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 1.20 by root, Sun Sep 5 09:32:59 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
… …
111 } while ((++i)->name); 137 } while ((++i)->name);
112 138
113 return CS_UNKNOWN; 139 return CS_UNKNOWN;
114} 140}
115 141
116struct rxvt_codeset_conv_unknown : rxvt_codeset_conv { 142static 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; } 143static uint32_t cs_unknown_from_unicode (unicode_t unicode) { return NOCHAR; }
119} rxvt_codeset_conv_unknown;
120 144
121struct rxvt_codeset_conv_us_ascii : rxvt_codeset_conv { 145static unicode_t cs_unicode_to_unicode (uint32_t enc) { return enc; }
146static uint32_t cs_unicode_from_unicode (unicode_t unicode) { return unicode; }
147
148#define cs_us_ascii_to_unicode cs_unicode_to_unicode
122 uint32_t from_unicode (uint32_t unicode) const { return unicode <= 127 ? unicode : NOCHAR; } 149static uint32_t cs_us_ascii_from_unicode (unicode_t unicode) { return unicode <= 127 ? unicode : NOCHAR; }
123} rxvt_codeset_conv_us_ascii;
124 150
125struct rxvt_codeset_conv_unicode : rxvt_codeset_conv { 151#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; } 152static 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;
144 153
145#define ENCODING_DEFAULT 154#define ENCODING_DEFAULT
146 155
147#include "table/iso8859_1.h" 156#include "table/iso8859_1.h"
148#include "table/iso8859_15.h" 157#include "table/iso8859_15.h"
… …
168 177
169//#define ENCODING_KR 178//#define ENCODING_KR
170 179
171#include "table/ksc5601_1987_0.h" 180#include "table/ksc5601_1987_0.h"
172 181
173//#define ENCODING_CN 182//#define ENCODING_ZH
174 183
175#include "table/gb2312_1980_0.h" 184#include "table/gb2312_1980_0.h"
176#include "table/big5.h" 185#include "table/big5.h"
177 186
178//#define ENCODING_CN_EXT 187//#define ENCODING_ZH_EXT
179 188
180#include "table/cns11643_1992_1.h" 189#include "table/cns11643_1992_1.h"
181#include "table/cns11643_1992_2.h" 190#include "table/cns11643_1992_2.h"
182#include "table/cns11643_1992_3.h" 191#include "table/cns11643_1992_3.h"
183#include "table/cns11643_1992_4.h" 192#include "table/cns11643_1992_4.h"
… …
193#include "table/viscii.h" 202#include "table/viscii.h"
194 203
195//#define ENCODING_JP 204//#define ENCODING_JP
196 205
197#include "table/jis0201_1976_0.h" 206#include "table/jis0201_1976_0.h"
198#include "table/jis0208_1983_0.h" 207#include "table/jis0208_1990_0.h"
199#include "table/jis0212_1990_0.h" 208#include "table/jis0212_1990_0.h"
200 209
201//#define ENCODING_JP_EXT 210//#define ENCODING_JP_EXT
202 211
203#include "table/jis0213_1.h" 212#include "table/jis0213_1.h"
204#include "table/jis0213_2.h" 213#include "table/jis0213_2.h"
205 214
215#if ENCODING_TO_UNICODE
216# define ENC(base) { cs_ ## base ## _from_unicode, cs_ ## base ## _to_unicode }
217#else
218# define ENC(base) { cs_ ## base ## _from_unicode }
219#endif
220
221
206// order must match table in encoding.h(!) 222// order must match table in encoding.h(!)
207const rxvt_codeset_conv *rxvt_codeset[NUM_CODESETS] = { 223const rxvt_codeset_conv rxvt_codeset[NUM_CODESETS] = {
208 &rxvt_codeset_conv_unknown, 224 ENC (unknown),
209 &rxvt_codeset_conv_special,
210 225
211 &rxvt_codeset_conv_us_ascii, 226 ENC (us_ascii),
212 227
213 &rxvt_codeset_conv_iso8859_1, 228 ENC (iso8859_1),
214 &rxvt_codeset_conv_iso8859_2, 229 ENC (iso8859_2),
215 &rxvt_codeset_conv_iso8859_3, 230 ENC (iso8859_3),
216 &rxvt_codeset_conv_iso8859_4, 231 ENC (iso8859_4),
217 &rxvt_codeset_conv_iso8859_5, 232 ENC (iso8859_5),
218 &rxvt_codeset_conv_iso8859_6, 233 ENC (iso8859_6),
219 &rxvt_codeset_conv_iso8859_7, 234 ENC (iso8859_7),
220 &rxvt_codeset_conv_iso8859_8, 235 ENC (iso8859_8),
221 &rxvt_codeset_conv_iso8859_9, 236 ENC (iso8859_9),
222 &rxvt_codeset_conv_iso8859_10, 237 ENC (iso8859_10),
223 &rxvt_codeset_conv_iso8859_11, 238 ENC (iso8859_11),
224 &rxvt_codeset_conv_iso8859_13, 239 ENC (iso8859_13),
225 &rxvt_codeset_conv_iso8859_14, 240 ENC (iso8859_14),
226 &rxvt_codeset_conv_iso8859_15, 241 ENC (iso8859_15),
227 &rxvt_codeset_conv_iso8859_16, 242 ENC (iso8859_16),
228 243
229 &rxvt_codeset_conv_koi8_r, 244 ENC (koi8_r),
230 &rxvt_codeset_conv_koi8_u, 245 ENC (koi8_u),
231 246
232 &rxvt_codeset_conv_jis0201_1976_0, 247 ENC (jis0201_1976_0),
233 &rxvt_codeset_conv_jis0208_1983_0, 248 ENC (jis0208_1990_0),
234 &rxvt_codeset_conv_jis0212_1990_0, 249 ENC (jis0212_1990_0),
235 250
236 &rxvt_codeset_conv_jis0213_1, 251 ENC (jis0213_1),
237 &rxvt_codeset_conv_jis0213_2, 252 ENC (jis0213_2),
238 253
239 &rxvt_codeset_conv_ksc5601_1987_0, 254 ENC (ksc5601_1987_0),
240 255
241 &rxvt_codeset_conv_gb2312_1980_0, 256 ENC (gb2312_1980_0),
242 257
243 &rxvt_codeset_conv_cns11643_1992_1, 258 ENC (cns11643_1992_1),
244 &rxvt_codeset_conv_cns11643_1992_2, 259 ENC (cns11643_1992_2),
245 &rxvt_codeset_conv_cns11643_1992_3, 260 ENC (cns11643_1992_3),
246 &rxvt_codeset_conv_cns11643_1992_4, 261 ENC (cns11643_1992_4),
247 &rxvt_codeset_conv_cns11643_1992_5, 262 ENC (cns11643_1992_5),
248 &rxvt_codeset_conv_cns11643_1992_6, 263 ENC (cns11643_1992_6),
249 &rxvt_codeset_conv_cns11643_1992_7, 264 ENC (cns11643_1992_7),
250 &rxvt_codeset_conv_cns11643_1992_f, 265 ENC (cns11643_1992_f),
251 &rxvt_codeset_conv_big5, 266 ENC (big5),
252 &rxvt_codeset_conv_big5_ext, 267 ENC (big5_ext),
253 &rxvt_codeset_conv_big5_plus, 268 ENC (big5_plus),
254 269
255 &rxvt_codeset_conv_viscii, 270 ENC (viscii),
256 271
257 &rxvt_codeset_conv_unicode_16, 272 ENC (unicode_16),
258 &rxvt_codeset_conv_unicode 273 ENC (unicode),
259}; 274};
260 275
276#if ENABLE_COMBINING
277# define ENCODING_COMPOSE
278#endif
279
280#include "table/compose.h"
281
282unicode_t
283rxvt_compose (unicode_t c1, unicode_t c2)
284{
285 int l = 0;
286 int r = sizeof (rxvt_compose_table) / sizeof (rxvt_compose_entry) - 1;
287 int m;
288
289 while (r > l)
290 {
291 m = (l + r) / 2;
292 rxvt_compose_entry &c = rxvt_compose_table[m];
293
294 if (c.c1 < c1 || (c.c1 == c1 && c.c2 < c2))
295 l = m + 1;
296 else if (c.c1 > c1 || (c.c1 == c1 && c.c2 > c2))
297 r = m - 1;
298 else
299 return c.r;
300 }
301
302 return NOCHAR;
303}
304
305#include "table/category.h"
306
307bool unicode::is_space (unicode_t c)
308{
309 return IS_SPACE (c)
310 || c == 0x09; // exclude tabs, too, as we store them in the buffer
311}

Diff Legend

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