Comment change.
[emacs.git] / lisp / language / japan-util.el
blob6575fdd43b97734d18f98eec03a35e2483061866
1 ;;; japan-util.el --- utilities for Japanese
3 ;; Copyright (C) 1995 Free Software Foundation, Inc.
4 ;; Copyright (C) 1995 Electrotechnical Laboratory, JAPAN.
6 ;; Keywords: mule, multilingual, Japanese
8 ;; This file is part of GNU Emacs.
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.
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.
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.
25 ;;; Code:
27 (defconst japanese-kana-table
28 '((?\e$B$"\e(B ?\e$B%"\e(B ?\e(I1\e(B) (?\e$B$$\e(B ?\e$B%$\e(B ?\e(I2\e(B) (?\e$B$&\e(B ?\e$B%&\e(B ?\e(I3\e(B) (?\e$B$(\e(B ?\e$B%(\e(B ?\e(I4\e(B) (?\e$B$*\e(B ?\e$B%*\e(B ?\e(I5\e(B)
29 (?\e$B$+\e(B ?\e$B%+\e(B ?\e(I6\e(B) (?\e$B$-\e(B ?\e$B%-\e(B ?\e(I7\e(B) (?\e$B$/\e(B ?\e$B%/\e(B ?\e(I8\e(B) (?\e$B$1\e(B ?\e$B%1\e(B ?\e(I9\e(B) (?\e$B$3\e(B ?\e$B%3\e(B ?\e(I:\e(B)
30 (?\e$B$5\e(B ?\e$B%5\e(B ?\e(I;\e(B) (?\e$B$7\e(B ?\e$B%7\e(B ?\e(I<\e(B) (?\e$B$9\e(B ?\e$B%9\e(B ?\e(I=\e(B) (?\e$B$;\e(B ?\e$B%;\e(B ?\e(I>\e(B) (?\e$B$=\e(B ?\e$B%=\e(B ?\e(I?\e(B)
31 (?\e$B$?\e(B ?\e$B%?\e(B ?\e(I@\e(B) (?\e$B$A\e(B ?\e$B%A\e(B ?\e(IA\e(B) (?\e$B$D\e(B ?\e$B%D\e(B ?\e(IB\e(B) (?\e$B$F\e(B ?\e$B%F\e(B ?\e(IC\e(B) (?\e$B$H\e(B ?\e$B%H\e(B ?\e(ID\e(B)
32 (?\e$B$J\e(B ?\e$B%J\e(B ?\e(IE\e(B) (?\e$B$K\e(B ?\e$B%K\e(B ?\e(IF\e(B) (?\e$B$L\e(B ?\e$B%L\e(B ?\e(IG\e(B) (?\e$B$M\e(B ?\e$B%M\e(B ?\e(IH\e(B) (?\e$B$N\e(B ?\e$B%N\e(B ?\e(II\e(B)
33 (?\e$B$O\e(B ?\e$B%O\e(B ?\e(IJ\e(B) (?\e$B$R\e(B ?\e$B%R\e(B ?\e(IK\e(B) (?\e$B$U\e(B ?\e$B%U\e(B ?\e(IL\e(B) (?\e$B$X\e(B ?\e$B%X\e(B ?\e(IM\e(B) (?\e$B$[\e(B ?\e$B%[\e(B ?\e(IN\e(B)
34 (?\e$B$^\e(B ?\e$B%^\e(B ?\e(IO\e(B) (?\e$B$_\e(B ?\e$B%_\e(B ?\e(IP\e(B) (?\e$B$`\e(B ?\e$B%`\e(B ?\e(IQ\e(B) (?\e$B$a\e(B ?\e$B%a\e(B ?\e(IR\e(B) (?\e$B$b\e(B ?\e$B%b\e(B ?\e(IS\e(B)
35 (?\e$B$d\e(B ?\e$B%d\e(B ?\e(IT\e(B) (?\e$B$f\e(B ?\e$B%f\e(B ?\e(IU\e(B) (?\e$B$h\e(B ?\e$B%h\e(B ?\e(IV\e(B)
36 (?\e$B$i\e(B ?\e$B%i\e(B ?\e(IW\e(B) (?\e$B$j\e(B ?\e$B%j\e(B ?\e(IX\e(B) (?\e$B$k\e(B ?\e$B%k\e(B ?\e(IY\e(B) (?\e$B$l\e(B ?\e$B%l\e(B ?\e(IZ\e(B) (?\e$B$m\e(B ?\e$B%m\e(B ?\e(I[\e(B)
37 (?\e$B$o\e(B ?\e$B%o\e(B ?\e(I\\e(B) (?\e$B$p\e(B ?\e$B%p\e(B nil) (?\e$B$q\e(B ?\e$B%q\e(B nil) (?\e$B$r\e(B ?\e$B%r\e(B ?\e(I&\e(B)
38 (?\e$B$s\e(B ?\e$B%s\e(B ?\e(I]\e(B)
39 (?\e$B$,\e(B ?\e$B%,\e(B "\e(I6^\e(B") (?\e$B$.\e(B ?\e$B%.\e(B "\e(I7^\e(B") (?\e$B$0\e(B ?\e$B%0\e(B "\e(I8^\e(B") (?\e$B$2\e(B ?\e$B%2\e(B "\e(I9^\e(B") (?\e$B$4\e(B ?\e$B%4\e(B "\e(I:^\e(B")
40 (?\e$B$6\e(B ?\e$B%6\e(B "\e(I;^\e(B") (?\e$B$8\e(B ?\e$B%8\e(B "\e(I<^\e(B") (?\e$B$:\e(B ?\e$B%:\e(B "\e(I=^\e(B") (?\e$B$<\e(B ?\e$B%<\e(B "\e(I>^\e(B") (?\e$B$>\e(B ?\e$B%>\e(B "\e(I?^\e(B")
41 (?\e$B$@\e(B ?\e$B%@\e(B "\e(I@^\e(B") (?\e$B$B\e(B ?\e$B%B\e(B "\e(IA^\e(B") (?\e$B$E\e(B ?\e$B%E\e(B "\e(IB^\e(B") (?\e$B$G\e(B ?\e$B%G\e(B "\e(IC^\e(B") (?\e$B$I\e(B ?\e$B%I\e(B "\e(ID^\e(B")
42 (?\e$B$P\e(B ?\e$B%P\e(B "\e(IJ^\e(B") (?\e$B$S\e(B ?\e$B%S\e(B "\e(IK^\e(B") (?\e$B$V\e(B ?\e$B%V\e(B "\e(IL^\e(B") (?\e$B$Y\e(B ?\e$B%Y\e(B "\e(IM^\e(B") (?\e$B$\\e(B ?\e$B%\\e(B "\e(IN^\e(B")
43 (?\e$B$Q\e(B ?\e$B%Q\e(B "\e(IJ_\e(B") (?\e$B$T\e(B ?\e$B%T\e(B "\e(IK_\e(B") (?\e$B$W\e(B ?\e$B%W\e(B "\e(IL_\e(B") (?\e$B$Z\e(B ?\e$B%Z\e(B "\e(IM_\e(B") (?\e$B$]\e(B ?\e$B%]\e(B "\e(IN_\e(B")
44 (?\e$B$!\e(B ?\e$B%!\e(B ?\e(I'\e(B) (?\e$B$#\e(B ?\e$B%#\e(B ?\e(I(\e(B) (?\e$B$%\e(B ?\e$B%%\e(B ?\e(I)\e(B) (?\e$B$'\e(B ?\e$B%'\e(B ?\e(I*\e(B) (?\e$B$)\e(B ?\e$B%)\e(B ?\e(I+\e(B)
45 (?\e$B$C\e(B ?\e$B%C\e(B ?\e(I/\e(B)
46 (?\e$B$c\e(B ?\e$B%c\e(B ?\e(I,\e(B) (?\e$B$e\e(B ?\e$B%e\e(B ?\e(I-\e(B) (?\e$B$g\e(B ?\e$B%g\e(B ?\e(I.\e(B)
47 (?\e$B$n\e(B ?\e$B%n\e(B nil)
48 (nil ?\e$B%t\e(B "\e(I3^\e(B") (nil ?\e$B%u\e(B nil) (nil ?\e$B%v\e(B nil))
49 "Japanese JISX0208 Kana character table.
50 Each element is of the form (HIRAGANA KATAKANA HANKAKU-KATAKANA), where
51 HIRAGANA and KATAKANA belong to `japanese-jisx0208',
52 HANKAKU-KATAKANA belongs to `japanese-jisx0201-kana'.")
54 ;; Put properties 'katakana, 'hiragana, and 'jix0201 to each Japanese
55 ;; kana characters for conversion among them.
56 (let ((l japanese-kana-table)
57 slot hiragana katakana jisx0201)
58 (while l
59 (setq slot (car l)
60 hiragana (car slot) katakana (nth 1 slot) jisx0201 (nth 2 slot)
61 l (cdr l))
62 (if hiragana
63 (progn
64 (put-char-code-property hiragana 'katakana katakana)
65 (put-char-code-property katakana 'hiragana hiragana)
66 (if jisx0201
67 (progn
68 (put-char-code-property hiragana 'jisx0201 jisx0201)
69 (if (integerp jisx0201)
70 (put-char-code-property jisx0201 'hiragana hiragana))))))
71 (if jisx0201
72 (progn
73 (put-char-code-property katakana 'jisx0201 jisx0201)
74 (if (integerp jisx0201)
75 (put-char-code-property jisx0201 'katakana katakana))))))
77 (defconst japanese-symbol-table
78 '((?\\e$B!!\e(B ?\ ) (?\e$B!"\e(B ?, ?\e(I$\e(B) (?\e$B!#\e(B ?. ?\e(I!\e(B) (?\e$B!$\e(B ?, ?\e(I$\e(B) (?\e$B!%\e(B ?. ?\e(I!\e(B) (?\e$B!&\e(B nil ?\e(I%\e(B)
79 (?\e$B!'\e(B ?:) (?\e$B!(\e(B ?\;) (?\e$B!)\e(B ??) (?\e$B!*\e(B ?!) (?\e$B!+\e(B nil ?\e(I^\e(B) (?\e$B!,\e(B nil ?\e(I_\e(B)
80 (?\e$B!-\e(B ?') (?\e$B!.\e(B ?`) (?\e$B!0\e(B ?^) (?\e$B!2\e(B ?_) (?\e$B!<\e(B ?-) (?\e$B!=\e(B ?-) (?\e$B!>\e(B ?-)
81 (?\e$B!?\e(B ?/) (?\e$B!@\e(B ?\\) (?\e$B!A\e(B ?~) (?\e$B!C\e(B ?|) (?\e$B!F\e(B ?`) (?\e$B!G\e(B ?') (?\e$B!H\e(B ?\") (?\e$B!I\e(B ?\")
82 (?\\e$B!J\e(B ?\() (?\\e$B!K\e(B ?\)) (?\\e$B!N\e(B ?[) (?\\e$B!O\e(B ?]) (?\\e$B!P\e(B ?{) (?\\e$B!Q\e(B ?})
83 (?\e$B!R\e(B ?<) (?\e$B!S\e(B ?>) (?\e$B!\\e(B ?+) (?\e$B!]\e(B ?-) (?\e$B!a\e(B ?=) (?\e$B!c\e(B ?<) (?\e$B!d\e(B ?>)
84 (?\e$B!l\e(B ?') (?\e$B!m\e(B ?\") (?\e$B!o\e(B ?\\) (?\e$B!p\e(B ?$) (?\e$B!s\e(B ?%) (?\e$B!t\e(B ?#) (?\e$B!u\e(B ?&) (?\e$B!v\e(B ?*)
85 (?\e$B!w\e(B ?@))
86 "Japanese JISX0208 symbol character table.
87 Each element is of the form (SYMBOL ASCII HANKAKU), where SYMBOL
88 belongs to `japanese-jisx0208', ASCII belongs to `ascii', and HANKAKU
89 belongs to `japanese-jisx0201-kana'.")
91 ;; Put properties 'jisx0208, 'jisx0201, and 'ascii to each Japanese
92 ;; symbol and ASCII characters for conversion among them.
93 (let ((l japanese-symbol-table)
94 slot jisx0208 ascii jisx0201)
95 (while l
96 (setq slot (car l)
97 jisx0208 (car slot) ascii (nth 1 slot) jisx0201 (nth 2 slot)
98 l (cdr l))
99 (if ascii
100 (progn
101 (put-char-code-property jisx0208 'ascii ascii)
102 (put-char-code-property ascii 'jisx0208 jisx0208)))
103 (if jisx0201
104 (progn
105 (put-char-code-property jisx0208 'jisx0201 jisx0201)
106 (put-char-code-property jisx0201 'jisx0208 jisx0208)))))
108 (defconst japanese-alpha-numeric-table
109 '((?\e$B#0\e(B . ?0) (?\e$B#1\e(B . ?1) (?\e$B#2\e(B . ?2) (?\e$B#3\e(B . ?3) (?\e$B#4\e(B . ?4)
110 (?\e$B#5\e(B . ?5) (?\e$B#6\e(B . ?6) (?\e$B#7\e(B . ?7) (?\e$B#8\e(B . ?8) (?\e$B#9\e(B . ?9)
111 (?\e$B#A\e(B . ?A) (?\e$B#B\e(B . ?B) (?\e$B#C\e(B . ?C) (?\e$B#D\e(B . ?D) (?\e$B#E\e(B . ?E)
112 (?\e$B#F\e(B . ?F) (?\e$B#G\e(B . ?G) (?\e$B#H\e(B . ?H) (?\e$B#I\e(B . ?I) (?\e$B#J\e(B . ?J)
113 (?\e$B#K\e(B . ?K) (?\e$B#L\e(B . ?L) (?\e$B#M\e(B . ?M) (?\e$B#N\e(B . ?N) (?\e$B#O\e(B . ?O)
114 (?\e$B#P\e(B . ?P) (?\e$B#Q\e(B . ?Q) (?\e$B#R\e(B . ?R) (?\e$B#S\e(B . ?S) (?\e$B#T\e(B . ?T)
115 (?\e$B#U\e(B . ?U) (?\e$B#V\e(B . ?V) (?\e$B#W\e(B . ?W) (?\e$B#X\e(B . ?X) (?\e$B#Y\e(B . ?Y) (?\e$B#Z\e(B . ?Z)
116 (?\e$B#a\e(B . ?a) (?\e$B#b\e(B . ?b) (?\e$B#c\e(B . ?c) (?\e$B#d\e(B . ?d) (?\e$B#e\e(B . ?e)
117 (?\e$B#f\e(B . ?f) (?\e$B#g\e(B . ?g) (?\e$B#h\e(B . ?h) (?\e$B#i\e(B . ?i) (?\e$B#j\e(B . ?j)
118 (?\e$B#k\e(B . ?k) (?\e$B#l\e(B . ?l) (?\e$B#m\e(B . ?m) (?\e$B#n\e(B . ?n) (?\e$B#o\e(B . ?o)
119 (?\e$B#p\e(B . ?p) (?\e$B#q\e(B . ?q) (?\e$B#r\e(B . ?r) (?\e$B#s\e(B . ?s) (?\e$B#t\e(B . ?t)
120 (?\e$B#u\e(B . ?u) (?\e$B#v\e(B . ?v) (?\e$B#w\e(B . ?w) (?\e$B#x\e(B . ?x) (?\e$B#y\e(B . ?y) (?\e$B#z\e(B . ?z))
121 "Japanese JISX0208 alpha numeric character table.
122 Each element is of the form (ALPHA-NUMERIC ASCII), where ALPHA-NUMERIC
123 belongs to `japanese-jisx0208', ASCII belongs to `ascii'.")
125 ;; Put properties 'jisx0208 and 'ascii to each Japanese alpha numeric
126 ;; and ASCII characters for conversion between them.
127 (let ((l japanese-alpha-numeric-table)
128 slot jisx0208 ascii)
129 (while l
130 (setq slot (car l)
131 jisx0208 (car slot) ascii (cdr slot)
132 l (cdr l))
133 (put-char-code-property jisx0208 'ascii ascii)
134 (put-char-code-property ascii 'jisx0208 jisx0208)))
136 ;; Convert string STR by FUNC and return a resulting string.
137 (defun japanese-string-conversion (str func &rest args)
138 (let ((buf (get-buffer-create " *Japanese work*")))
139 (save-excursion
140 (set-buffer buf)
141 (erase-buffer)
142 (insert str)
143 (apply func 1 (point) args)
144 (buffer-string))))
146 ;;;###autoload
147 (defun japanese-katakana (obj &optional hankaku)
148 "Convert argument to Katakana and return that.
149 The argument may be a character or string. The result has the same type.
150 The argument object is not altered--the value is a copy.
151 Optional argument HANKAKU t means to convert to `hankaku' Katakana
152 \(`japanese-jisx0201-kana'), in which case return value
153 may be a string even if OBJ is a character if two Katakanas are
154 necessary to represent OBJ."
155 (if (stringp obj)
156 (japanese-string-conversion obj 'japanese-katakana-region hankaku)
157 (or (get-char-code-property obj (if hankaku 'jisx0201 'katakana))
158 obj)))
160 ;;;###autoload
161 (defun japanese-hiragana (obj)
162 "Convert argument to Hiragana and return that.
163 The argument may be a character or string. The result has the same type.
164 The argument object is not altered--the value is a copy."
165 (if (stringp obj)
166 (japanese-string-conversion obj 'japanese-hiragana-region)
167 (or (get-char-code-property obj 'hiragana)
168 obj)))
170 ;;;###autoload
171 (defun japanese-hankaku (obj &optional ascii-only)
172 "Convert argument to `hankaku' and return that.
173 The argument may be a character or string. The result has the same type.
174 The argument object is not altered--the value is a copy.
175 Optional argument ASCII-ONLY non-nil means to return only ASCII character."
176 (if (stringp obj)
177 (japanese-string-conversion obj 'japanese-hankaku-region ascii-only)
178 (or (get-char-code-property obj 'ascii)
179 (and (not ascii-only)
180 (get-char-code-property obj 'jisx0201))
181 obj)))
183 ;;;###autoload
184 (defun japanese-zenkaku (obj)
185 "Convert argument to `zenkaku' and return that.
186 The argument may be a character or string. The result has the same type.
187 The argument object is not altered--the value is a copy."
188 (if (stringp obj)
189 (japanese-string-conversion obj 'japanese-zenkaku-region)
190 (or (get-char-code-property obj 'jisx0208)
191 obj)))
193 ;;;###autoload
194 (defun japanese-katakana-region (from to &optional hankaku)
195 "Convert Japanese `hiragana' chars in the region to `katakana' chars.
196 Optional argument HANKAKU t means to convert to `hankaku katakana' character
197 of which charset is `japanese-jisx0201-kana'."
198 (interactive "r\nP")
199 (save-restriction
200 (narrow-to-region from to)
201 (goto-char (point-min))
202 (while (re-search-forward "\\cH\\|\\cK" nil t)
203 (let* ((hira (preceding-char))
204 (kata (japanese-katakana hira hankaku)))
205 (if kata
206 (progn
207 (delete-region (match-beginning 0) (match-end 0))
208 (insert kata)))))))
210 ;;;###autoload
211 (defun japanese-hiragana-region (from to)
212 "Convert Japanese `katakana' chars in the region to `hiragana' chars."
213 (interactive "r")
214 (save-restriction
215 (narrow-to-region from to)
216 (goto-char (point-min))
217 (while (re-search-forward "\\cK\\|\\ck" nil t)
218 (let* ((kata (preceding-char))
219 (hira (japanese-hiragana kata)))
220 (if hira
221 (progn
222 (delete-region (match-beginning 0) (match-end 0))
223 (insert hira)))))))
225 ;;;###autoload
226 (defun japanese-hankaku-region (from to &optional ascii-only)
227 "Convert Japanese `zenkaku' chars in the region to `hankaku' chars.
228 `Zenkaku' chars belong to `japanese-jisx0208'
229 `Hankaku' chars belong to `ascii' or `japanese-jisx0201-kana'.
230 Optional argument ASCII-ONLY non-nil means to convert only to ASCII char."
231 (interactive "r\nP")
232 (save-restriction
233 (narrow-to-region from to)
234 (goto-char (point-min))
235 (while (re-search-forward "\\cj" nil t)
236 (let* ((zenkaku (preceding-char))
237 (hankaku (japanese-hankaku zenkaku ascii-only)))
238 (if hankaku
239 (progn
240 (delete-region (match-beginning 0) (match-end 0))
241 (insert hankaku)))))))
243 ;;;###autoload
244 (defun japanese-zenkaku-region (from to)
245 "Convert hankaku' chars in the region to Japanese `zenkaku' chars.
246 `Zenkaku' chars belong to `japanese-jisx0208'
247 `Hankaku' chars belong to `ascii' or `japanese-jisx0201-kana'."
248 (interactive "r")
249 (save-restriction
250 (narrow-to-region from to)
251 (goto-char (point-min))
252 (while (re-search-forward "\\ca\\|\\ck" nil t)
253 (let* ((hankaku (preceding-char))
254 (zenkaku (japanese-zenkaku hankaku)))
255 (if zenkaku
256 (progn
257 (delete-region (match-beginning 0) (match-end 0))
258 (insert zenkaku)))))))
260 ;;;###autoload
261 (defun read-hiragana-string (prompt &optional initial-input)
262 "Read a Hiragana string from the minibuffer, prompting with string PROMPT.
263 If non-nil, second arg INITIAL-INPUT is a string to insert before reading."
264 (read-multilingual-string prompt initial-input
265 "Japanese" "quail-ja-hiragana"))
268 (provide 'language/japan-util)
270 ;;; Local Variables:
271 ;;; generated-autoload-file: "../loaddefs.el"
272 ;;; End:
273 ;;; japan-util.el ends here