Chinese character set GB2312 used | |
32 ;; widely in Internet. It is very similar to 7-bit environment of | |
33 ;; ISO-2022. The difference is that HZ uses the sequence "~{" and | |
34 ;; "~}" for designating GB2312 and ASCII respectively, hence, it | |
35 ;; doesn't uses ESC (0x1B) code. | |
36 | |
37 ;; ZW is another encoding method for Chinese character set GB2312. It | |
38 ;; encodes Chinese characters line by line by starting each line with | |
39 ;; the sequence "zW". It also uses only 7-bit as HZ. | |
40 | |
41 ;; ISO-2022 escape sequence to designate GB2312. | |
42 (defvar iso2022-gb-designation "\e$A") | |
43 ;; HZ escape sequence to designate GB2312. | |
44 (defvar hz-gb-designnation "~{") | |
45 ;; ISO-2022 escape sequence to designate ASCII. | |
46 (defvar iso2022-ascii-designation "\e(B") | |
47 ;; HZ escape sequence to designate ASCII. | |
48 (defvar hz-ascii-designnation "~}") | |
49 ;; Regexp of ZW sequence to start GB2312. | |
50 (defvar zw-start-gb "^zW") | |
51 ;; Regexp for start of GB2312 in an encoding mixture of HZ and ZW. | |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
52 (defvar hz/zw-start-gb |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
53 (concat hz-gb-designnation "\\|" zw-start-gb "\\|[^\0-\177]")) |
17052 | 54 |
55 (defvar decode-hz-line-continuation nil | |
56 "Flag to tell if we should care line continuation convention of Hz.") | |
57 | |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
58 (defconst hz-set-msb-table |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
59 (let ((str (make-string 127 0)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
60 (i 0)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
61 (while (< i 33) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
62 (aset str i i) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
63 (setq i (1+ i))) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
64 (while (< i 127) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
65 (aset str i (+ i 128)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
66 (setq i (1+ i))) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
67 str)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
68 |
17052 | 69 ;;;###autoload |
70 (defun decode-hz-region (beg end) | |
71 "Decode HZ/ZW encoded text in the current region. | |
72 Return the length of resulting text." | |
73 (interactive "r") | |
74 (save-excursion | |
75 (save-restriction | |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
76 (let (pos ch) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
77 (narrow-to-region beg end) |
17052 | 78 |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
79 ;; We, at first, convert HZ/ZW to `euc-china', |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
80 ;; then decode it. |
17052 | 81 |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
82 ;; "~\n" -> "\n", "~~" -> "~" |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
83 (goto-char (point-min)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
84 (while (search-forward "~" nil t) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
85 (setq ch (following-char)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
86 (if (or (= ch ?\n) (= ch ?~)) (delete-char -1))) |
17052 | 87 |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
88 ;; "^zW...\n" -> Chinese GB2312 |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
89 ;; "~{...~}" -> Chinese GB2312 |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
90 (goto-char (point-min)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
91 (setq beg nil) |
17052 | 92 (while (re-search-forward hz/zw-start-gb nil t) |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
93 (setq pos (match-beginning 0) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
94 ch (char-after pos)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
95 ;; Record the first position to start conversion. |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
96 (or beg (setq beg pos)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
97 (end-of-line) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
98 (setq end (point)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
99 (if (>= ch 128) ; 8bit GB2312 |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
100 nil |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
101 (goto-char pos) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
102 (delete-char 2) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
103 (setq end (- end 2)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
104 (if (= ch ?z) ; ZW -> euc-china |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
105 (progn |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
106 (translate-region (point) end hz-set-msb-table) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
107 (goto-char end)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
108 (if (search-forward hz-ascii-designnation |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
109 (if decode-hz-line-continuation nil end) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
110 t) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
111 (delete-char -2)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
112 (setq end (point)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
113 (translate-region pos (point) hz-set-msb-table)))) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
114 (if beg |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
115 (decode-coding-region beg end 'euc-china))) |
17052 | 116 (- (point-max) (point-min))))) |
117 | |
118 ;;;###autoload | |
119 (defun decode-hz-buffer () | |
120 "Decode HZ/ZW encoded text in the current buffer." | |
121 (interactive) | |
122 (decode-hz-region (point-min) (point-max))) | |
123 | |
124 ;;;###autoload | |
125 (defun encode-hz-region (beg end) | |
126 "Encode the text in the current region to HZ. | |
127 Return the length of resulting text." | |
128 (interactive "r") | |
129 (save-excursion | |
130 (save-restriction | |
131 (narrow-to-region beg end) | |
132 | |
133 ;; "~" -> "~~" | |
134 (goto-char (point-min)) | |
135 (while (search-forward "~" nil t) (insert ?~)) | |
136 | |
137 ;; Chinese GB2312 -> "~{...~}" | |
138 (goto-char (point-min)) | |
139 (if (re-search-forward "\\cc" nil t) | |
20838
3d67d591c066
(encode-hz-region): Do not bind
Kenichi Handa <handa@m17n.org>
parents:
20738
diff
changeset
|
140 (let (pos) |
17052 | 141 (goto-char (setq pos (match-beginning 0))) |
18553
62e17ab4e33e
Use true coding system names instead of
Kenichi Handa <handa@m17n.org>
parents:
18377
diff
changeset
|
142 (encode-coding-region pos (point-max) 'iso-2022-7bit) |
17052 | 143 (goto-char pos) |
144 (while (search-forward iso2022-gb-designation nil t) | |
145 (delete-char -3) | |
146 (insert hz-gb-designnation)) | |
147 (goto-char pos) | |
148 (while (search-forward iso2022-ascii-designation nil t) | |
149 (delete-char -3) | |
150 (insert hz-ascii-designnation)))) | |
151 (- (point-max) (point-min))))) | |
152 | |
153 ;;;###autoload | |
154 (defun encode-hz-buffer () | |
155 "Encode the text in the current buffer to HZ." | |
156 (interactive) | |
157 (encode-hz-region (point-min) (point-max))) | |
158 | |
159 ;; | |
18309
bd8b521f5218
Provide XXX-util instead of
Kenichi Handa <handa@m17n.org>
parents:
17993
diff
changeset
|
160 (provide 'china-util) |
17052 | 161 |
162 ;;; china-util.el ends here |