ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/rxvt-unicode/src/encoding.C
(Generate patch)

Comparing rxvt-unicode/src/encoding.C (file contents):
Revision 1.3 by pcg, Wed Mar 3 17:15:49 2004 UTC vs.
Revision 1.21 by root, Fri Dec 10 19:55:55 2004 UTC

1/*--------------------------------*-C-*---------------------------------*
2 * File: encoding.C
3 *----------------------------------------------------------------------*
4 *
5 * All portions of code are copyright by their respective author/s.
6 * Copyright (c) 2003-2004 Marc Lehmann <pcg@goof.com>
7 *
8 * This program is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation; either version 2 of the License, or
11 * (at your option) any later version.
12 *
13 * This program is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
17 *
18 * You should have received a copy of the GNU General Public License
19 * along with this program; if not, write to the Free Software
20 * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
21 *----------------------------------------------------------------------*/
22
1#include "../config.h" 23#include "../config.h"
2 24
3#include "encoding.h" 25#include "encoding.h"
4 26
5#include <cstdlib> 27#include <cstdlib>
… …
26 { "ISO885914", CS_ISO8859_14 }, 48 { "ISO885914", CS_ISO8859_14 },
27 { "ISO885915", CS_ISO8859_15 }, 49 { "ISO885915", CS_ISO8859_15 },
28 { "FCD885915", CS_ISO8859_15 }, 50 { "FCD885915", CS_ISO8859_15 },
29 { "ISO885916", CS_ISO8859_16 }, 51 { "ISO885916", CS_ISO8859_16 },
30 52
53 { "TIS620*", CS_ISO8859_11 }, // close enough
54
31 { "ISO10646*", CS_UNICODE }, 55 { "ISO10646*", CS_UNICODE },
32 { "UNICODE", CS_UNICODE }, 56 { "UNICODE", CS_UNICODE },
33 { "UTF8", CS_UNICODE }, 57 { "UTF8", CS_UNICODE },
34 58
35 { "ASCII", CS_US_ASCII }, 59 { "ASCII", CS_US_ASCII },
… …
40 { "GOST1976874*", CS_KOI8_R }, 64 { "GOST1976874*", CS_KOI8_R },
41 { "KOI8RU", CS_KOI8_U }, 65 { "KOI8RU", CS_KOI8_U },
42 { "KOI8U", CS_KOI8_U }, 66 { "KOI8U", CS_KOI8_U },
43 67
44 { "VISCII*", CS_VISCII }, 68 { "VISCII*", CS_VISCII },
45 { "TIS62025291", CS_VISCII }, // close enough
46 69
47 { "JISX0201*", CS_JIS0201_1976_0 }, 70 { "JISX0201*", CS_JIS0201_1976_0 },
71 { "JISC6226*", CS_JIS0208_1990_0 }, // also wrongly matches -1987-0? (check Encode::JP)
48 { "JISX0208*", CS_JIS0208_1983_0 }, // also wrongly matches -1990-0 (check Encode::JP) 72 { "JISX0208*", CS_JIS0208_1990_0 }, // also wrongly matches -1987-0? (check Encode::JP)
49 { "JISX0212*", CS_JIS0212_1990_0 }, 73 { "JISX0212*", CS_JIS0212_1990_0 },
74 { "JISX021320001", CS_JIS0213_1 },
75 { "JISX021320002", CS_JIS0213_2 },
50 { "JISX0221*", CS_UNICODE }, 76 { "JISX0221*", CS_UNICODE }, // _very_ close
51 77
52 { "KSC5601*", CS_KSC5601_1987_0 }, 78 { "KSC5601*", CS_KSC5601_1987_0 },
53 { "KSX1001*", CS_KSC5601_1987_0 }, 79 { "KSX1001*", CS_KSC5601_1987_0 },
54 { "KSC5700*", CS_UNICODE }, // unicode plus extensions 80 { "KSC5700*", CS_UNICODE }, // unicode plus extensions
55 81
56 { "BIG5P*", CS_BIG5_PLUS }, 82 { "BIG5P*", CS_BIG5_PLUS },
57 { "BIG5ETEN*", CS_BIG5_EXT }, 83 { "BIG5ETEN*", CS_BIG5_EXT },
58 { "BIG5*", CS_BIG5 }, 84 { "BIG5*", CS_BIG5 },
59 { "GB2312*", CS_GB2312_1980_0 }, 85 { "GB2312*", CS_GB2312_1980_0 },
86 { "GBK*", CS_GBK_0 },
60 { "GB6345*", CS_GB2312_1980_0 }, // slightly different to gb2312?? 87 { "GB6345*", CS_GB2312_1980_0 }, // slightly different to gb2312??
61 { "GB8565*", CS_GB2312_1980_0 }, // a superset of gb2312?? 88 { "GB8565*", CS_GB2312_1980_0 }, // a superset of gb2312??
62 { "GB13000*", CS_UNICODE }, 89 { "GB13000*", CS_UNICODE },
63 { "CNS1164319921", CS_CNS11643_1992_1 }, 90 { "CNS1164319921", CS_CNS11643_1992_1 },
64 { "CNS1164319922", CS_CNS11643_1992_2 }, 91 { "CNS1164319922", CS_CNS11643_1992_2 },
… …
111 } while ((++i)->name); 138 } while ((++i)->name);
112 139
113 return CS_UNKNOWN; 140 return CS_UNKNOWN;
114} 141}
115 142
116struct rxvt_codeset_conv_unknown : rxvt_codeset_conv { 143static unicode_t cs_unknown_to_unicode (uint32_t enc) { return NOCHAR; }
117 uint32_t to_unicode (uint32_t enc) const { return NOCHAR; }
118 uint32_t from_unicode (uint32_t unicode) const { return NOCHAR; } 144static uint32_t cs_unknown_from_unicode (unicode_t unicode) { return NOCHAR; }
119} rxvt_codeset_conv_unknown;
120 145
121struct rxvt_codeset_conv_us_ascii : rxvt_codeset_conv { 146static unicode_t cs_unicode_to_unicode (uint32_t enc) { return enc; }
147static uint32_t cs_unicode_from_unicode (unicode_t unicode) { return unicode; }
148
149#define cs_us_ascii_to_unicode cs_unicode_to_unicode
122 uint32_t from_unicode (uint32_t unicode) const { return unicode <= 127 ? unicode : NOCHAR; } 150static uint32_t cs_us_ascii_from_unicode (unicode_t unicode) { return unicode <= 127 ? unicode : NOCHAR; }
123} rxvt_codeset_conv_us_ascii;
124 151
125struct rxvt_codeset_conv_unicode : rxvt_codeset_conv { 152#define cs_us_ascii_to_unicode_16 cs_unicode_to_unicode
126 /* transparent */
127} rxvt_codeset_conv_unicode;
128
129struct rxvt_codeset_conv_unicode_16 : rxvt_codeset_conv {
130 uint32_t to_unicode (uint32_t enc) const { return enc; }
131 uint32_t from_unicode (uint32_t unicode) const { return unicode <= 65535 ? unicode : NOCHAR; } 153static uint32_t cs_unicode_16_from_unicode (unicode_t unicode) { return unicode <= 65535 ? unicode : NOCHAR; }
132} rxvt_codeset_conv_unicode_16;
133
134/* block character set, must conform to the dec special pseudofont in defaultfont.C */
135struct rxvt_codeset_conv_special : rxvt_codeset_conv {
136 uint32_t to_unicode (uint32_t enc) const {
137 return enc;
138 }
139
140 uint32_t from_unicode (uint32_t unicode) const {
141 return unicode;
142 }
143} rxvt_codeset_conv_special;
144 154
145#define ENCODING_DEFAULT 155#define ENCODING_DEFAULT
146 156
147#include "table/iso8859_1.h" 157#include "table/iso8859_1.h"
148#include "table/iso8859_15.h" 158#include "table/iso8859_15.h"
… …
168 178
169//#define ENCODING_KR 179//#define ENCODING_KR
170 180
171#include "table/ksc5601_1987_0.h" 181#include "table/ksc5601_1987_0.h"
172 182
173//#define ENCODING_CN 183//#define ENCODING_ZH
174 184
185#include "table/big5.h"
186#include "table/gbk_0.h"
175#include "table/gb2312_1980_0.h" 187#include "table/gb2312_1980_0.h"
176#include "table/big5.h"
177 188
178//#define ENCODING_CN_EXT 189//#define ENCODING_ZH_EXT
179 190
180#include "table/cns11643_1992_1.h" 191#include "table/cns11643_1992_1.h"
181#include "table/cns11643_1992_2.h" 192#include "table/cns11643_1992_2.h"
182#include "table/cns11643_1992_3.h" 193#include "table/cns11643_1992_3.h"
183#include "table/cns11643_1992_4.h" 194#include "table/cns11643_1992_4.h"
… …
193#include "table/viscii.h" 204#include "table/viscii.h"
194 205
195//#define ENCODING_JP 206//#define ENCODING_JP
196 207
197#include "table/jis0201_1976_0.h" 208#include "table/jis0201_1976_0.h"
198#include "table/jis0208_1983_0.h" 209#include "table/jis0208_1990_0.h"
199#include "table/jis0212_1990_0.h" 210#include "table/jis0212_1990_0.h"
200 211
201//#define ENCODING_JP_EXT 212//#define ENCODING_JP_EXT
202 213
203#include "table/jis0213_1.h" 214#include "table/jis0213_1.h"
204#include "table/jis0213_2.h" 215#include "table/jis0213_2.h"
205 216
217#if ENCODING_TO_UNICODE
218# define ENC(base) { cs_ ## base ## _from_unicode, cs_ ## base ## _to_unicode }
219#else
220# define ENC(base) { cs_ ## base ## _from_unicode }
221#endif
222
223
206// order must match table in encoding.h(!) 224// order must match table in encoding.h(!)
207const rxvt_codeset_conv *rxvt_codeset[NUM_CODESETS] = { 225const rxvt_codeset_conv rxvt_codeset[NUM_CODESETS] = {
208 &rxvt_codeset_conv_unknown, 226 ENC (unknown),
209 &rxvt_codeset_conv_special,
210 227
211 &rxvt_codeset_conv_us_ascii, 228 ENC (us_ascii),
212 229
213 &rxvt_codeset_conv_iso8859_1, 230 ENC (iso8859_1),
214 &rxvt_codeset_conv_iso8859_2, 231 ENC (iso8859_2),
215 &rxvt_codeset_conv_iso8859_3, 232 ENC (iso8859_3),
216 &rxvt_codeset_conv_iso8859_4, 233 ENC (iso8859_4),
217 &rxvt_codeset_conv_iso8859_5, 234 ENC (iso8859_5),
218 &rxvt_codeset_conv_iso8859_6, 235 ENC (iso8859_6),
219 &rxvt_codeset_conv_iso8859_7, 236 ENC (iso8859_7),
220 &rxvt_codeset_conv_iso8859_8, 237 ENC (iso8859_8),
221 &rxvt_codeset_conv_iso8859_9, 238 ENC (iso8859_9),
222 &rxvt_codeset_conv_iso8859_10, 239 ENC (iso8859_10),
223 &rxvt_codeset_conv_iso8859_11, 240 ENC (iso8859_11),
224 &rxvt_codeset_conv_iso8859_13, 241 ENC (iso8859_13),
225 &rxvt_codeset_conv_iso8859_14, 242 ENC (iso8859_14),
226 &rxvt_codeset_conv_iso8859_15, 243 ENC (iso8859_15),
227 &rxvt_codeset_conv_iso8859_16, 244 ENC (iso8859_16),
228 245
229 &rxvt_codeset_conv_koi8_r, 246 ENC (koi8_r),
230 &rxvt_codeset_conv_koi8_u, 247 ENC (koi8_u),
231 248
232 &rxvt_codeset_conv_jis0201_1976_0, 249 ENC (jis0201_1976_0),
233 &rxvt_codeset_conv_jis0208_1983_0, 250 ENC (jis0208_1990_0),
234 &rxvt_codeset_conv_jis0212_1990_0, 251 ENC (jis0212_1990_0),
235 252
236 &rxvt_codeset_conv_jis0213_1, 253 ENC (jis0213_1),
237 &rxvt_codeset_conv_jis0213_2, 254 ENC (jis0213_2),
238 255
239 &rxvt_codeset_conv_ksc5601_1987_0, 256 ENC (ksc5601_1987_0),
240 257
241 &rxvt_codeset_conv_gb2312_1980_0, 258 ENC (gb2312_1980_0),
259 ENC (gbk_0),
242 260
243 &rxvt_codeset_conv_cns11643_1992_1, 261 ENC (cns11643_1992_1),
244 &rxvt_codeset_conv_cns11643_1992_2, 262 ENC (cns11643_1992_2),
245 &rxvt_codeset_conv_cns11643_1992_3, 263 ENC (cns11643_1992_3),
246 &rxvt_codeset_conv_cns11643_1992_4, 264 ENC (cns11643_1992_4),
247 &rxvt_codeset_conv_cns11643_1992_5, 265 ENC (cns11643_1992_5),
248 &rxvt_codeset_conv_cns11643_1992_6, 266 ENC (cns11643_1992_6),
249 &rxvt_codeset_conv_cns11643_1992_7, 267 ENC (cns11643_1992_7),
250 &rxvt_codeset_conv_cns11643_1992_f, 268 ENC (cns11643_1992_f),
251 &rxvt_codeset_conv_big5, 269 ENC (big5),
252 &rxvt_codeset_conv_big5_ext, 270 ENC (big5_ext),
253 &rxvt_codeset_conv_big5_plus, 271 ENC (big5_plus),
254 272
255 &rxvt_codeset_conv_viscii, 273 ENC (viscii),
256 274
257 &rxvt_codeset_conv_unicode_16, 275 ENC (unicode_16),
258 &rxvt_codeset_conv_unicode 276 ENC (unicode),
259}; 277};
260 278
279#if ENABLE_COMBINING
280# define ENCODING_COMPOSE
281#endif
282
283#include "table/compose.h"
284
285unicode_t
286rxvt_compose (unicode_t c1, unicode_t c2)
287{
288 int l = 0;
289 int r = sizeof (rxvt_compose_table) / sizeof (rxvt_compose_entry) - 1;
290 int m;
291
292 while (r > l)
293 {
294 m = (l + r) / 2;
295 rxvt_compose_entry &c = rxvt_compose_table[m];
296
297 if (c.c1 < c1 || (c.c1 == c1 && c.c2 < c2))
298 l = m + 1;
299 else if (c.c1 > c1 || (c.c1 == c1 && c.c2 > c2))
300 r = m - 1;
301 else
302 return c.r;
303 }
304
305 return NOCHAR;
306}
307
308#include "table/category.h"
309
310bool unicode::is_space (unicode_t c)
311{
312 return IS_SPACE (c)
313 || c == 0x09; // exclude tabs, too, as we store them in the buffer
314}

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines