1 /*
2 * "streamable kanji code filter and converter"
3 * Copyright (c) 1998-2002 HappySize, Inc. All rights reserved.
4 *
5 * LICENSE NOTICES
6 *
7 * This file is part of "streamable kanji code filter and converter",
8 * which is distributed under the terms of GNU Lesser General Public
9 * License (version 2) as published by the Free Software Foundation.
10 *
11 * This software is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with "streamable kanji code filter and converter";
18 * if not, write to the Free Software Foundation, Inc., 59 Temple Place,
19 * Suite 330, Boston, MA 02111-1307 USA
20 *
21 * The author of this file: Rui Hirokawa <hirokawa@php.net>
22 *
23 */
24 /*
25 * The source code included in this files was separated from mbfilter_tw.c
26 * by moriyoshi koizumi <moriyoshi@php.net> on 4 dec 2002.
27 *
28 */
29
30 #include "mbfilter.h"
31 #include "mbfilter_big5.h"
32
33 #include "unicode_table_big5.h"
34
35 static int mbfl_filt_ident_big5(int c, mbfl_identify_filter *filter);
36
37 static const unsigned char mblen_table_big5[] = { /* 0x81-0xFE */
38 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
39 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
40 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
41 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
42 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
43 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
44 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
45 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
46 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
47 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
48 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
49 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
50 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
51 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
52 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
53 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1
54 };
55
56 static const char *mbfl_encoding_big5_aliases[] = {"CN-BIG5", "BIG-FIVE", "BIGFIVE", NULL};
57
58 const mbfl_encoding mbfl_encoding_big5 = {
59 mbfl_no_encoding_big5,
60 "BIG-5",
61 "BIG5",
62 (const char *(*)[])&mbfl_encoding_big5_aliases,
63 mblen_table_big5,
64 MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE,
65 &vtbl_big5_wchar,
66 &vtbl_wchar_big5
67 };
68
69 const mbfl_encoding mbfl_encoding_cp950 = {
70 mbfl_no_encoding_cp950,
71 "CP950",
72 "BIG5",
73 NULL,
74 mblen_table_big5,
75 MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE,
76 &vtbl_cp950_wchar,
77 &vtbl_wchar_cp950
78 };
79
80 const struct mbfl_identify_vtbl vtbl_identify_big5 = {
81 mbfl_no_encoding_big5,
82 mbfl_filt_ident_common_ctor,
83 mbfl_filt_ident_big5
84 };
85
86 const struct mbfl_identify_vtbl vtbl_identify_cp950 = {
87 mbfl_no_encoding_cp950,
88 mbfl_filt_ident_common_ctor,
89 mbfl_filt_ident_big5
90 };
91
92 const struct mbfl_convert_vtbl vtbl_big5_wchar = {
93 mbfl_no_encoding_big5,
94 mbfl_no_encoding_wchar,
95 mbfl_filt_conv_common_ctor,
96 NULL,
97 mbfl_filt_conv_big5_wchar,
98 mbfl_filt_conv_common_flush,
99 NULL,
100 };
101
102 const struct mbfl_convert_vtbl vtbl_wchar_big5 = {
103 mbfl_no_encoding_wchar,
104 mbfl_no_encoding_big5,
105 mbfl_filt_conv_common_ctor,
106 NULL,
107 mbfl_filt_conv_wchar_big5,
108 mbfl_filt_conv_common_flush,
109 NULL
110 };
111
112 const struct mbfl_convert_vtbl vtbl_cp950_wchar = {
113 mbfl_no_encoding_cp950,
114 mbfl_no_encoding_wchar,
115 mbfl_filt_conv_common_ctor,
116 NULL,
117 mbfl_filt_conv_big5_wchar,
118 mbfl_filt_conv_common_flush,
119 NULL,
120 };
121
122 const struct mbfl_convert_vtbl vtbl_wchar_cp950 = {
123 mbfl_no_encoding_wchar,
124 mbfl_no_encoding_cp950,
125 mbfl_filt_conv_common_ctor,
126 NULL,
127 mbfl_filt_conv_wchar_big5,
128 mbfl_filt_conv_common_flush,
129 NULL,
130 };
131
132 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
133
134 /* 63 + 94 = 157 or 94 */
135 static unsigned short cp950_pua_tbl[][4] = {
136 {0xe000,0xe310,0xfa40,0xfefe},
137 {0xe311,0xeeb7,0x8e40,0xa0fe},
138 {0xeeb8,0xf6b0,0x8140,0x8dfe},
139 {0xf6b1,0xf70e,0xc6a1,0xc6fe},
140 {0xf70f,0xf848,0xc740,0xc8fe},
141 };
142
is_in_cp950_pua(int c1,int c)143 static inline int is_in_cp950_pua(int c1, int c) {
144 if ((c1 >= 0xfa && c1 <= 0xfe) || (c1 >= 0x8e && c1 <= 0xa0) ||
145 (c1 >= 0x81 && c1 <= 0x8d) || (c1 >= 0xc7 && c1 <= 0xc8)) {
146 return (c >=0x40 && c <= 0x7e) || (c >= 0xa1 && c <= 0xfe);
147 }
148 if (c1 == 0xc6) {
149 return c >= 0xa1 && c <= 0xfe;
150 }
151 return 0;
152 }
153
154 /*
155 * Big5 => wchar
156 */
157 int
mbfl_filt_conv_big5_wchar(int c,mbfl_convert_filter * filter)158 mbfl_filt_conv_big5_wchar(int c, mbfl_convert_filter *filter)
159 {
160 int k;
161 int c1, w, c2;
162
163 switch (filter->status) {
164 case 0:
165 if (filter->from->no_encoding == mbfl_no_encoding_cp950) {
166 c1 = 0x80;
167 } else {
168 c1 = 0xa0;
169 }
170
171 if (c >= 0 && c <= 0x80) { /* latin */
172 CK((*filter->output_function)(c, filter->data));
173 } else if (c == 0xff) {
174 CK((*filter->output_function)(0xf8f8, filter->data));
175 } else if (c > c1 && c < 0xff) { /* dbcs lead byte */
176 filter->status = 1;
177 filter->cache = c;
178 } else {
179 w = c & MBFL_WCSGROUP_MASK;
180 w |= MBFL_WCSGROUP_THROUGH;
181 CK((*filter->output_function)(w, filter->data));
182 }
183 break;
184
185 case 1: /* dbcs second byte */
186 filter->status = 0;
187 c1 = filter->cache;
188 if ((c > 0x39 && c < 0x7f) | (c > 0xa0 && c < 0xff)) {
189 if (c < 0x7f){
190 w = (c1 - 0xa1)*157 + (c - 0x40);
191 } else {
192 w = (c1 - 0xa1)*157 + (c - 0xa1) + 0x3f;
193 }
194 if (w >= 0 && w < big5_ucs_table_size) {
195 w = big5_ucs_table[w];
196 } else {
197 w = 0;
198 }
199
200 if (filter->from->no_encoding == mbfl_no_encoding_cp950) {
201 /* PUA for CP950 */
202 if (w <= 0 && is_in_cp950_pua(c1, c)) {
203 c2 = c1 << 8 | c;
204 for (k = 0; k < sizeof(cp950_pua_tbl)/(sizeof(unsigned short)*4); k++) {
205 if (c2 >= cp950_pua_tbl[k][2] && c2 <= cp950_pua_tbl[k][3]) {
206 break;
207 }
208 }
209
210 if ((cp950_pua_tbl[k][2] & 0xff) == 0x40) {
211 w = 157*(c1 - (cp950_pua_tbl[k][2]>>8)) + c - (c >= 0xa1 ? 0x62 : 0x40)
212 + cp950_pua_tbl[k][0];
213 } else {
214 w = c2 - cp950_pua_tbl[k][2] + cp950_pua_tbl[k][0];
215 }
216 }
217 }
218
219 if (w <= 0) {
220 w = (c1 << 8) | c;
221 w &= MBFL_WCSPLANE_MASK;
222 w |= MBFL_WCSPLANE_BIG5;
223 }
224 CK((*filter->output_function)(w, filter->data));
225 } else if ((c >= 0 && c < 0x21) || c == 0x7f) { /* CTLs */
226 CK((*filter->output_function)(c, filter->data));
227 } else {
228 w = (c1 << 8) | c;
229 w &= MBFL_WCSGROUP_MASK;
230 w |= MBFL_WCSGROUP_THROUGH;
231 CK((*filter->output_function)(w, filter->data));
232 }
233 break;
234
235 default:
236 filter->status = 0;
237 break;
238 }
239
240 return c;
241 }
242
243 /*
244 * wchar => Big5
245 */
246 int
mbfl_filt_conv_wchar_big5(int c,mbfl_convert_filter * filter)247 mbfl_filt_conv_wchar_big5(int c, mbfl_convert_filter *filter)
248 {
249 int k;
250 int c1, s, c2;
251
252 s = 0;
253 if (c >= ucs_a1_big5_table_min && c < ucs_a1_big5_table_max) {
254 s = ucs_a1_big5_table[c - ucs_a1_big5_table_min];
255 } else if (c >= ucs_a2_big5_table_min && c < ucs_a2_big5_table_max) {
256 s = ucs_a2_big5_table[c - ucs_a2_big5_table_min];
257 } else if (c >= ucs_a3_big5_table_min && c < ucs_a3_big5_table_max) {
258 s = ucs_a3_big5_table[c - ucs_a3_big5_table_min];
259 } else if (c >= ucs_i_big5_table_min && c < ucs_i_big5_table_max) {
260 s = ucs_i_big5_table[c - ucs_i_big5_table_min];
261 } else if (c >= ucs_pua_big5_table_min && c < ucs_pua_big5_table_max) {
262 s = ucs_pua_big5_table[c - ucs_pua_big5_table_min];
263 } else if (c >= ucs_r1_big5_table_min && c < ucs_r1_big5_table_max) {
264 s = ucs_r1_big5_table[c - ucs_r1_big5_table_min];
265 } else if (c >= ucs_r2_big5_table_min && c < ucs_r2_big5_table_max) {
266 s = ucs_r2_big5_table[c - ucs_r2_big5_table_min];
267 }
268
269 if (filter->to->no_encoding == mbfl_no_encoding_cp950) {
270 if (c >= 0xe000 && c <= 0xf848) { /* PUA for CP950 */
271 for (k = 0; k < sizeof(cp950_pua_tbl)/(sizeof(unsigned short)*4); k++) {
272 if (c <= cp950_pua_tbl[k][1]) {
273 break;
274 }
275 }
276 c1 = c - cp950_pua_tbl[k][0];
277 if ((cp950_pua_tbl[k][2] & 0xff) == 0x40) {
278 c2 = cp950_pua_tbl[k][2] >> 8;
279 s = ((c1 / 157) + c2) << 8; c1 %= 157;
280 s |= c1 + (c1 >= 0x3f ? 0x62 : 0x40);
281 } else {
282 s = c1 + cp950_pua_tbl[k][2];
283 }
284 }
285
286 if (c == 0x80) {
287 s = 0x80;
288 } else if (c == 0xf8f8) {
289 s = 0xff;
290 } else if (c == 0x256d) {
291 s = 0xa27e;
292 } else if (c == 0x256e) {
293 s = 0xa2a1;
294 } else if (c == 0x256f) {
295 s = 0xa2a3;
296 } else if (c == 0x2570) {
297 s = 0xa2a2;
298 }
299 }
300
301 if (s <= 0) {
302 c1 = c & ~MBFL_WCSPLANE_MASK;
303 if (c1 == MBFL_WCSPLANE_BIG5) {
304 s = c & MBFL_WCSPLANE_MASK;
305 }
306 if (c == 0) {
307 s = 0;
308 } else if (s <= 0) {
309 s = -1;
310 }
311 }
312 if (s >= 0) {
313 if (s <= 0x80 || s == 0xff) { /* latin */
314 CK((*filter->output_function)(s, filter->data));
315 } else {
316 CK((*filter->output_function)((s >> 8) & 0xff, filter->data));
317 CK((*filter->output_function)(s & 0xff, filter->data));
318 }
319 } else {
320 CK(mbfl_filt_conv_illegal_output(c, filter));
321 }
322
323 return c;
324 }
325
mbfl_filt_ident_big5(int c,mbfl_identify_filter * filter)326 static int mbfl_filt_ident_big5(int c, mbfl_identify_filter *filter)
327 {
328 int c1;
329 if (filter->encoding->no_encoding == mbfl_no_encoding_cp950) {
330 c1 = 0x80;
331 } else {
332 c1 = 0xa0;
333 }
334
335 if (filter->status) { /* kanji second char */
336 if (c < 0x40 || (c > 0x7e && c < 0xa1) ||c > 0xfe) { /* bad */
337 filter->flag = 1;
338 }
339 filter->status = 0;
340 } else if (c >= 0 && c < 0x80) { /* latin ok */
341 ;
342 } else if (c > c1 && c < 0xff) { /* DBCS lead byte */
343 filter->status = 1;
344 } else { /* bad */
345 filter->flag = 1;
346 }
347
348 return c;
349 }
350
