17301
|
1 ;;; language/tibet-util.el -- Support for inputting Tibetan characters
|
|
2
|
|
3 ;; Copyright (C) 1995 Free Software Foundation, Inc.
|
|
4 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
|
|
5
|
|
6 ;; Keywords: multilingual, Tibetan
|
|
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 ;; Author: Toru TOMABECHI, <Toru.Tomabechi@orient.unil.ch>
|
|
26
|
|
27 ;; Created: Feb. 17. 1997
|
|
28
|
|
29 ;; History:
|
|
30 ;; 1997.03.13 Modification in treatment of text properties;
|
|
31 ;; Support for some special signs and punctuations.
|
|
32
|
|
33 ;;; Code:
|
|
34
|
|
35 ;;; This function makes a transcription string for
|
|
36 ;;; re-composing a character.
|
|
37
|
|
38 ;;;###autoload
|
|
39 (defun tibetan-tibetan-to-transcription (ch)
|
|
40 "Return a transcription string of Tibetan character CH"
|
|
41 (let ((char ch)
|
|
42 (l (append tibetan-consonant-transcription-alist
|
|
43 tibetan-vowel-transcription-alist
|
|
44 tibetan-precomposed-transcription-alist
|
|
45 tibetan-subjoined-transcription-alist))
|
|
46 decomp-l t-char trans str result)
|
|
47 (if (eq (char-charset char) 'composition)
|
|
48 (setq decomp-l (decompose-composite-char char 'list nil))
|
|
49 (setq decomp-l (cons char nil)))
|
|
50 (setq str "")
|
|
51 (while decomp-l
|
|
52 (setq t-char (char-to-string (car decomp-l)))
|
|
53 (setq trans (car (rassoc t-char l)))
|
|
54 (setq str (concat str trans))
|
|
55 (setq decomp-l (cdr decomp-l)))
|
|
56 (setq result str)))
|
|
57
|
|
58 ;;; This function translates transcription string into a string of
|
|
59 ;;; Tibetan characters.
|
|
60
|
|
61 ;;;###autoload
|
|
62 (defun tibetan-transcription-to-tibetan (transcription)
|
|
63 "Translate Roman transcription into a sequence of Tibetan components."
|
|
64 (let ((trans transcription)
|
|
65 (lp tibetan-precomposed-transcription-alist)
|
|
66 (l (append tibetan-consonant-transcription-alist
|
|
67 tibetan-vowel-transcription-alist
|
|
68 tibetan-subjoined-transcription-alist))
|
|
69 (case-fold-search nil)
|
|
70 substr t-char p-str t-str result)
|
|
71 (setq substr "")
|
|
72 (setq p-str "")
|
|
73 (setq t-str "")
|
|
74 (cond ((string-match tibetan-precomposed-regexp trans)
|
|
75 (setq substr (substring trans (match-beginning 0) (match-end 0)))
|
|
76 (setq trans (substring trans (match-end 0)))
|
|
77 (setq t-char (cdr (assoc substr lp)))
|
|
78 (setq p-str t-char)))
|
|
79 (while (string-match tibetan-regexp trans)
|
|
80 (setq substr (substring trans (match-beginning 0) (match-end 0)))
|
|
81 (setq trans (substring trans 0 (match-beginning 0)))
|
|
82 (setq t-char
|
|
83 (cdr (assoc substr l)))
|
|
84 (setq t-str (concat t-char t-str)))
|
|
85 (setq result (concat p-str t-str))))
|
|
86
|
|
87
|
|
88 ;;;
|
|
89 ;;; Functions for composing Tibetan character.
|
|
90 ;;;
|
|
91 ;;; A Tibetan syllable is typically structured as follows:
|
|
92 ;;;
|
|
93 ;;; [Prefix] C [C+] V [M] [Suffix [Post suffix]]
|
|
94 ;;;
|
|
95 ;;; where C's are all vertically stacked, V appears below or above
|
|
96 ;;; consonant cluster and M is always put above the C[C+]V combination.
|
|
97 ;;; (Sanskrit visarga, though it is a vowel modifier, is considered
|
|
98 ;;; to be a punctuation.)
|
|
99 ;;;
|
|
100 ;;; Here are examples of the words "bsgrubs" and "h'uM"
|
|
101 ;;;
|
|
102 ;;; $(7"72%q`"U1"7"G(B 2$(7"H`#A`"U0"_1(B
|
|
103 ;;;
|
|
104 ;;; M
|
|
105 ;;; b s b s h
|
|
106 ;;; g '
|
|
107 ;;; r u
|
|
108 ;;; u
|
|
109 ;;;
|
|
110 ;;; Consonants ''', 'w', 'y', 'r' take special forms when they are used
|
|
111 ;;; as subjoined consonant. Consonant 'r' takes another special form
|
|
112 ;;; when used as superjoined as in "rka", and so on, while it does not
|
|
113 ;;; change its form when conjoined with subjoined ''', 'w' or 'y'
|
|
114 ;;; as in "rwa", "rya".
|
|
115 ;;;
|
|
116 ;;;
|
|
117 ;;; As a Tibetan input method should avoid using conversion key,
|
|
118 ;;; we use a "Tibetan glyph -> transcription -> Tibetan glyph"
|
|
119 ;;; translation at each key input.
|
|
120 ;;;
|
|
121 ;;; 1st stage - Check the preceding char.
|
|
122 ;;; If the preceding char is Tibetan and composable, then
|
|
123 ;;;
|
|
124 ;;; 2nd stage - Translate the preceding char into transcription
|
|
125 ;;;
|
|
126 ;;; 3rd stage - Concatenate the transcription of preceding char
|
|
127 ;;; and the current input key.
|
|
128 ;;;
|
|
129 ;;; 4th stage - Re-translate the concatenated transcription into
|
|
130 ;;; a sequence of Tibetan letters.
|
|
131 ;;;
|
|
132 ;;; 5th stage - Convert leading consonants into one single precomposed char
|
|
133 ;;; if possible.
|
|
134 ;;;
|
|
135 ;;; 6th stage - Compose the consonants into one composite glyph.
|
|
136 ;;;
|
|
137 ;;; (If the current input is a vowel sign or a vowel modifier,
|
|
138 ;;; then it is composed with preceding char without checking
|
|
139 ;;; except when the preceding char is a punctuation or a digit.)
|
|
140 ;;;
|
|
141 ;;;
|
|
142
|
|
143 ;;; This function is used to avoid composition
|
|
144 ;;; between Tibetan and non-Tibetan chars.
|
|
145
|
|
146 ;;;###autoload
|
|
147 (defun tibetan-char-examin (ch)
|
|
148 "Check if char CH is Tibetan character.
|
|
149 Returns non-nil if CH is Tibetan. Otherwise, returns nil."
|
|
150 (let ((chr ch))
|
|
151 (if (eq (char-charset chr) 'composition)
|
|
152 (string-match "\\cq+" (decompose-composite-char chr))
|
|
153 (string-match "\\cq" (char-to-string chr)))))
|
|
154
|
|
155 ;;; This is used to avoid composition between digits, signs, punctuations
|
|
156 ;;; and word constituents.
|
|
157
|
|
158 ;;;###autoload
|
|
159 (defun tibetan-composable-examin (ch)
|
|
160 "Check if Tibetan char CH is composable.
|
|
161 Returns t if CH is a composable char \(i.e. neither punctuation nor digit)."
|
|
162 (let ((chr ch)
|
|
163 chstr)
|
|
164 (if (eq (char-charset chr) 'composition)
|
|
165 (setq chstr (decompose-composite-char chr))
|
|
166 (setq chstr (char-to-string chr)))
|
|
167 (not (string-match "[$(7!1(B-$(7!o"f$(8!;!=!?!@!A!D"`(B]" chstr))))
|
|
168
|
|
169
|
|
170 ;;; This checks if a character to be composed contains already
|
|
171 ;;; one or more vowels / vowel modifiers. If the character contains
|
|
172 ;;; them, then no more consonant should be added.
|
|
173
|
|
174 ;;;###autoload
|
|
175 (defun tibetan-complete-char-examin (ch)
|
|
176 "Check if composite char CH contains one or more vowel/vowel modifiers.
|
|
177 Returns non-nil, if CH contains vowel/vowel modifiers."
|
|
178 (let ((chr ch)
|
|
179 chstr)
|
|
180 (if (eq (char-charset chr) 'composition)
|
|
181 (setq chstr (decompose-composite-char chr))
|
|
182 (setq chstr (char-to-string chr)))
|
|
183 (string-match "[$(7!g!e"Q(B-$(7"^"_(B-$(7"l(B]" chstr)))
|
|
184
|
|
185 ;;; This function makes a composite character consisting of two characters
|
|
186 ;;; vertically stacked.
|
|
187
|
|
188 ;;;###autoload
|
|
189 (defun tibetan-vertical-stacking (first second upward)
|
|
190 "Return a vertically stacked composite char consisting of FIRST and SECOND.
|
|
191 If UPWARD is non-nil, then SECOND is put above FIRST."
|
|
192 (if upward
|
|
193 (compose-chars first '(tc . bc) second)
|
|
194 (compose-chars first '(bc . tc) second)))
|
|
195
|
|
196 ;;; This function makes a composite char from a string.
|
|
197 ;;; Note that this function returns a string, not a char.
|
|
198
|
|
199 ;;;###autoload
|
|
200 (defun tibetan-compose-string (str)
|
|
201 "Compose a sequence of Tibetan character components into a composite character.
|
|
202 Returns a string containing a composite character."
|
|
203 (let ((t-str str)
|
|
204 f-str s-str f-ch s-ch rest composed result)
|
|
205 ;;Make sure no redundant vowel sign is present.
|
|
206 (if (string-match
|
|
207 "^\\(.+\\)\\($(7"Q(B\\)\\([$(7!I!g!e"Q(B-$(7"^"_(B-$(7"l(B]\\)" t-str)
|
|
208 (setq t-str (concat
|
|
209 (match-string 1 t-str)
|
|
210 (match-string 3 t-str))))
|
|
211 (if (string-match
|
|
212 "^\\(.+\\)\\([$(7!I!g!e"Q(B-$(7"^"_(B-$(7"l(B]\\)\\($(7"Q(B\\)" t-str)
|
|
213 (setq t-str (concat
|
|
214 (match-string 1 t-str)
|
|
215 (match-string 2 t-str))))
|
|
216 ;;Start conversion.
|
|
217 (setq result "")
|
|
218 ;; Consecutive base/precomposed consonants are reduced to the last one.
|
|
219 (while (string-match "^\\([$(7"!(B-$(7"J$!(B-$(7%u(B]\\)\\([$(7"!(B-$(7"@"B(B-$(7"J$!(B-$(7%u(B].*\\)" t-str)
|
|
220 (setq result (concat result (match-string 1 t-str)))
|
|
221 (setq t-str (match-string 2 t-str)))
|
|
222 ;; Vowel/vowel modifier, subjoined consonants are added one by one
|
|
223 ;; to the preceding element.
|
|
224 (while
|
|
225 (string-match "^\\(.\\)\\([$(7"A#!(B-$(7#J!I!g!e"Q(B-$(7"^"_(B-$(7"l(B]\\)\\(.*\\)" t-str)
|
|
226 (setq f-str (match-string 1 t-str))
|
|
227 (setq f-ch (string-to-char f-str))
|
|
228 (setq s-str (match-string 2 t-str))
|
|
229 ;;Special treatment for 'a chung.
|
|
230 ;;If 'a follows a consonant, then turned into its subjoined form.
|
|
231 (if (and (string-match "$(7"A(B" s-str)
|
|
232 (not (tibetan-complete-char-examin f-ch)))
|
|
233 (setq s-str "$(7#A(B"))
|
|
234 (setq s-ch (string-to-char s-str))
|
|
235 (setq rest (match-string 3 t-str))
|
|
236 (cond ((string-match "\\c2" s-str);; upper vowel sign
|
|
237 (setq composed
|
|
238 (tibetan-vertical-stacking f-ch s-ch t)))
|
|
239 ((string-match "\\c3" s-str);; lower vowel sign
|
|
240 (setq composed
|
|
241 (tibetan-vertical-stacking f-ch s-ch nil)))
|
|
242 ;;Automatic conversion of ra-mgo (superscribed r).
|
|
243 ;;'r' is converted if followed by a subjoined consonant
|
|
244 ;;other than w, ', y, r.
|
|
245 ((and (string-match "$(7"C(B" f-str)
|
|
246 (not (string-match "[$(7#>#A#B#C(B]" s-str)))
|
|
247 (setq f-ch ?$(7#P(B)
|
|
248 (setq composed
|
|
249 (tibetan-vertical-stacking f-ch s-ch nil)))
|
|
250 ((not (tibetan-complete-char-examin f-ch))
|
|
251 ;;Initial base consonant is tranformed, if followed by
|
|
252 ;;a subjoined consonant, except when it is followed
|
|
253 ;;by a subscribed 'a.
|
|
254 (if (and (string-match "[$(7"!(B-$(7"="?"@"D(B-$(7"J(B]" f-str)
|
|
255 (not (string-match "$(7#A(B" s-str)))
|
|
256 (setq f-ch
|
|
257 (string-to-char
|
|
258 (cdr (assoc f-str tibetan-base-to-subjoined-alist)))))
|
|
259 (setq composed
|
|
260 (tibetan-vertical-stacking f-ch s-ch nil)))
|
|
261 (t
|
|
262 (setq composed s-str)
|
|
263 (setq result (concat result f-str))))
|
|
264 (setq t-str (concat composed rest)))
|
|
265 (setq result (concat result t-str))))
|
|
266
|
|
267 ;;; quail <-> conversion interface.
|
|
268
|
|
269 (defun tibetan-composition (pc key)
|
|
270 "Interface to quail input method.
|
|
271 Takes two arguments: char PC and string KEY, where PC is the preceding
|
|
272 character to be composed with current input KEY.
|
|
273 Returns a string which is the result of composition."
|
|
274 (let (trans cur-ch t-str result)
|
|
275 ;; Make a tibetan character corresponding to current input key.
|
|
276 (setq cur-ch (tibetan-transcription-to-tibetan key))
|
|
277 ;; Check if the preceding character is Tibetan and composable.
|
|
278 (cond ((and (tibetan-char-examin pc)
|
|
279 (tibetan-composable-examin pc))
|
|
280 ;;If Tibetan char corresponding to the current input key exists,
|
|
281 (cond (cur-ch
|
|
282 ;; Then,
|
|
283 ;; Convert the preceding character into transcription,
|
|
284 ;; and concatenate it with the current input key,
|
|
285 (setq trans (tibetan-tibetan-to-transcription pc))
|
|
286 (setq trans (concat trans key))
|
|
287 ;; Concatenated transcription is converted to
|
|
288 ;; a sequence of Tibetan characters,
|
|
289 (setq t-str (tibetan-transcription-to-tibetan trans))
|
|
290 ;; And it is composed into a composite character.
|
|
291 (setq result (tibetan-compose-string t-str)))
|
|
292 ;; Else,
|
|
293 (t
|
|
294 ;; Simply concatenate the preceding character and
|
|
295 ;; the current input key.
|
|
296 (setq result (char-to-string pc))
|
|
297 (setq result (concat result key)))))
|
|
298 ;; If the preceding char is not Tibetan or not composable,
|
|
299 (t
|
|
300 ;; pc = 0 means the point is at the beginning of buffer.
|
|
301 (if (not (eq pc 0))
|
|
302 (setq result (char-to-string pc)))
|
|
303 (if cur-ch
|
|
304 (setq result (concat result cur-ch))
|
|
305 (setq result (concat result key))))
|
|
306 )))
|
|
307
|
|
308
|
|
309 ;;;###autoload
|
|
310 (defun tibetan-decompose-region (beg end)
|
|
311 "Decompose Tibetan characters in the region BEG END into their components.
|
|
312 Components are: base and subjoined consonants, vowel signs, vowel modifiers.
|
|
313 One column punctuations are converted to their 2 column equivalents."
|
|
314 (interactive "r")
|
|
315 (let (ch-str ch-beg ch-end)
|
|
316 (save-excursion
|
|
317 (save-restriction
|
|
318 (narrow-to-region beg end)
|
|
319 (goto-char (point-min))
|
|
320 ;; \\cq = Tibetan character
|
|
321 (while (re-search-forward "\\cq" nil t)
|
|
322 (setq ch-str (buffer-substring-no-properties
|
|
323 (match-beginning 0) (match-end 0)))
|
|
324 ;; Save the points. Maybe, using save-match-data is preferable.
|
|
325 ;; But in order not to lose the trace(because the body is too long),
|
|
326 ;; we save the points in variables.
|
|
327 (setq ch-beg (match-beginning 0))
|
|
328 (setq ch-end (match-end 0))
|
|
329 ;; Here starts the decomposition.
|
|
330 (cond
|
|
331 ;; 1 column punctuations -> 2 column equivalent
|
|
332 ((string-match "[$(8!D!;!=!?!@!A"`(B]" ch-str)
|
|
333 (setq ch-str
|
|
334 (car (rassoc ch-str tibetan-precomposition-rule-alist))))
|
|
335 ;; Decomposition of composite character.
|
|
336 ((eq (char-charset (string-to-char ch-str)) 'composition)
|
|
337 ;; Make a string which consists of a sequence of
|
|
338 ;; components.
|
|
339 (setq ch-str (decompose-composite-char (string-to-char ch-str)))
|
|
340 ;; Converts nyi zla into base elements.
|
|
341 (cond ((string= ch-str "$(7#R#S#S#S(B")
|
|
342 (setq ch-str "$(7!4!5!5(B"))
|
|
343 ((string= ch-str "$(7#R#S#S(B")
|
|
344 (setq ch-str "$(7!4!5(B"))
|
|
345 ((string= ch-str "$(7#R#S!I(B")
|
|
346 (setq ch-str "$(7!6(B"))
|
|
347 ((string= ch-str "$(7#R#S(B")
|
|
348 (setq ch-str "$(7!4(B")))))
|
|
349 ;; If the sequence of components starts with a subjoined consonants,
|
|
350 (if (string-match "^\\([$(7#!(B-$(7#J(B]\\)\\(.*\\)$" ch-str)
|
|
351 ;; then the first components is converted to its base form.
|
|
352 (setq ch-str
|
|
353 (concat (car (rassoc (match-string 1 ch-str)
|
|
354 tibetan-base-to-subjoined-alist))
|
|
355 (match-string 2 ch-str))))
|
|
356 ;; If the sequence of components starts with a precomposed character,
|
|
357 (if (string-match "^\\([$(7$!(B-$(7%u(B]\\)\\(.*\\)$" ch-str)
|
|
358 ;; then it is converted into a sequence of components.
|
|
359 (setq ch-str
|
|
360 (concat (car (rassoc (match-string 1 ch-str)
|
|
361 tibetan-precomposition-rule-alist))
|
|
362 (match-string 2 ch-str))))
|
|
363 ;; Special treatment for superscribed r.
|
|
364 (if (string-match "^$(7#P(B\\(.*\\)$" ch-str)
|
|
365 (setq ch-str (concat "$(7"C(B" (match-string 1 ch-str))))
|
|
366 ;; Finally, the result of decomposition is inserted, and
|
|
367 ;; the composite character is deleted.
|
|
368 (insert-and-inherit ch-str)
|
|
369 (delete-region ch-beg ch-end))))))
|
|
370
|
|
371 ;;;###autoload
|
|
372 (defun tibetan-compose-region (beg end)
|
|
373 "Make composite chars from Tibetan character components in the region BEG END.
|
|
374 Two column punctuations are converted to their 1 column equivalents."
|
|
375 (interactive "r")
|
|
376 (let (str result)
|
|
377 (save-excursion
|
|
378 (save-restriction
|
|
379 (narrow-to-region beg end)
|
|
380 (goto-char (point-min))
|
|
381 ;; First, sequence of components which has a precomposed equivalent
|
|
382 ;; is converted.
|
|
383 (while (re-search-forward
|
|
384 tibetan-precomposition-rule-regexp nil t)
|
|
385 (setq str (buffer-substring-no-properties
|
|
386 (match-beginning 0) (match-end 0)))
|
|
387 (save-match-data
|
|
388 (insert-and-inherit
|
|
389 (cdr (assoc str tibetan-precomposition-rule-alist))))
|
|
390 (delete-region (match-beginning 0) (match-end 0)))
|
|
391 (goto-char (point-min))
|
|
392 ;; Then, composable elements are put into a composite character.
|
|
393 (while (re-search-forward
|
|
394 "[$(7"!(B-$(7"J$!(B-$(7%u(B]+[$(7#!(B-$(7#J!I!g!e"Q(B-$(7"^"_(B-$(7"l(B]+"
|
|
395 nil t)
|
|
396 (setq str (buffer-substring-no-properties
|
|
397 (match-beginning 0) (match-end 0)))
|
|
398 (save-match-data
|
|
399 (setq result (tibetan-compose-string str))
|
|
400 (insert-and-inherit result))
|
|
401 (delete-region (match-beginning 0) (match-end 0)))))))
|
|
402
|
|
403 ;;;
|
|
404 ;;; This variable is used to avoid repeated decomposition.
|
|
405 ;;;
|
|
406 (setq-default tibetan-decomposed nil)
|
|
407
|
|
408 ;;;###autoload
|
|
409 (defun tibetan-decompose-buffer ()
|
|
410 "Decomposes Tibetan characters in the buffer into their components.
|
|
411 See also docstring of the function tibetan-decompose-region."
|
|
412 (interactive)
|
|
413 (make-local-variable 'tibetan-decomposed)
|
|
414 (cond ((not tibetan-decomposed)
|
|
415 (tibetan-decompose-region (point-min) (point-max))
|
|
416 (setq tibetan-decomposed t))))
|
|
417
|
|
418 ;;;###autoload
|
|
419 (defun tibetan-compose-buffer ()
|
|
420 "Composes Tibetan character components in the buffer.
|
|
421 See also docstring of the function tibetan-compose-region."
|
|
422 (interactive)
|
|
423 (make-local-variable 'tibetan-decomposed)
|
|
424 (tibetan-compose-region (point-min) (point-max))
|
|
425 (setq tibetan-decomposed nil))
|
|
426
|
|
427 ;;;###autoload
|
|
428 (defun tibetan-post-read-conversion (len)
|
|
429 (save-excursion
|
|
430 (save-restriction
|
|
431 (let ((buffer-modified-p (buffer-modified-p)))
|
|
432 (narrow-to-region (point) (+ (point) len))
|
|
433 (tibetan-compose-region (point-min) (point-max))
|
|
434 (set-buffer-modified-p buffer-modified-p)
|
|
435 (point-max))))
|
|
436 (make-local-variable 'tibetan-decomposed)
|
|
437 (setq tibetan-decomposed nil))
|
|
438
|
|
439
|
|
440 ;;;###autoload
|
|
441 (defun tibetan-pre-write-conversion (from to)
|
|
442 (setq tibetan-decomposed-temp tibetan-decomposed)
|
|
443 (let ((old-buf (current-buffer))
|
|
444 (work-buf (get-buffer-create " *tibetan-work*")))
|
|
445 (set-buffer work-buf)
|
|
446 (erase-buffer)
|
|
447 (insert-buffer-substring old-buf from to)
|
|
448 (if (not tibetan-decomposed-temp)
|
|
449 (tibetan-decompose-region (point-min) (point-max)))))
|
|
450
|
|
451 (provide 'language/tibet-util)
|
|
452
|
|
453 ;;; language/tibet-util.el ends here.
|