|
771
|
1 ;;; unicode.el --- Unicode support -*- coding: iso-2022-7bit; -*-
|
|
|
2
|
|
778
|
3 ;; Copyright (C) 2001, 2002 Ben Wing.
|
|
771
|
4
|
|
|
5 ;; Keywords: multilingual, Unicode
|
|
|
6
|
|
|
7 ;; This file is part of XEmacs.
|
|
|
8
|
|
|
9 ;; XEmacs is free software; you can redistribute it and/or modify it
|
|
|
10 ;; under the terms of the GNU General Public License as published by
|
|
|
11 ;; the Free Software Foundation; either version 2, or (at your option)
|
|
|
12 ;; any later version.
|
|
|
13
|
|
|
14 ;; XEmacs is distributed in the hope that it will be useful, but
|
|
|
15 ;; WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
16 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
17 ;; General Public License for more details.
|
|
|
18
|
|
|
19 ;; You should have received a copy of the GNU General Public License
|
|
|
20 ;; along with XEmacs; see the file COPYING. If not, write to the Free
|
|
|
21 ;; Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
|
|
|
22 ;; 02111-1307, USA.
|
|
|
23
|
|
|
24 ;;; Synched up with: Not in FSF.
|
|
|
25
|
|
|
26 ;;; Commentary:
|
|
|
27
|
|
|
28 ;; Lisp support for Unicode, e.g. initialize the translation tables.
|
|
|
29
|
|
|
30 ;;; Code:
|
|
|
31
|
|
3659
|
32 ;; GNU Emacs has the charsets:
|
|
778
|
33
|
|
3659
|
34 ;; mule-unicode-2500-33ff
|
|
|
35 ;; mule-unicode-e000-ffff
|
|
|
36 ;; mule-unicode-0100-24ff
|
|
778
|
37
|
|
3659
|
38 ;; built-in. This is hack--and an incomplete hack at that--against the
|
|
|
39 ;; spirit and the letter of standard ISO 2022 character sets. Instead of
|
|
|
40 ;; this, we have the jit-ucs-charset-N Mule character sets, created in
|
|
|
41 ;; unicode.c on encountering a Unicode code point that we don't recognise,
|
|
|
42 ;; and saved in ISO 2022 coding systems using the UTF-8 escape described in
|
|
|
43 ;; ISO-IR 196.
|
|
778
|
44
|
|
4083
|
45 (eval-when-compile (when (featurep 'mule) (require 'ccl)))
|
|
|
46
|
|
2367
|
47 ;; accessed in loadup.el, mule-cmds.el; see discussion in unicode.c
|
|
|
48 (defvar load-unicode-tables-at-dump-time (eq system-type 'windows-nt)
|
|
|
49 "[INTERNAL] Whether to load the Unicode tables at dump time.
|
|
|
50 Setting this at run-time does nothing.")
|
|
|
51
|
|
771
|
52 ;; NOTE: This takes only a fraction of a second on my Pentium III
|
|
|
53 ;; 700Mhz even with a totally optimization-disabled XEmacs.
|
|
|
54 (defun load-unicode-tables ()
|
|
|
55 "Initialize the Unicode translation tables for all standard charsets."
|
|
780
|
56 (let ((parse-args
|
|
|
57 '(("unicode/unicode-consortium"
|
|
877
|
58 ;; Due to the braindamaged way Mule treats the ASCII and Control-1
|
|
|
59 ;; charsets' types, trying to load them results in out-of-range
|
|
|
60 ;; warnings at unicode.c:1439. They're no-ops anyway, they're
|
|
|
61 ;; hardwired in unicode.c (unicode_to_ichar, ichar_to_unicode).
|
|
|
62 ;; ("8859-1.TXT" ascii #x00 #x7F #x0)
|
|
|
63 ;; ("8859-1.TXT" control-1 #x80 #x9F #x-80)
|
|
|
64 ;; The 8859-1.TXT G1 assignments are half no-ops, hardwired in
|
|
|
65 ;; unicode.c ichar_to_unicode, but not in unicode_to_ichar.
|
|
780
|
66 ("8859-1.TXT" latin-iso8859-1 #xA0 #xFF #x-80)
|
|
|
67 ;; "8859-10.TXT"
|
|
|
68 ;; "8859-13.TXT"
|
|
|
69 ("8859-14.TXT" latin-iso8859-14 #xA0 #xFF #x-80)
|
|
|
70 ("8859-15.TXT" latin-iso8859-15 #xA0 #xFF #x-80)
|
|
2575
|
71 ("8859-16.TXT" latin-iso8859-16 #xA0 #xFF #x-80)
|
|
780
|
72 ("8859-2.TXT" latin-iso8859-2 #xA0 #xFF #x-80)
|
|
|
73 ("8859-3.TXT" latin-iso8859-3 #xA0 #xFF #x-80)
|
|
|
74 ("8859-4.TXT" latin-iso8859-4 #xA0 #xFF #x-80)
|
|
|
75 ("8859-5.TXT" cyrillic-iso8859-5 #xA0 #xFF #x-80)
|
|
|
76 ("8859-6.TXT" arabic-iso8859-6 #xA0 #xFF #x-80)
|
|
|
77 ("8859-7.TXT" greek-iso8859-7 #xA0 #xFF #x-80)
|
|
|
78 ("8859-8.TXT" hebrew-iso8859-8 #xA0 #xFF #x-80)
|
|
|
79 ("8859-9.TXT" latin-iso8859-9 #xA0 #xFF #x-80)
|
|
|
80 ;; charset for Big5 does not matter; specifying `big5' will
|
|
|
81 ;; automatically make the right thing happen
|
|
|
82 ("BIG5.TXT" chinese-big5-1 nil nil nil big5)
|
|
|
83 ("CNS11643.TXT" chinese-cns11643-1 #x10000 #x1FFFF #x-10000)
|
|
|
84 ("CNS11643.TXT" chinese-cns11643-2 #x20000 #x2FFFF #x-20000)
|
|
|
85 ;; "CP1250.TXT"
|
|
|
86 ;; "CP1251.TXT"
|
|
|
87 ;; "CP1252.TXT"
|
|
|
88 ;; "CP1253.TXT"
|
|
|
89 ;; "CP1254.TXT"
|
|
|
90 ;; "CP1255.TXT"
|
|
|
91 ;; "CP1256.TXT"
|
|
|
92 ;; "CP1257.TXT"
|
|
|
93 ;; "CP1258.TXT"
|
|
|
94 ;; "CP874.TXT"
|
|
|
95 ;; "CP932.TXT"
|
|
|
96 ;; "CP936.TXT"
|
|
|
97 ;; "CP949.TXT"
|
|
|
98 ;; "CP950.TXT"
|
|
|
99 ;; "GB12345.TXT"
|
|
|
100 ("GB2312.TXT" chinese-gb2312)
|
|
2297
|
101 ;; "HANGUL.TXT"
|
|
|
102 ;; #### shouldn't JIS X 0201's upper limit be 7f?
|
|
780
|
103 ("JIS0201.TXT" latin-jisx0201 #x21 #x80)
|
|
|
104 ("JIS0201.TXT" katakana-jisx0201 #xA0 #xFF #x-80)
|
|
|
105 ("JIS0208.TXT" japanese-jisx0208 nil nil nil ignore-first-column)
|
|
|
106 ("JIS0212.TXT" japanese-jisx0212)
|
|
|
107 ;; "JOHAB.TXT"
|
|
|
108 ;; "KOI8-R.TXT"
|
|
|
109 ;; "KSC5601.TXT"
|
|
|
110 ;; note that KSC5601.TXT as currently distributed is NOT what
|
|
|
111 ;; it claims to be! see comments in KSX1001.TXT.
|
|
|
112 ("KSX1001.TXT" korean-ksc5601)
|
|
|
113 ;; "OLD5601.TXT"
|
|
|
114 ;; "SHIFTJIS.TXT"
|
|
|
115 )
|
|
|
116 ("unicode/mule-ucs"
|
|
2297
|
117 ;; #### we don't support surrogates?!??
|
|
780
|
118 ;; use these instead of the above ones once we support surrogates
|
|
|
119 ;;("chinese-cns11643-1.txt" chinese-cns11643-1)
|
|
|
120 ;;("chinese-cns11643-2.txt" chinese-cns11643-2)
|
|
|
121 ;;("chinese-cns11643-3.txt" chinese-cns11643-3)
|
|
|
122 ;;("chinese-cns11643-4.txt" chinese-cns11643-4)
|
|
|
123 ;;("chinese-cns11643-5.txt" chinese-cns11643-5)
|
|
|
124 ;;("chinese-cns11643-6.txt" chinese-cns11643-6)
|
|
|
125 ;;("chinese-cns11643-7.txt" chinese-cns11643-7)
|
|
|
126 ("chinese-sisheng.txt" chinese-sisheng)
|
|
|
127 ("ethiopic.txt" ethiopic)
|
|
|
128 ("indian-is13194.txt" indian-is13194)
|
|
|
129 ("ipa.txt" ipa)
|
|
|
130 ("thai-tis620.txt" thai-tis620)
|
|
|
131 ("tibetan.txt" tibetan)
|
|
|
132 ("vietnamese-viscii-lower.txt" vietnamese-viscii-lower)
|
|
|
133 ("vietnamese-viscii-upper.txt" vietnamese-viscii-upper)
|
|
|
134 )
|
|
|
135 ("unicode/other"
|
|
|
136 ("lao.txt" lao)
|
|
|
137 )
|
|
771
|
138 )))
|
|
780
|
139 (mapcar #'(lambda (tables)
|
|
|
140 (let ((undir
|
|
|
141 (expand-file-name (car tables) data-directory)))
|
|
|
142 (mapcar #'(lambda (args)
|
|
1318
|
143 (apply 'load-unicode-mapping-table
|
|
780
|
144 (expand-file-name (car args) undir)
|
|
|
145 (cdr args)))
|
|
|
146 (cdr tables))))
|
|
4145
|
147 parse-args)
|
|
|
148 ;; The default-unicode-precedence-list. We set this here to default to
|
|
|
149 ;; *not* mapping various European characters to East Asian characters;
|
|
|
150 ;; otherwise the default-unicode-precedence-list is numerically ordered
|
|
|
151 ;; by charset ID.
|
|
|
152 (set-default-unicode-precedence-list
|
|
|
153 '(ascii control-1 latin-iso8859-1 latin-iso8859-2 latin-iso8859-15
|
|
|
154 greek-iso8859-7 hebrew-iso8859-8 ipa cyrillic-iso8859-5
|
|
|
155 latin-iso8859-16 latin-iso8859-3 latin-iso8859-4 latin-iso8859-9
|
|
|
156 vietnamese-viscii-lower vietnamese-viscii-upper arabic-iso8859-6
|
|
|
157 jit-ucs-charset-0 japanese-jisx0208 japanese-jisx0208-1978
|
|
|
158 japanese-jisx0212 japanese-jisx0213-1 japanese-jisx0213-2
|
|
|
159 chinese-gb2312 chinese-sisheng chinese-big5-1 chinese-big5-2
|
|
|
160 indian-is13194 korean-ksc5601 chinese-cns11643-1 chinese-cns11643-2
|
|
|
161 chinese-isoir165 arabic-1-column arabic-2-column arabic-digit
|
|
|
162 composite ethiopic indian-1-column indian-2-column jit-ucs-charset-0
|
|
|
163 katakana-jisx0201 lao thai-tis620 thai-xtis tibetan tibetan-1-column
|
|
|
164 latin-jisx0201 chinese-cns11643-3 chinese-cns11643-4
|
|
|
165 chinese-cns11643-5 chinese-cns11643-6 chinese-cns11643-7))))
|
|
771
|
166
|
|
|
167 (make-coding-system
|
|
|
168 'utf-16 'unicode
|
|
|
169 "UTF-16"
|
|
|
170 '(mnemonic "UTF-16"
|
|
3767
|
171 documentation
|
|
771
|
172 "UTF-16 Unicode encoding -- the standard (almost-) fixed-width
|
|
|
173 two-byte encoding, with surrogates. It will be fixed-width if all
|
|
|
174 characters are in the BMP (Basic Multilingual Plane -- first 65536
|
|
|
175 codepoints). Cannot represent characters with codepoints above
|
|
|
176 0x10FFFF (a little more than 1,000,000). Unicode and ISO guarantee
|
|
|
177 never to encode any characters outside this range -- all the rest are
|
|
|
178 for private, corporate or internal use."
|
|
3767
|
179 unicode-type utf-16))
|
|
771
|
180
|
|
2574
|
181 (define-coding-system-alias 'utf-16-be 'utf-16)
|
|
|
182
|
|
771
|
183 (make-coding-system
|
|
|
184 'utf-16-bom 'unicode
|
|
|
185 "UTF-16 w/BOM"
|
|
|
186 '(mnemonic "UTF16-BOM"
|
|
3767
|
187 documentation
|
|
771
|
188 "UTF-16 Unicode encoding with byte order mark (BOM) at the beginning.
|
|
|
189 The BOM is Unicode character U+FEFF -- i.e. the first two bytes are
|
|
|
190 0xFE and 0xFF, respectively, or reversed in a little-endian
|
|
|
191 representation. It has been sanctioned by the Unicode Consortium for
|
|
|
192 use at the beginning of a Unicode stream as a marker of the byte order
|
|
|
193 of the stream, and commonly appears in Unicode files under Microsoft
|
|
|
194 Windows, where it also functions as a magic cookie identifying a
|
|
|
195 Unicode file. The character is called \"ZERO WIDTH NO-BREAK SPACE\"
|
|
|
196 and is suitable as a byte-order marker because:
|
|
|
197
|
|
|
198 -- it has no displayable representation
|
|
|
199 -- due to its semantics it never normally appears at the beginning
|
|
|
200 of a stream
|
|
|
201 -- its reverse U+FFFE is not a legal Unicode character
|
|
|
202 -- neither byte sequence is at all likely in any other standard
|
|
|
203 encoding, particularly at the beginning of a stream
|
|
|
204
|
|
|
205 This coding system will insert a BOM at the beginning of a stream when
|
|
|
206 writing and strip it off when reading."
|
|
3767
|
207 unicode-type utf-16
|
|
771
|
208 need-bom t))
|
|
|
209
|
|
|
210 (make-coding-system
|
|
|
211 'utf-16-little-endian 'unicode
|
|
|
212 "UTF-16 Little Endian"
|
|
|
213 '(mnemonic "UTF16-LE"
|
|
|
214 documentation
|
|
|
215 "Little-endian version of UTF-16 Unicode encoding.
|
|
|
216 See `utf-16' coding system."
|
|
3767
|
217 unicode-type utf-16
|
|
771
|
218 little-endian t))
|
|
|
219
|
|
2574
|
220 (define-coding-system-alias 'utf-16-le 'utf-16-little-endian)
|
|
|
221
|
|
771
|
222 (make-coding-system
|
|
|
223 'utf-16-little-endian-bom 'unicode
|
|
|
224 "UTF-16 Little Endian w/BOM"
|
|
|
225 '(mnemonic "MSW-Unicode"
|
|
|
226 documentation
|
|
|
227 "Little-endian version of UTF-16 Unicode encoding, with byte order mark.
|
|
|
228 Standard encoding for representing Unicode under MS Windows. See
|
|
|
229 `utf-16-bom' coding system."
|
|
3767
|
230 unicode-type utf-16
|
|
771
|
231 little-endian t
|
|
|
232 need-bom t))
|
|
|
233
|
|
|
234 (make-coding-system
|
|
|
235 'ucs-4 'unicode
|
|
|
236 "UCS-4"
|
|
|
237 '(mnemonic "UCS4"
|
|
|
238 documentation
|
|
|
239 "UCS-4 Unicode encoding -- fully fixed-width four-byte encoding."
|
|
3767
|
240 unicode-type ucs-4))
|
|
771
|
241
|
|
|
242 (make-coding-system
|
|
|
243 'ucs-4-little-endian 'unicode
|
|
|
244 "UCS-4 Little Endian"
|
|
|
245 '(mnemonic "UCS4-LE"
|
|
|
246 documentation
|
|
2297
|
247 ;; #### I don't think this is permitted by ISO 10646, only Unicode.
|
|
|
248 ;; Call it UTF-32 instead?
|
|
771
|
249 "Little-endian version of UCS-4 Unicode encoding. See `ucs-4' coding system."
|
|
3767
|
250 unicode-type ucs-4
|
|
771
|
251 little-endian t))
|
|
|
252
|
|
|
253 (make-coding-system
|
|
4096
|
254 'utf-32 'unicode
|
|
|
255 "UTF-32"
|
|
|
256 '(mnemonic "UTF32"
|
|
|
257 documentation
|
|
|
258 "UTF-32 Unicode encoding -- fixed-width four-byte encoding,
|
|
|
259 characters less than #x10FFFF are not supported. "
|
|
|
260 unicode-type utf-32))
|
|
|
261
|
|
|
262 (make-coding-system
|
|
|
263 'utf-32-little-endian 'unicode
|
|
|
264 "UTF-32 Little Endian"
|
|
|
265 '(mnemonic "UTF32-LE"
|
|
|
266 documentation
|
|
|
267 "Little-endian version of UTF-32 Unicode encoding.
|
|
|
268
|
|
|
269 A fixed-width four-byte encoding, characters less than #x10FFFF are not
|
|
|
270 supported. "
|
|
|
271 unicode-type ucs-4 little-endian t))
|
|
|
272
|
|
|
273 (make-coding-system
|
|
771
|
274 'utf-8 'unicode
|
|
|
275 "UTF-8"
|
|
|
276 '(mnemonic "UTF8"
|
|
3767
|
277 documentation "
|
|
|
278 UTF-8 Unicode encoding -- ASCII-compatible 8-bit variable-width encoding
|
|
2297
|
279 sharing the following principles with the Mule-internal encoding:
|
|
771
|
280
|
|
|
281 -- All ASCII characters (codepoints 0 through 127) are represented
|
|
|
282 by themselves (i.e. using one byte, with the same value as the
|
|
|
283 ASCII codepoint), and these bytes are disjoint from bytes
|
|
|
284 representing non-ASCII characters.
|
|
|
285
|
|
|
286 This means that any 8-bit clean application can safely process
|
|
|
287 UTF-8-encoded text as it were ASCII, with no corruption (e.g. a
|
|
|
288 '/' byte is always a slash character, never the second byte of
|
|
|
289 some other character, as with Big5, so a pathname encoded in
|
|
|
290 UTF-8 can safely be split up into components and reassembled
|
|
|
291 again using standard ASCII processes).
|
|
|
292
|
|
|
293 -- Leading bytes and non-leading bytes in the encoding of a
|
|
|
294 character are disjoint, so moving backwards is easy.
|
|
|
295
|
|
|
296 -- Given only the leading byte, you know how many following bytes
|
|
|
297 are present.
|
|
|
298 "
|
|
3767
|
299 unicode-type utf-8))
|
|
771
|
300
|
|
985
|
301 (make-coding-system
|
|
|
302 'utf-8-bom 'unicode
|
|
|
303 "UTF-8 w/BOM"
|
|
|
304 '(mnemonic "MSW-UTF8"
|
|
|
305 documentation
|
|
|
306 "UTF-8 Unicode encoding, with byte order mark.
|
|
|
307 Standard encoding for representing UTF-8 under MS Windows."
|
|
3767
|
308 unicode-type utf-8
|
|
985
|
309 little-endian t
|
|
|
310 need-bom t))
|
|
|
311
|
|
2633
|
312 (defun decode-char (quote-ucs code &optional restriction)
|
|
3659
|
313 "FSF compatibility--return Mule character with Unicode codepoint CODE.
|
|
2633
|
314 The second argument must be 'ucs, the third argument is ignored. "
|
|
4096
|
315 ;; We're prepared to accept invalid Unicode in unicode-to-char, but not in
|
|
|
316 ;; this function, which is the API that should actually be used, since
|
|
|
317 ;; it's available in GNU and in Mule-UCS.
|
|
|
318 (check-argument-range code #x0 #x10FFFF)
|
|
3506
|
319 (assert (eq quote-ucs 'ucs) t
|
|
2633
|
320 "Sorry, decode-char doesn't yet support anything but the UCS. ")
|
|
|
321 (unicode-to-char code))
|
|
|
322
|
|
|
323 (defun encode-char (char quote-ucs &optional restriction)
|
|
3659
|
324 "FSF compatibility--return the Unicode code point of CHAR.
|
|
2633
|
325 The second argument must be 'ucs, the third argument is ignored. "
|
|
3506
|
326 (assert (eq quote-ucs 'ucs) t
|
|
2633
|
327 "Sorry, encode-char doesn't yet support anything but the UCS. ")
|
|
|
328 (char-to-unicode char))
|
|
|
329
|
|
3667
|
330 (when (featurep 'mule)
|
|
4083
|
331 (let ((prog
|
|
|
332 (eval-when-compile
|
|
|
333 (let ((pre-existing
|
|
|
334 ;; This is the compiled CCL program from the assert
|
|
|
335 ;; below. Since this file is dumped and ccl.el isn't (and
|
|
|
336 ;; even when it was, it was dumped much later than this
|
|
|
337 ;; one), we can't compile the program at dump time. We can
|
|
|
338 ;; check at byte compile time that the program is as
|
|
|
339 ;; expected, though.
|
|
4145
|
340 [1 16 131127 7 98872 65823 1307 5 -65536 65313 64833 1028
|
|
|
341 147513 8 82009 255 22]))
|
|
4083
|
342 (when (featurep 'mule)
|
|
|
343 ;; Check that the pre-existing constant reflects the intended
|
|
|
344 ;; CCL program.
|
|
|
345 (assert
|
|
|
346 (equal pre-existing
|
|
|
347 (ccl-compile
|
|
|
348 `(1
|
|
|
349 (;; mule-to-unicode's first argument is the
|
|
|
350 ;; charset ID, the second its first byte
|
|
|
351 ;; left shifted by 7 bits masked with its
|
|
|
352 ;; second byte.
|
|
|
353 (r1 = (r1 << 7))
|
|
|
354 (r1 = (r1 | r2))
|
|
|
355 (mule-to-unicode r0 r1)
|
|
|
356 (if (r0 & ,(lognot #xFFFF))
|
|
|
357 ;; Redisplay looks in r1 and r2 for the first
|
|
|
358 ;; and second bytes of the X11 font,
|
|
|
359 ;; respectively. For non-BMP characters we
|
|
|
360 ;; display U+FFFD.
|
|
|
361 ((r1 = #xFF)
|
|
|
362 (r2 = #xFD))
|
|
|
363 ((r1 = (r0 >> 8))
|
|
|
364 (r2 = (r0 & #xFF))))))))
|
|
|
365 nil
|
|
|
366 "The pre-compiled CCL program appears broken. "))
|
|
|
367 pre-existing))))
|
|
|
368 (defconst ccl-encode-to-ucs-2 prog
|
|
3667
|
369 "CCL program to transform Mule characters to UCS-2.")
|
|
4083
|
370 (put 'ccl-encode-to-ucs-2 'ccl-program-idx
|
|
4145
|
371 (register-ccl-program 'ccl-encode-to-ucs-2 prog)))
|
|
|
372
|
|
|
373 ;; Now, create jit-ucs-charset-0 entries for those characters in Windows
|
|
|
374 ;; Glyph List 4 that would otherwise end up in East Asian character sets.
|
|
|
375 ;;
|
|
|
376 ;; WGL4 is a character repertoire from Microsoft that gives a guideline
|
|
|
377 ;; for font implementors as to what characters are sufficient for
|
|
|
378 ;; pan-European support. The intention of this code is to avoid the
|
|
|
379 ;; situation where these characters end up mapping to East Asian XEmacs
|
|
|
380 ;; characters, which generally clash strongly with European characters
|
|
|
381 ;; both in font choice and character width; jit-ucs-charset-0 is a
|
|
|
382 ;; single-width character set which comes before the East Asian character
|
|
|
383 ;; sets in the default-unicode-precedence-list above.
|
|
|
384 (loop for (ucs ascii-or-latin-1)
|
|
|
385 in '((#x2013 ?-) ;; U+2013 EN DASH
|
|
|
386 (#x2014 ?-) ;; U+2014 EM DASH
|
|
|
387 (#x2105 ?%) ;; U+2105 CARE OF
|
|
|
388 (#x203e ?-) ;; U+203E OVERLINE
|
|
|
389 (#x221f ?|) ;; U+221F RIGHT ANGLE
|
|
|
390 (#x2584 ?|) ;; U+2584 LOWER HALF BLOCK
|
|
|
391 (#x2588 ?|) ;; U+2588 FULL BLOCK
|
|
|
392 (#x258c ?|) ;; U+258C LEFT HALF BLOCK
|
|
|
393 (#x2550 ?|) ;; U+2550 BOX DRAWINGS DOUBLE HORIZONTAL
|
|
|
394 (#x255e ?|) ;; U+255E BOX DRAWINGS VERTICAL SINGLE AND RIGHT DOUBLE
|
|
|
395 (#x256a ?|) ;; U+256A BOX DRAWINGS VERTICAL SINGLE & HORIZONTAL DOUBLE
|
|
|
396 (#x2561 ?|) ;; U+2561 BOX DRAWINGS VERTICAL SINGLE AND LEFT DOUBLE
|
|
|
397 (#x2215 ?/) ;; U+2215 DIVISION SLASH
|
|
|
398 (#x02c9 ?`) ;; U+02C9 MODIFIER LETTER MACRON
|
|
|
399 (#x2211 ?s) ;; U+2211 N-ARY SUMMATION
|
|
|
400 (#x220f ?s) ;; U+220F N-ARY PRODUCT
|
|
|
401 (#x2248 ?=) ;; U+2248 ALMOST EQUAL TO
|
|
|
402 (#x2264 ?=) ;; U+2264 LESS-THAN OR EQUAL TO
|
|
|
403 (#x2265 ?=) ;; U+2265 GREATER-THAN OR EQUAL TO
|
|
|
404 (#x201c ?') ;; U+201C LEFT DOUBLE QUOTATION MARK
|
|
|
405 (#x2026 ?.) ;; U+2026 HORIZONTAL ELLIPSIS
|
|
|
406 (#x2212 ?-) ;; U+2212 MINUS SIGN
|
|
|
407 (#x2260 ?=) ;; U+2260 NOT EQUAL TO
|
|
|
408 (#x221e ?=) ;; U+221E INFINITY
|
|
|
409 (#x2642 ?=) ;; U+2642 MALE SIGN
|
|
|
410 (#x2640 ?=) ;; U+2640 FEMALE SIGN
|
|
|
411 (#x2032 ?=) ;; U+2032 PRIME
|
|
|
412 (#x2033 ?=) ;; U+2033 DOUBLE PRIME
|
|
|
413 (#x25cb ?=) ;; U+25CB WHITE CIRCLE
|
|
|
414 (#x25cf ?=) ;; U+25CF BLACK CIRCLE
|
|
|
415 (#x25a1 ?=) ;; U+25A1 WHITE SQUARE
|
|
|
416 (#x25a0 ?=) ;; U+25A0 BLACK SQUARE
|
|
|
417 (#x25b2 ?=) ;; U+25B2 BLACK UP-POINTING TRIANGLE
|
|
|
418 (#x25bc ?=) ;; U+25BC BLACK DOWN-POINTING TRIANGLE
|
|
|
419 (#x2192 ?=) ;; U+2192 RIGHTWARDS ARROW
|
|
|
420 (#x2190 ?=) ;; U+2190 LEFTWARDS ARROW
|
|
|
421 (#x2191 ?=) ;; U+2191 UPWARDS ARROW
|
|
|
422 (#x2193 ?=) ;; U+2193 DOWNWARDS ARROW
|
|
|
423 (#x2229 ?=) ;; U+2229 INTERSECTION
|
|
|
424 (#x2202 ?=) ;; U+2202 PARTIAL DIFFERENTIAL
|
|
|
425 (#x2261 ?=) ;; U+2261 IDENTICAL TO
|
|
|
426 (#x221a ?=) ;; U+221A SQUARE ROOT
|
|
|
427 (#x222b ?=) ;; U+222B INTEGRAL
|
|
|
428 (#x2030 ?=) ;; U+2030 PER MILLE SIGN
|
|
|
429 (#x266a ?=) ;; U+266A EIGHTH NOTE
|
|
|
430 (#x2020 ?*) ;; U+2020 DAGGER
|
|
|
431 (#x2021 ?*) ;; U+2021 DOUBLE DAGGER
|
|
|
432 (#x2500 ?|) ;; U+2500 BOX DRAWINGS LIGHT HORIZONTAL
|
|
|
433 (#x2502 ?|) ;; U+2502 BOX DRAWINGS LIGHT VERTICAL
|
|
|
434 (#x250c ?|) ;; U+250C BOX DRAWINGS LIGHT DOWN AND RIGHT
|
|
|
435 (#x2510 ?|) ;; U+2510 BOX DRAWINGS LIGHT DOWN AND LEFT
|
|
|
436 (#x2518 ?|) ;; U+2518 BOX DRAWINGS LIGHT UP AND LEFT
|
|
|
437 (#x2514 ?|) ;; U+2514 BOX DRAWINGS LIGHT UP AND RIGHT
|
|
|
438 (#x251c ?|) ;; U+251C BOX DRAWINGS LIGHT VERTICAL AND RIGHT
|
|
|
439 (#x252c ?|) ;; U+252C BOX DRAWINGS LIGHT DOWN AND HORIZONTAL
|
|
|
440 (#x2524 ?|) ;; U+2524 BOX DRAWINGS LIGHT VERTICAL AND LEFT
|
|
|
441 (#x2534 ?|) ;; U+2534 BOX DRAWINGS LIGHT UP AND HORIZONTAL
|
|
|
442 (#x253c ?|) ;; U+253C BOX DRAWINGS LIGHT VERTICAL AND HORIZONTAL
|
|
|
443 (#x02da ?^) ;; U+02DA RING ABOVE
|
|
|
444 (#x2122 ?\xa9) ;; U+2122 TRADE MARK SIGN, ?,A)(B
|
|
|
445
|
|
|
446 (#x0132 ?\xe6) ;; U+0132 LATIN CAPITAL LIGATURE IJ, ?,Af(B
|
|
|
447 (#x013f ?\xe6) ;; U+013F LATIN CAPITAL LETTER L WITH MIDDLE DOT, ?,Af(B
|
|
|
448
|
|
|
449 (#x0133 ?\xe6) ;; U+0133 LATIN SMALL LIGATURE IJ, ?,Af(B
|
|
|
450 (#x0140 ?\xe6) ;; U+0140 LATIN SMALL LETTER L WITH MIDDLE DOT, ?,Af(B
|
|
|
451 (#x0149 ?\xe6) ;; U+0149 LATIN SMALL LETTER N PRECEDED BY APOSTROPH,?,Af(B
|
|
|
452
|
|
|
453 (#x2194 ?|) ;; U+2194 LEFT RIGHT ARROW
|
|
|
454 (#x2660 ?*) ;; U+2660 BLACK SPADE SUIT
|
|
|
455 (#x2665 ?*) ;; U+2665 BLACK HEART SUIT
|
|
|
456 (#x2663 ?*) ;; U+2663 BLACK CLUB SUIT
|
|
|
457 (#x2592 ?|) ;; U+2592 MEDIUM SHADE
|
|
|
458 (#x2195 ?|) ;; U+2195 UP DOWN ARROW
|
|
|
459
|
|
|
460 (#x2113 ?\xb9) ;; U+2113 SCRIPT SMALL L, ?,A9(B
|
|
|
461 (#x215b ?\xbe) ;; U+215B VULGAR FRACTION ONE EIGHTH, ?,A>(B
|
|
|
462 (#x215c ?\xbe) ;; U+215C VULGAR FRACTION THREE EIGHTHS, ?,A>(B
|
|
|
463 (#x215d ?\xbe) ;; U+215D VULGAR FRACTION FIVE EIGHTHS, ?,A>(B
|
|
|
464 (#x215e ?\xbe) ;; U+215E VULGAR FRACTION SEVEN EIGHTHS, ?,A>(B
|
|
|
465 (#x207f ?\xbe) ;; U+207F SUPERSCRIPT LATIN SMALL LETTER N, ?,A>(B
|
|
|
466
|
|
|
467 ;; These are not in WGL 4, but are IPA characters that should not
|
|
|
468 ;; be double width. They are the only IPA characters that both
|
|
|
469 ;; occur in packages/mule-packages/leim/ipa.el and end up in East
|
|
|
470 ;; Asian character sets when that file is loaded in an XEmacs
|
|
|
471 ;; without packages.
|
|
|
472 (#x2197 ?|) ;; U+2197 NORTH EAST ARROW
|
|
|
473 (#x2199 ?|) ;; U+2199 SOUTH WEST ARROW
|
|
|
474 (#x2191 ?|) ;; U+2191 UPWARDS ARROW
|
|
|
475 (#x207f ?\xb9));; U+207F SUPERSCRIPT LATIN SMALL LETTER N, ?,A9(B
|
|
|
476 with decoded = nil
|
|
|
477 with syntax-table = (standard-syntax-table)
|
|
|
478 ;; This creates jit-ucs-charset-0 entries because:
|
|
|
479 ;;
|
|
|
480 ;; 1. If the tables are dumped, it is run at dump time before they are
|
|
|
481 ;; dumped, and as such before the relevant conversions are available
|
|
|
482 ;; (they are made available in mule/general-late.el).
|
|
|
483 ;;
|
|
|
484 ;; 2. If the tables are not dumped, it is run at dump time, long before
|
|
|
485 ;; any of the other mappings are available.
|
|
|
486 ;;
|
|
|
487 do
|
|
|
488 (setq decoded (decode-char 'ucs ucs))
|
|
|
489 (assert (eq (char-charset decoded)
|
|
|
490 'jit-ucs-charset-0) nil
|
|
|
491 "Unexpected Unicode decoding behavior. ")
|
|
|
492 (modify-syntax-entry decoded
|
|
|
493 (string
|
|
|
494 (char-syntax ascii-or-latin-1))
|
|
|
495 syntax-table))
|
|
|
496
|
|
|
497 ;; Create all the Unicode error sequences, normally as jit-ucs-charset-0
|
|
|
498 ;; characters starting at U+200000 (which isn't a valid Unicode code
|
|
4202
|
499 ;; point). Make them available to user code.
|
|
|
500 (defvar unicode-error-default-translation-table
|
|
|
501 (loop
|
|
|
502 with char-table = (make-char-table 'char)
|
|
|
503 for i from ?\x00 to ?\xFF
|
|
|
504 do
|
|
|
505 (put-char-table (aref
|
|
|
506 ;; #xd800 is the first leading surrogate;
|
|
|
507 ;; trailing surrogates must be in the range
|
|
|
508 ;; #xdc00-#xdfff. These examples are not, so we
|
|
|
509 ;; intentionally provoke an error sequence.
|
|
|
510 (decode-coding-string (format "\xd8\x00\x00%c" i)
|
|
|
511 'utf-16-be)
|
|
|
512 3)
|
|
|
513 i
|
|
|
514 char-table)
|
|
|
515 finally return char-table)
|
|
|
516 "Translation table mapping Unicode error sequences to Latin-1 chars.
|
|
4145
|
517
|
|
4202
|
518 To transform XEmacs Unicode error sequences to the Latin-1 characters that
|
|
|
519 correspond to the octets on disk, you can use this variable. ")
|
|
4145
|
520
|
|
|
521 (defvar unicode-error-sequence-regexp-range
|
|
4202
|
522 (format "%c%c-%c"
|
|
|
523 (aref (decode-coding-string "\xd8\x00\x00\x00" 'utf-16-be) 0)
|
|
|
524 (aref (decode-coding-string "\xd8\x00\x00\x00" 'utf-16-be) 3)
|
|
|
525 (aref (decode-coding-string "\xd8\x00\x00\xFF" 'utf-16-be) 3))
|
|
4145
|
526 "Regular expression range to match Unicode error sequences in XEmacs.
|
|
|
527
|
|
4202
|
528 Invalid Unicode sequences on input are represented as XEmacs
|
|
|
529 characters with values stored as the keys in
|
|
|
530 `unicode-error-default-translation-table', one character for each
|
|
|
531 invalid octet. You can use this variable (with `re-search-forward' or
|
|
|
532 `skip-chars-forward') to search for such characters; see also
|
|
|
533 `unicode-error-translate-region'. ")
|
|
|
534
|
|
|
535 ;; Check that the lookup table is correct, and that all the actual error
|
|
|
536 ;; sequences are caught by the regexp.
|
|
|
537 (with-temp-buffer
|
|
|
538 (loop
|
|
|
539 for i from ?\x00 to ?\xFF
|
|
|
540 with to-check = (make-string 20 ?\x20)
|
|
|
541 with res = t
|
|
|
542 do
|
|
|
543 (delete-region (point-min) (point-max))
|
|
|
544 (insert to-check)
|
|
|
545 (goto-char 10)
|
|
|
546 (insert (decode-coding-string (format "\xd8\x00\x00%c" i)
|
|
|
547 'utf-16-be))
|
|
|
548 (backward-char)
|
|
|
549 (assert (= i (get-char-table (char-after (point))
|
|
|
550 unicode-error-default-translation-table))
|
|
|
551 (format "Char ?\\x%x not the expected error sequence!"
|
|
|
552 i))
|
|
|
553
|
|
|
554 (goto-char (point-min))
|
|
|
555 (assert (re-search-forward (concat "["
|
|
|
556 unicode-error-sequence-regexp-range
|
|
|
557 "]"))
|
|
|
558 nil
|
|
|
559 (format "Could not find char ?\\x%x in buffer" i))))
|
|
|
560
|
|
|
561 (defun frob-unicode-errors-region (frob-function begin end &optional buffer)
|
|
|
562 "Call FROB-FUNCTION on the Unicode error sequences between BEGIN and END.
|
|
|
563
|
|
|
564 Optional argument BUFFER specifies the buffer that should be examined for
|
|
|
565 such sequences. "
|
|
|
566 (check-argument-type #'functionp frob-function)
|
|
|
567 (check-argument-range begin (point-min buffer) (point-max buffer))
|
|
|
568 (check-argument-range end (point-min buffer) (point-max buffer))
|
|
|
569 (save-excursion
|
|
|
570 (save-restriction
|
|
|
571 (if buffer (set-buffer buffer))
|
|
|
572 (narrow-to-region begin end)
|
|
|
573 (goto-char (point-min))
|
|
|
574 (while end
|
|
|
575 (setq begin
|
|
|
576 (progn
|
|
|
577 (skip-chars-forward
|
|
|
578 (concat "^" unicode-error-sequence-regexp-range))
|
|
|
579 (point))
|
|
|
580 end (and (not (= (point) (point-max)))
|
|
|
581 (progn
|
|
|
582 (skip-chars-forward
|
|
|
583 unicode-error-sequence-regexp-range)
|
|
|
584 (point))))
|
|
|
585 (if end
|
|
|
586 (funcall frob-function begin end))))))
|
|
|
587
|
|
|
588 (defun unicode-error-translate-region (begin end &optional buffer table)
|
|
|
589 "Translate the Unicode error sequences in BUFFER between BEGIN and END.
|
|
|
590
|
|
|
591 The error sequences are transformed, by default, into the ASCII,
|
|
|
592 control-1 and latin-iso8859-1 characters with the numeric values
|
|
|
593 corresponding to the incorrect octets encountered. This is achieved
|
|
|
594 by using `unicode-error-default-translation-table' (which see) for
|
|
|
595 TABLE; you can change this by supplying another character table,
|
|
|
596 mapping from the error sequences to the desired characters. "
|
|
|
597 (unless table (setq table unicode-error-default-translation-table))
|
|
|
598 (frob-unicode-errors-region
|
|
|
599 (lambda (start finish)
|
|
|
600 (translate-region start finish table))
|
|
|
601 begin end buffer)))
|
|
3667
|
602
|
|
771
|
603 ;; #### UTF-7 is not yet implemented, and it's tricky to do. There's
|
|
|
604 ;; an implementation in appendix A.1 of the Unicode Standard, Version
|
|
|
605 ;; 2.0, but I don't know its licensing characteristics.
|
|
|
606
|
|
|
607 ; (make-coding-system
|
|
|
608 ; 'utf-7 'unicode
|
|
|
609 ; "UTF-7"
|
|
|
610 ; '(mnemonic "UTF7"
|
|
3659
|
611 ; documentation; "UTF-7 Unicode encoding -- 7-bit-ASCII modal Internet-mail-compatible
|
|
771
|
612 ; encoding especially designed for headers, with the following
|
|
|
613 ; properties:
|
|
|
614
|
|
|
615 ; -- Only characters that are considered safe for passing through any mail
|
|
|
616 ; gateway without damage are used.
|
|
|
617
|
|
|
618 ; -- This is a modal encoding, with two states. The first, default
|
|
|
619 ; state encodes the most common Unicode characters (upper and
|
|
|
620 ; lowercase letters, digits, and 9 common punctuation marks) as
|
|
|
621 ; themselves, and the second state, entered using '+' and
|
|
|
622 ; terminated with '-' or any character disallowed in state 2,
|
|
|
623 ; encodes any Unicode characters by first converting to UTF-16,
|
|
|
624 ; most significant byte first, and then to a slightly modified
|
|
|
625 ; Base64 encoding. (Thus, UTF-7 has the same limitations on the
|
|
|
626 ; characters it can encode as UTF-16.)
|
|
|
627
|
|
|
628 ; -- The modified Base64 encoding deviates from standard Base64 in
|
|
|
629 ; that it omits the `=' pad character. This is eliminated so as to
|
|
|
630 ; avoid conflicts with the use of `=' as an escape in the
|
|
|
631 ; Quoted-Printable encoding and the related Q encoding for headers:
|
|
|
632 ; With this modification, non-whitespace chars in UTF-7 will be
|
|
|
633 ; represented in Quoted-Printable and in Q as-is, with no further
|
|
|
634 ; encoding.
|
|
|
635
|
|
|
636 ; For more information, see Appendix A.1 of The Unicode Standard 2.0, or
|
|
|
637 ; wherever it is in v3.0."
|
|
3767
|
638 ; unicode-type utf-7))
|
