Mercurial > hg > xemacs-beta

comparison lisp/language/viet-util.el @ 197:acd284d43ca1 r20-3b25

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Import from CVS: tag r20-3b25
author cvs
date Mon, 13 Aug 2007 10:00:02 +0200
parents
children
comparison
equal deleted inserted replaced
196:58e0786448ca 197:acd284d43ca1
1 ;;; viet-util.el --- utilities for Vietnamese
2
3 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
4 ;; Licensed to the Free Software Foundation.
5
6 ;; Keywords: mule, multilingual, Vietnamese
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
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.
24
25 ;;; Commentary:
26
27 ;; Vietnamese uses ASCII characters and additional 134 unique
28 ;; characters (these are Latin alphabets with various diacritical and
29 ;; tone marks). As far as I know, Vietnamese now has 4 different ways
30 ;; for representing these characters: VISCII, VSCII, VIQR, and
31 ;; Unicode. VISCII and VSCII are simple 1-byte code which assigns 134
32 ;; unique characters in control-code area (0x00..0x1F) and right half
33 ;; area (0x80..0xFF). VIQR is a menmonic encoding specification
34 ;; representing diacritical marks by following ASCII characters.
35
36 ;;; Code:
37
38 ;;;###autoload
39 (defun setup-vietnamese-environment ()
40 "Setup multilingual environment (MULE) for Vietnamese VISCII users."
41 (interactive)
42 (setup-8-bit-environment "Vietnamese" nil 'vietnamese-viscii
43 "vietnamese-viqr")
44 (setq coding-category-raw-text 'vietnamese-viscii))
45
46 ;; VIQR is a menmonic encoding specification for Vietnamese.
47 ;; It represents diacritical marks by ASCII characters as follows:
48 ;; ------------+----------+--------
49 ;; mark | mnemonic | example
50 ;; ------------+----------+---------
51 ;; breve | ( | a( -> ,1e(B
52 ;; circumflex | ^ | a^ -> ,1b(B
53 ;; horn | + | o+ -> ,1=(B
54 ;; ------------+----------+---------
55 ;; acute | ' | a' -> ,1a(B
56 ;; grave | ` | a` -> ,1`(B
57 ;; hook above | ? | a? -> ,1d(B
58 ;; tilde | ~ | a~ -> ,1c(B
59 ;; dot below | . | a. -> ,1U(B
60 ;; ------------+----------+---------
61 ;; d bar | dd | dd -> ,1p(B
62 ;; ------------+----------+---------
63
64 (defvar viet-viqr-alist
65 '(;; lowercase
66 (?,1!(B . "a('") ; 161
67 (?,1"(B . "a(`") ; 162
68 (?,1#(B . "a(.") ; 163
69 (?,1$(B . "a^'") ; 164
70 (?,1%(B . "a^`") ; 165
71 (?,1&(B . "a^?") ; 166
72 (?,1'(B . "a^.") ; 167
73 (?,1((B . "e~") ; 168
74 (?,1)(B . "e.") ; 169
75 (?,1*(B . "e^'") ; 170
76 (?,1+(B . "e^`") ; 171
77 (?,1,(B . "e^?") ; 172
78 (?,1-(B . "e^~") ; 173
79 (?,1.(B . "e^.") ; 174
80 (?,1/(B . "o^'") ; 175
81 (?,10(B . "o^`") ; 176
82 (?,11(B . "o^?") ; 177
83 (?,12(B . "o^~") ; 178
84 (?,15(B . "o^.") ; 181
85 (?,16(B . "o+`") ; 182
86 (?,17(B . "o+?") ; 183
87 (?,18(B . "i.") ; 184
88 (?,1=(B . "o+") ; 189
89 (?,1>(B . "o+'") ; 190
90 (?,1F(B . "a(?") ; 198
91 (?,1G(B . "a(~") ; 199
92 (?,1O(B . "y`") ; 207
93 (?,1Q(B . "u+'") ; 209
94 (?,1U(B . "a.") ; 213
95 (?,1V(B . "y?") ; 214
96 (?,1W(B . "u+`") ; 215
97 (?,1X(B . "u+?") ; 216
98 (?,1[(B . "y~") ; 219
99 (?,1\(B . "y.") ; 220
100 (?,1^(B . "o+~") ; 222
101 (?,1_(B . "u+") ; 223
102 (?,1`(B . "a`") ; 224
103 (?,1a(B . "a'") ; 225
104 (?,1b(B . "a^") ; 226
105 (?,1c(B . "a~") ; 227
106 (?,1d(B . "a?") ; 228
107 (?,1e(B . "a(") ; 229
108 (?,1f(B . "u+~") ; 230
109 (?,1g(B . "a^~") ; 231
110 (?,1h(B . "e`") ; 232
111 (?,1i(B . "e'") ; 233
112 (?,1j(B . "e^") ; 234
113 (?,1k(B . "e?") ; 235
114 (?,1l(B . "i`") ; 236
115 (?,1m(B . "i'") ; 237
116 (?,1n(B . "i~") ; 238
117 (?,1o(B . "i?") ; 239
118 (?,1p(B . "dd") ; 240
119 (?,1q(B . "u+.") ; 241
120 (?,1r(B . "o`") ; 242
121 (?,1s(B . "o'") ; 243
122 (?,1t(B . "o^") ; 244
123 (?,1u(B . "o~") ; 245
124 (?,1v(B . "o?") ; 246
125 (?,1w(B . "o.") ; 247
126 (?,1x(B . "u.") ; 248
127 (?,1y(B . "u`") ; 249
128 (?,1z(B . "u'") ; 250
129 (?,1{(B . "u~") ; 251
130 (?,1|(B . "u?") ; 252
131 (?,1}(B . "y'") ; 253
132 (?,1~(B . "o+.") ; 254
133
134 ;; upper case
135 (?,2!(B . "A('") ; 161
136 (?,2"(B . "A(`") ; 162
137 (?,2#(B . "A(.") ; 163
138 (?,2$(B . "A^'") ; 164
139 (?,2%(B . "A^`") ; 165
140 (?,2&(B . "A^?") ; 166
141 (?,2'(B . "A^.") ; 167
142 (?,2((B . "E~") ; 168
143 (?,2)(B . "E.") ; 169
144 (?,2*(B . "E^'") ; 170
145 (?,2+(B . "E^`") ; 171
146 (?,2,(B . "E^?") ; 172
147 (?,2-(B . "E^~") ; 173
148 (?,2.(B . "E^.") ; 174
149 (?,2/(B . "O^'") ; 175
150 (?,20(B . "O^`") ; 176
151 (?,21(B . "O^?") ; 177
152 (?,22(B . "O^~") ; 178
153 (?,25(B . "O^.") ; 181
154 (?,26(B . "O+`") ; 182
155 (?,27(B . "O+?") ; 183
156 (?,28(B . "I.") ; 184
157 (?,2=(B . "O+") ; 189
158 (?,2>(B . "O+'") ; 190
159 (?,2F(B . "A(?") ; 198
160 (?,2G(B . "A(~") ; 199
161 (?,2O(B . "Y`") ; 207
162 (?,2Q(B . "U+'") ; 209
163 (?,2U(B . "A.") ; 213
164 (?,2V(B . "Y?") ; 214
165 (?,2W(B . "U+`") ; 215
166 (?,2X(B . "U+?") ; 216
167 (?,2[(B . "Y~") ; 219
168 (?,2\(B . "Y.") ; 220
169 (?,2^(B . "O+~") ; 222
170 (?,2_(B . "U+") ; 223
171 (?,2`(B . "A`") ; 224
172 (?,2a(B . "A'") ; 225
173 (?,2b(B . "A^") ; 226
174 (?,2c(B . "A~") ; 227
175 (?,2d(B . "A?") ; 228
176 (?,2e(B . "A(") ; 229
177 (?,2f(B . "U+~") ; 230
178 (?,2g(B . "A^~") ; 231
179 (?,2h(B . "E`") ; 232
180 (?,2i(B . "E'") ; 233
181 (?,2j(B . "E^") ; 234
182 (?,2k(B . "E?") ; 235
183 (?,2l(B . "I`") ; 236
184 (?,2m(B . "I'") ; 237
185 (?,2n(B . "I~") ; 238
186 (?,2o(B . "I?") ; 239
187 (?,2p(B . "DD") ; 240
188 (?,2p(B . "dD") ; 240
189 (?,2p(B . "Dd") ; 240
190 (?,2q(B . "U+.") ; 241
191 (?,2r(B . "O`") ; 242
192 (?,2s(B . "O'") ; 243
193 (?,2t(B . "O^") ; 244
194 (?,2u(B . "O~") ; 245
195 (?,2v(B . "O?") ; 246
196 (?,2w(B . "O.") ; 247
197 (?,2x(B . "U.") ; 248
198 (?,2y(B . "U`") ; 249
199 (?,2z(B . "U'") ; 250
200 (?,2{(B . "U~") ; 251
201 (?,2|(B . "U?") ; 252
202 (?,2}(B . "Y'") ; 253
203 (?,2~(B . "O+.") ; 254
204
205 ;; escape from composition
206 (?\( . "\\(") ; breve (left parenthesis)
207 (?^ . "\\^") ; circumflex (caret)
208 (?+ . "\\+") ; horn (plus sign)
209 (?' . "\\'") ; acute (apostrophe)
210 (?` . "\\`") ; grave (backquote)
211 (?? . "\\?") ; hook above (question mark)
212 (?~ . "\\~") ; tilde (tilde)
213 (?. . "\\.") ; dot below (period)
214 (?d . "\\d") ; d-bar (d)
215 (?\\ . "\\\\") ; literal backslash
216 )
217 "Alist of Vietnamese characters vs corresponding `VIQR' string.")
218
219 ;; Regular expression matching single Vietnamese character represented
220 ;; by VIQR.
221 (defconst viqr-regexp
222 "[aeiouyAEIOUY]\\([(^+]?['`?~.]\\|[(^+]\\)\\|[Dd][Dd]")
223
224 ;;;###autoload
225 (defun viet-decode-viqr-region (from to)
226 "Convert `VIQR' mnemonics of the current region to Vietnamese characaters.
227 When called from a program, expects two arguments,
228 positions (integers or markers) specifying the stretch of the region."
229 (interactive "r")
230 (save-restriction
231 (narrow-to-region from to)
232 (goto-char (point-min))
233 (while (re-search-forward viqr-regexp nil t)
234 (let* ((viqr (buffer-substring (match-beginning 0) (match-end 0)))
235 (ch (car (rassoc viqr viet-viqr-alist))))
236 (if ch
237 (progn
238 (delete-region (match-beginning 0) (match-end 0))
239 (insert ch)))))))
240
241 ;;;###autoload
242 (defun viet-decode-viqr-buffer ()
243 "Convert `VIQR' mnemonics of the current buffer to Vietnamese characaters."
244 (interactive)
245 (viet-decode-viqr-region (point-min) (point-max)))
246
247 ;;;###autoload
248 (defun viet-encode-viqr-region (from to)
249 "Convert Vietnamese characaters of the current region to `VIQR' mnemonics.
250 When called from a program, expects two arguments,
251 positions (integers or markers) specifying the stretch of the region."
252 (interactive "r")
253 (save-restriction
254 (narrow-to-region from to)
255 (goto-char (point-min))
256 (while (re-search-forward "\\cv" nil t)
257 (let* ((ch (preceding-char))
258 (viqr (cdr (assq ch viet-viqr-alist))))
259 (if viqr
260 (progn
261 (delete-char -1)
262 (insert viqr)))))))
263
264 ;;;###autoload
265 (defun viet-encode-viqr-buffer ()
266 "Convert Vietnamese characaters of the current buffer to `VIQR' mnemonics."
267 (interactive)
268 (viet-encode-viqr-region (point-min) (point-max)))
269
270 ;;;###autoload
271 (defun viqr-post-read-conversion (len)
272 (save-excursion
273 (save-restriction
274 (narrow-to-region (point) (+ (point) len))
275 (let ((buffer-modified-p (buffer-modified-p)))
276 (viet-decode-viqr-region (point-min) (point-max))
277 (set-buffer-modified-p buffer-modified-p)
278 (- (point-max) (point-min))))))
279
280 ;;;###autoload
281 (defun viqr-pre-write-conversion (from to)
282 (let ((old-buf (current-buffer))
283 (work-buf (get-buffer-create " *viet-work*")))
284 (set-buffer work-buf)
285 (erase-buffer)
286 (if (stringp from)
287 (insert from)
288 (insert-buffer-substring old-buf from to))
289 (viet-encode-viqr-region (point-min) (point-max))
290 ;; Should return nil as annotations.
291 nil))
292
293 ;;;
294 (provide 'viet-util)
295
296 ;;; viet-util.el ends here