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 1.15 by pcg, Mon Mar 15 01:27:46 2004 UTC vs.
Revision 1.24 by root, Mon Feb 20 22:40:34 2006 UTC

1/*--------------------------------*-C-*---------------------------------*
2 * File: encoding.C
3 *----------------------------------------------------------------------*
4 *
5 * All portions of code are copyright by their respective author/s.
6 * Copyright (c) 2003-2006 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>
… …
35 { "UTF8", CS_UNICODE }, 57 { "UTF8", CS_UNICODE },
36 58
37 { "ASCII", CS_US_ASCII }, 59 { "ASCII", CS_US_ASCII },
38 { "USASCII", CS_US_ASCII }, 60 { "USASCII", CS_US_ASCII },
39 { "ANSIX341968", CS_US_ASCII }, 61 { "ANSIX341968", CS_US_ASCII },
62 { "ISO646.1991-IRV", CS_US_ASCII }, // older versions used the currency sign
40 63
41 { "KOI8R", CS_KOI8_R }, 64 { "KOI8R*", CS_KOI8_R },
42 { "GOST1976874*", CS_KOI8_R }, 65 { "GOST1976874*", CS_KOI8_R },
43 { "KOI8RU", CS_KOI8_U }, 66 { "KOI8RU", CS_KOI8_U },
44 { "KOI8U", CS_KOI8_U }, 67 { "KOI8U", CS_KOI8_U },
45 68
46 { "VISCII*", CS_VISCII }, 69 { "VISCII*", CS_VISCII },
47 70
48 { "JISX0201*", CS_JIS0201_1976_0 }, 71 { "JISX0201*", CS_JIS0201_1976_0 },
72 { "JISC6226*", CS_JIS0208_1990_0 }, // also wrongly matches -1987-0? (check Encode::JP)
49 { "JISX0208*", CS_JIS0208_1990_0 }, // also wrongly matches -1987-0? (check Encode::JP) 73 { "JISX0208*", CS_JIS0208_1990_0 }, // also wrongly matches -1987-0? (check Encode::JP)
50 { "JISX0212*", CS_JIS0212_1990_0 }, 74 { "JISX0212*", CS_JIS0212_1990_0 },
51 { "JISX021320001", CS_JIS0213_1 }, 75 { "JISX021320001", CS_JIS0213_1 },
52 { "JISX021320002", CS_JIS0213_2 }, 76 { "JISX021320002", CS_JIS0213_2 },
53 { "JISX0221*", CS_UNICODE }, // _very_ close 77 { "JISX0221*", CS_UNICODE }, // _very_ close
… …
58 82
59 { "BIG5P*", CS_BIG5_PLUS }, 83 { "BIG5P*", CS_BIG5_PLUS },
60 { "BIG5ETEN*", CS_BIG5_EXT }, 84 { "BIG5ETEN*", CS_BIG5_EXT },
61 { "BIG5*", CS_BIG5 }, 85 { "BIG5*", CS_BIG5 },
62 { "GB2312*", CS_GB2312_1980_0 }, 86 { "GB2312*", CS_GB2312_1980_0 },
87 { "GBK*", CS_GBK_0 },
63 { "GB6345*", CS_GB2312_1980_0 }, // slightly different to gb2312?? 88 { "GB6345*", CS_GB2312_1980_0 }, // slightly different to gb2312??
64 { "GB8565*", CS_GB2312_1980_0 }, // a superset of gb2312?? 89 { "GB8565*", CS_GB2312_1980_0 }, // a superset of gb2312??
65 { "GB13000*", CS_UNICODE }, 90 { "GB13000*", CS_UNICODE },
66 { "CNS1164319921", CS_CNS11643_1992_1 }, 91 { "CNS1164319921", CS_CNS11643_1992_1 },
67 { "CNS1164319922", CS_CNS11643_1992_2 }, 92 { "CNS1164319922", CS_CNS11643_1992_2 },
… …
114 } while ((++i)->name); 139 } while ((++i)->name);
115 140
116 return CS_UNKNOWN; 141 return CS_UNKNOWN;
117} 142}
118 143
119struct rxvt_codeset_conv_unknown : rxvt_codeset_conv {
120 unicode_t to_unicode (uint32_t enc) const { return NOCHAR; } 144static unicode_t cs_unknown_to_unicode (uint32_t enc) { return NOCHAR; }
121 uint32_t from_unicode (unicode_t unicode) const { return NOCHAR; } 145static uint32_t cs_unknown_from_unicode (unicode_t unicode) { return NOCHAR; }
122} rxvt_codeset_conv_unknown;
123 146
124struct rxvt_codeset_conv_us_ascii : rxvt_codeset_conv { 147static unicode_t cs_unicode_to_unicode (uint32_t enc) { return enc; }
148static uint32_t cs_unicode_from_unicode (unicode_t unicode) { return unicode; }
149
150#define cs_us_ascii_to_unicode cs_unicode_to_unicode
125 uint32_t from_unicode (unicode_t unicode) const { return unicode <= 127 ? unicode : NOCHAR; } 151static uint32_t cs_us_ascii_from_unicode (unicode_t unicode) { return unicode <= 127 ? unicode : NOCHAR; }
126} rxvt_codeset_conv_us_ascii;
127 152
128struct rxvt_codeset_conv_unicode : rxvt_codeset_conv { 153#define cs_us_ascii_to_unicode_16 cs_unicode_to_unicode
129 /* transparent */
130} rxvt_codeset_conv_unicode;
131
132struct rxvt_codeset_conv_unicode_16 : rxvt_codeset_conv {
133 unicode_t to_unicode (uint32_t enc) const { return enc; }
134 uint32_t from_unicode (unicode_t unicode) const { return unicode <= 65535 ? unicode : NOCHAR; } 154static uint32_t cs_unicode_16_from_unicode (unicode_t unicode) { return unicode <= 65535 ? unicode : NOCHAR; }
135} rxvt_codeset_conv_unicode_16;
136 155
137#define ENCODING_DEFAULT 156#define ENCODING_DEFAULT
138 157
139#include "table/iso8859_1.h" 158#include "table/iso8859_1.h"
140#include "table/iso8859_15.h" 159#include "table/iso8859_15.h"
… …
160 179
161//#define ENCODING_KR 180//#define ENCODING_KR
162 181
163#include "table/ksc5601_1987_0.h" 182#include "table/ksc5601_1987_0.h"
164 183
165//#define ENCODING_CN 184//#define ENCODING_ZH
166 185
186#include "table/big5.h"
187#include "table/gbk_0.h"
167#include "table/gb2312_1980_0.h" 188#include "table/gb2312_1980_0.h"
168#include "table/big5.h"
169 189
170//#define ENCODING_CN_EXT 190//#define ENCODING_ZH_EXT
171 191
172#include "table/cns11643_1992_1.h" 192#include "table/cns11643_1992_1.h"
173#include "table/cns11643_1992_2.h" 193#include "table/cns11643_1992_2.h"
174#include "table/cns11643_1992_3.h" 194#include "table/cns11643_1992_3.h"
175#include "table/cns11643_1992_4.h" 195#include "table/cns11643_1992_4.h"
… …
193//#define ENCODING_JP_EXT 213//#define ENCODING_JP_EXT
194 214
195#include "table/jis0213_1.h" 215#include "table/jis0213_1.h"
196#include "table/jis0213_2.h" 216#include "table/jis0213_2.h"
197 217
218#if ENCODING_TO_UNICODE
219# define ENC(base) { cs_ ## base ## _from_unicode, cs_ ## base ## _to_unicode }
220#else
221# define ENC(base) { cs_ ## base ## _from_unicode }
222#endif
223
224
198// order must match table in encoding.h(!) 225// order must match table in encoding.h(!)
199const rxvt_codeset_conv *rxvt_codeset[NUM_CODESETS] = { 226const rxvt_codeset_conv rxvt_codeset[NUM_CODESETS] = {
200 &rxvt_codeset_conv_unknown, 227 ENC (unknown),
201 228
202 &rxvt_codeset_conv_us_ascii, 229 ENC (us_ascii),
203 230
204 &rxvt_codeset_conv_iso8859_1, 231 ENC (iso8859_1),
205 &rxvt_codeset_conv_iso8859_2, 232 ENC (iso8859_2),
206 &rxvt_codeset_conv_iso8859_3, 233 ENC (iso8859_3),
207 &rxvt_codeset_conv_iso8859_4, 234 ENC (iso8859_4),
208 &rxvt_codeset_conv_iso8859_5, 235 ENC (iso8859_5),
209 &rxvt_codeset_conv_iso8859_6, 236 ENC (iso8859_6),
210 &rxvt_codeset_conv_iso8859_7, 237 ENC (iso8859_7),
211 &rxvt_codeset_conv_iso8859_8, 238 ENC (iso8859_8),
212 &rxvt_codeset_conv_iso8859_9, 239 ENC (iso8859_9),
213 &rxvt_codeset_conv_iso8859_10, 240 ENC (iso8859_10),
214 &rxvt_codeset_conv_iso8859_11, 241 ENC (iso8859_11),
215 &rxvt_codeset_conv_iso8859_13, 242 ENC (iso8859_13),
216 &rxvt_codeset_conv_iso8859_14, 243 ENC (iso8859_14),
217 &rxvt_codeset_conv_iso8859_15, 244 ENC (iso8859_15),
218 &rxvt_codeset_conv_iso8859_16, 245 ENC (iso8859_16),
219 246
220 &rxvt_codeset_conv_koi8_r, 247 ENC (koi8_r),
221 &rxvt_codeset_conv_koi8_u, 248 ENC (koi8_u),
222 249
223 &rxvt_codeset_conv_jis0201_1976_0, 250 ENC (jis0201_1976_0),
224 &rxvt_codeset_conv_jis0208_1990_0, 251 ENC (jis0208_1990_0),
225 &rxvt_codeset_conv_jis0212_1990_0, 252 ENC (jis0212_1990_0),
226 253
227 &rxvt_codeset_conv_jis0213_1, 254 ENC (jis0213_1),
228 &rxvt_codeset_conv_jis0213_2, 255 ENC (jis0213_2),
229 256
230 &rxvt_codeset_conv_ksc5601_1987_0, 257 ENC (ksc5601_1987_0),
231 258
232 &rxvt_codeset_conv_gb2312_1980_0, 259 ENC (gb2312_1980_0),
260 ENC (gbk_0),
233 261
234 &rxvt_codeset_conv_cns11643_1992_1, 262 ENC (cns11643_1992_1),
235 &rxvt_codeset_conv_cns11643_1992_2, 263 ENC (cns11643_1992_2),
236 &rxvt_codeset_conv_cns11643_1992_3, 264 ENC (cns11643_1992_3),
237 &rxvt_codeset_conv_cns11643_1992_4, 265 ENC (cns11643_1992_4),
238 &rxvt_codeset_conv_cns11643_1992_5, 266 ENC (cns11643_1992_5),
239 &rxvt_codeset_conv_cns11643_1992_6, 267 ENC (cns11643_1992_6),
240 &rxvt_codeset_conv_cns11643_1992_7, 268 ENC (cns11643_1992_7),
241 &rxvt_codeset_conv_cns11643_1992_f, 269 ENC (cns11643_1992_f),
242 &rxvt_codeset_conv_big5, 270 ENC (big5),
243 &rxvt_codeset_conv_big5_ext, 271 ENC (big5_ext),
244 &rxvt_codeset_conv_big5_plus, 272 ENC (big5_plus),
245 273
246 &rxvt_codeset_conv_viscii, 274 ENC (viscii),
247 275
248 &rxvt_codeset_conv_unicode_16, 276 ENC (unicode_16),
249 &rxvt_codeset_conv_unicode 277 ENC (unicode),
250}; 278};
251 279
252#if ENABLE_COMBINING 280#if ENABLE_COMBINING
253# define ENCODING_COMPOSE 281# define ENCODING_COMPOSE
254#endif 282#endif
… …
280 308
281#include "table/category.h" 309#include "table/category.h"
282 310
283bool unicode::is_space (unicode_t c) 311bool unicode::is_space (unicode_t c)
284{ 312{
285 return IS_SPACE (c); 313 return IS_SPACE (c)
314 || c == 0x09; // exclude tabs, too, as we store them in the buffer
286} 315}

Diff Legend

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