Mercurial > emacs
annotate lisp/language/china-util.el @ 26981:bf4bfaf9f594
*** empty log message ***
author | Dave Love <fx@gnu.org> |
---|---|
date | Wed, 22 Dec 1999 23:15:17 +0000 |
parents | 708271862495 |
children | 60eb71a9f901 |
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 |
25 ;;; Code: | |
26 | |
17993
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
27 ;;;###autoload |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
28 (defun setup-chinese-gb-environment () |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
29 "Setup multilingual environment (MULE) for Chinese GB2312 users." |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
30 (interactive) |
22987
708271862495
(setup-XXX-environment): Just call set-language-environment. If
Kenichi Handa <handa@m17n.org>
parents:
20942
diff
changeset
|
31 (set-language-environment "Chinese-GB")) |
17993
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
32 |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
33 ;;;###autoload |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
34 (defun setup-chinese-big5-environment () |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
35 "Setup multilingual environment (MULE) for Chinese Big5 users." |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
36 (interactive) |
22987
708271862495
(setup-XXX-environment): Just call set-language-environment. If
Kenichi Handa <handa@m17n.org>
parents:
20942
diff
changeset
|
37 (set-language-environment "Chinese-BIG5")) |
17993
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
38 |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
39 ;;;###autoload |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
40 (defun setup-chinese-cns-environment () |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
41 "Setup multilingual environment (MULE) for Chinese CNS11643 family users." |
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
42 (interactive) |
22987
708271862495
(setup-XXX-environment): Just call set-language-environment. If
Kenichi Handa <handa@m17n.org>
parents:
20942
diff
changeset
|
43 (set-language-environment "Chinese-CNS")) |
17993
73869115ae0a
Most of setup-LANGUAGE-environment functions are
Kenichi Handa <handa@m17n.org>
parents:
17315
diff
changeset
|
44 |
17052 | 45 ;; Hz/ZW encoding stuffs |
46 | |
47 ;; HZ is an encoding method for Chinese character set GB2312 used | |
48 ;; widely in Internet. It is very similar to 7-bit environment of | |
49 ;; ISO-2022. The difference is that HZ uses the sequence "~{" and | |
50 ;; "~}" for designating GB2312 and ASCII respectively, hence, it | |
51 ;; doesn't uses ESC (0x1B) code. | |
52 | |
53 ;; ZW is another encoding method for Chinese character set GB2312. It | |
54 ;; encodes Chinese characters line by line by starting each line with | |
55 ;; the sequence "zW". It also uses only 7-bit as HZ. | |
56 | |
57 ;; ISO-2022 escape sequence to designate GB2312. | |
58 (defvar iso2022-gb-designation "\e$A") | |
59 ;; HZ escape sequence to designate GB2312. | |
60 (defvar hz-gb-designnation "~{") | |
61 ;; ISO-2022 escape sequence to designate ASCII. | |
62 (defvar iso2022-ascii-designation "\e(B") | |
63 ;; HZ escape sequence to designate ASCII. | |
64 (defvar hz-ascii-designnation "~}") | |
65 ;; Regexp of ZW sequence to start GB2312. | |
66 (defvar zw-start-gb "^zW") | |
67 ;; 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
|
68 (defvar hz/zw-start-gb |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
69 (concat hz-gb-designnation "\\|" zw-start-gb "\\|[^\0-\177]")) |
17052 | 70 |
71 (defvar decode-hz-line-continuation nil | |
72 "Flag to tell if we should care line continuation convention of Hz.") | |
73 | |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
74 (defconst hz-set-msb-table |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
75 (let ((str (make-string 127 0)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
76 (i 0)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
77 (while (< i 33) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
78 (aset str i i) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
79 (setq i (1+ i))) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
80 (while (< i 127) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
81 (aset str i (+ i 128)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
82 (setq i (1+ i))) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
83 str)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
84 |
17052 | 85 ;;;###autoload |
86 (defun decode-hz-region (beg end) | |
87 "Decode HZ/ZW encoded text in the current region. | |
88 Return the length of resulting text." | |
89 (interactive "r") | |
90 (save-excursion | |
91 (save-restriction | |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
92 (let (pos ch) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
93 (narrow-to-region beg end) |
17052 | 94 |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
95 ;; We, at first, convert HZ/ZW to `euc-china', |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
96 ;; then decode it. |
17052 | 97 |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
98 ;; "~\n" -> "\n", "~~" -> "~" |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
99 (goto-char (point-min)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
100 (while (search-forward "~" nil t) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
101 (setq ch (following-char)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
102 (if (or (= ch ?\n) (= ch ?~)) (delete-char -1))) |
17052 | 103 |
20738
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
104 ;; "^zW...\n" -> Chinese GB2312 |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
105 ;; "~{...~}" -> Chinese GB2312 |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
106 (goto-char (point-min)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
107 (setq beg nil) |
17052 | 108 (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
|
109 (setq pos (match-beginning 0) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
110 ch (char-after pos)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
111 ;; Record the first position to start conversion. |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
112 (or beg (setq beg pos)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
113 (end-of-line) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
114 (setq end (point)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
115 (if (>= ch 128) ; 8bit GB2312 |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
116 nil |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
117 (goto-char pos) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
118 (delete-char 2) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
119 (setq end (- end 2)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
120 (if (= ch ?z) ; ZW -> euc-china |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
121 (progn |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
122 (translate-region (point) end hz-set-msb-table) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
123 (goto-char end)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
124 (if (search-forward hz-ascii-designnation |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
125 (if decode-hz-line-continuation nil end) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
126 t) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
127 (delete-char -2)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
128 (setq end (point)) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
129 (translate-region pos (point) hz-set-msb-table)))) |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
130 (if beg |
f510736ce3d1
In setup-LANGUAGE-environment functions,
Kenichi Handa <handa@m17n.org>
parents:
19428
diff
changeset
|
131 (decode-coding-region beg end 'euc-china))) |
17052 | 132 (- (point-max) (point-min))))) |
133 | |
134 ;;;###autoload | |
135 (defun decode-hz-buffer () | |
136 "Decode HZ/ZW encoded text in the current buffer." | |
137 (interactive) | |
138 (decode-hz-region (point-min) (point-max))) | |
139 | |
140 ;;;###autoload | |
141 (defun encode-hz-region (beg end) | |
142 "Encode the text in the current region to HZ. | |
143 Return the length of resulting text." | |
144 (interactive "r") | |
145 (save-excursion | |
146 (save-restriction | |
147 (narrow-to-region beg end) | |
148 | |
149 ;; "~" -> "~~" | |
150 (goto-char (point-min)) | |
151 (while (search-forward "~" nil t) (insert ?~)) | |
152 | |
153 ;; Chinese GB2312 -> "~{...~}" | |
154 (goto-char (point-min)) | |
155 (if (re-search-forward "\\cc" nil t) | |
20838
3d67d591c066
(encode-hz-region): Do not bind
Kenichi Handa <handa@m17n.org>
parents:
20738
diff
changeset
|
156 (let (pos) |
17052 | 157 (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
|
158 (encode-coding-region pos (point-max) 'iso-2022-7bit) |
17052 | 159 (goto-char pos) |
160 (while (search-forward iso2022-gb-designation nil t) | |
161 (delete-char -3) | |
162 (insert hz-gb-designnation)) | |
163 (goto-char pos) | |
164 (while (search-forward iso2022-ascii-designation nil t) | |
165 (delete-char -3) | |
166 (insert hz-ascii-designnation)))) | |
167 (- (point-max) (point-min))))) | |
168 | |
169 ;;;###autoload | |
170 (defun encode-hz-buffer () | |
171 "Encode the text in the current buffer to HZ." | |
172 (interactive) | |
173 (encode-hz-region (point-min) (point-max))) | |
174 | |
175 ;; | |
18309
bd8b521f5218
Provide XXX-util instead of
Kenichi Handa <handa@m17n.org>
parents:
17993
diff
changeset
|
176 (provide 'china-util) |
17052 | 177 |
178 ;;; china-util.el ends here |