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 #ifdef HAVE_CONFIG_H
31 #include "config.h"
32 #endif
33
34 #include "mbfilter.h"
35 #include "mbfilter_big5.h"
36
37 #include "unicode_table_big5.h"
38
39 static int mbfl_filt_ident_big5(int c, mbfl_identify_filter *filter);
40
41 static const unsigned char mblen_table_big5[] = { /* 0x81-0xFE */
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, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
47 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
48 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
49 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
50 1, 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, 2,
54 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
55 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
56 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
57 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 1
58 };
59
60 static const char *mbfl_encoding_big5_aliases[] = {"CN-BIG5", "BIG-FIVE", "BIGFIVE", NULL};
61
62 const mbfl_encoding mbfl_encoding_big5 = {
63 mbfl_no_encoding_big5,
64 "BIG-5",
65 "BIG5",
66 (const char *(*)[])&mbfl_encoding_big5_aliases,
67 mblen_table_big5,
68 MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE
69 };
70
71 const mbfl_encoding mbfl_encoding_cp950 = {
72 mbfl_no_encoding_cp950,
73 "CP950",
74 "BIG5",
75 NULL,
76 mblen_table_big5,
77 MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE
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_common_dtor,
84 mbfl_filt_ident_big5
85 };
86
87 const struct mbfl_identify_vtbl vtbl_identify_cp950 = {
88 mbfl_no_encoding_cp950,
89 mbfl_filt_ident_common_ctor,
90 mbfl_filt_ident_common_dtor,
91 mbfl_filt_ident_big5
92 };
93
94 const struct mbfl_convert_vtbl vtbl_big5_wchar = {
95 mbfl_no_encoding_big5,
96 mbfl_no_encoding_wchar,
97 mbfl_filt_conv_common_ctor,
98 mbfl_filt_conv_common_dtor,
99 mbfl_filt_conv_big5_wchar,
100 mbfl_filt_conv_common_flush
101 };
102
103 const struct mbfl_convert_vtbl vtbl_wchar_big5 = {
104 mbfl_no_encoding_wchar,
105 mbfl_no_encoding_big5,
106 mbfl_filt_conv_common_ctor,
107 mbfl_filt_conv_common_dtor,
108 mbfl_filt_conv_wchar_big5,
109 mbfl_filt_conv_common_flush
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 mbfl_filt_conv_common_dtor,
117 mbfl_filt_conv_big5_wchar,
118 mbfl_filt_conv_common_flush
119 };
120
121 const struct mbfl_convert_vtbl vtbl_wchar_cp950 = {
122 mbfl_no_encoding_wchar,
123 mbfl_no_encoding_cp950,
124 mbfl_filt_conv_common_ctor,
125 mbfl_filt_conv_common_dtor,
126 mbfl_filt_conv_wchar_big5,
127 mbfl_filt_conv_common_flush
128 };
129
130 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
131
132 /* 63 + 94 = 157 or 94 */
133 static unsigned short cp950_pua_tbl[][4] = {
134 {0xe000,0xe310,0xfa40,0xfefe},
135 {0xe311,0xeeb7,0x8e40,0xa0fe},
136 {0xeeb8,0xf6b0,0x8140,0x8dfe},
137 {0xf6b1,0xf70e,0xc6a1,0xc6fe},
138 {0xf70f,0xf848,0xc740,0xc8fe},
139 };
140
141 /*
142 * Big5 => wchar
143 */
144 int
mbfl_filt_conv_big5_wchar(int c,mbfl_convert_filter * filter)145 mbfl_filt_conv_big5_wchar(int c, mbfl_convert_filter *filter)
146 {
147 int k;
148 int c1, w, c2;
149
150 switch (filter->status) {
151 case 0:
152 if (filter->from->no_encoding == mbfl_no_encoding_cp950) {
153 c1 = 0x80;
154 } else {
155 c1 = 0xa0;
156 }
157
158 if (c >= 0 && c <= 0x80) { /* latin */
159 CK((*filter->output_function)(c, filter->data));
160 } else if (c == 0xff) {
161 CK((*filter->output_function)(0xf8f8, filter->data));
162 } else if (c > c1 && c < 0xff) { /* dbcs lead byte */
163 filter->status = 1;
164 filter->cache = c;
165 } else {
166 w = c & MBFL_WCSGROUP_MASK;
167 w |= MBFL_WCSGROUP_THROUGH;
168 CK((*filter->output_function)(w, filter->data));
169 }
170 break;
171
172 case 1: /* dbcs second byte */
173 filter->status = 0;
174 c1 = filter->cache;
175 if ((c > 0x39 && c < 0x7f) | (c > 0xa0 && c < 0xff)) {
176 if (c < 0x7f){
177 w = (c1 - 0xa1)*157 + (c - 0x40);
178 } else {
179 w = (c1 - 0xa1)*157 + (c - 0xa1) + 0x3f;
180 }
181 if (w >= 0 && w < big5_ucs_table_size) {
182 w = big5_ucs_table[w];
183 } else {
184 w = 0;
185 }
186
187 if (filter->from->no_encoding == mbfl_no_encoding_cp950) {
188 /* PUA for CP950 */
189 if (w <= 0 &&
190 (((c1 >= 0xfa && c1 <= 0xfe) || (c1 >= 0x8e && c1 <= 0xa0) ||
191 (c1 >= 0x81 && c1 <= 0x8d) ||(c1 >= 0xc7 && c1 <= 0xc8))
192 && ((c > 0x39 && c < 0x7f) || (c > 0xa0 && c < 0xff))) ||
193 ((c1 == 0xc6) && (c > 0xa0 && c < 0xff))) {
194 c2 = c1 << 8 | c;
195 for (k = 0; k < sizeof(cp950_pua_tbl)/(sizeof(unsigned short)*4); k++) {
196 if (c2 >= cp950_pua_tbl[k][2] && c2 <= cp950_pua_tbl[k][3]) {
197 break;
198 }
199 }
200
201 if ((cp950_pua_tbl[k][2] & 0xff) == 0x40) {
202 w = 157*(c1 - (cp950_pua_tbl[k][2]>>8)) + c - (c >= 0xa1 ? 0x62 : 0x40)
203 + cp950_pua_tbl[k][0];
204 } else {
205 w = c2 - cp950_pua_tbl[k][2] + cp950_pua_tbl[k][0];
206 }
207 }
208 }
209
210 if (w <= 0) {
211 w = (c1 << 8) | c;
212 w &= MBFL_WCSPLANE_MASK;
213 w |= MBFL_WCSPLANE_BIG5;
214 }
215 CK((*filter->output_function)(w, filter->data));
216 } else if ((c >= 0 && c < 0x21) || c == 0x7f) { /* CTLs */
217 CK((*filter->output_function)(c, filter->data));
218 } else {
219 w = (c1 << 8) | c;
220 w &= MBFL_WCSGROUP_MASK;
221 w |= MBFL_WCSGROUP_THROUGH;
222 CK((*filter->output_function)(w, filter->data));
223 }
224 break;
225
226 default:
227 filter->status = 0;
228 break;
229 }
230
231 return c;
232 }
233
234 /*
235 * wchar => Big5
236 */
237 int
mbfl_filt_conv_wchar_big5(int c,mbfl_convert_filter * filter)238 mbfl_filt_conv_wchar_big5(int c, mbfl_convert_filter *filter)
239 {
240 int k;
241 int c1, s, c2;
242
243 s = 0;
244 if (c >= ucs_a1_big5_table_min && c < ucs_a1_big5_table_max) {
245 s = ucs_a1_big5_table[c - ucs_a1_big5_table_min];
246 } else if (c >= ucs_a2_big5_table_min && c < ucs_a2_big5_table_max) {
247 s = ucs_a2_big5_table[c - ucs_a2_big5_table_min];
248 } else if (c >= ucs_a3_big5_table_min && c < ucs_a3_big5_table_max) {
249 s = ucs_a3_big5_table[c - ucs_a3_big5_table_min];
250 } else if (c >= ucs_i_big5_table_min && c < ucs_i_big5_table_max) {
251 s = ucs_i_big5_table[c - ucs_i_big5_table_min];
252 } else if (c >= ucs_pua_big5_table_min && c < ucs_pua_big5_table_max) {
253 s = ucs_pua_big5_table[c - ucs_pua_big5_table_min];
254 } else if (c >= ucs_r1_big5_table_min && c < ucs_r1_big5_table_max) {
255 s = ucs_r1_big5_table[c - ucs_r1_big5_table_min];
256 } else if (c >= ucs_r2_big5_table_min && c < ucs_r2_big5_table_max) {
257 s = ucs_r2_big5_table[c - ucs_r2_big5_table_min];
258 }
259
260 if (filter->to->no_encoding == mbfl_no_encoding_cp950) {
261 if (c >= 0xe000 && c <= 0xf848) { /* PUA for CP950 */
262 for (k = 0; k < sizeof(cp950_pua_tbl)/(sizeof(unsigned short)*4); k++) {
263 if (c <= cp950_pua_tbl[k][1]) {
264 break;
265 }
266 }
267 c1 = c - cp950_pua_tbl[k][0];
268 if ((cp950_pua_tbl[k][2] & 0xff) == 0x40) {
269 c2 = cp950_pua_tbl[k][2] >> 8;
270 s = ((c1 / 157) + c2) << 8; c1 %= 157;
271 s |= c1 + (c1 >= 0x3f ? 0x62 : 0x40);
272 } else {
273 s = c1 + cp950_pua_tbl[k][2];
274 }
275 }
276
277 if (c == 0x80) {
278 s = 0x80;
279 } else if (c == 0xf8f8) {
280 s = 0xff;
281 } else if (c == 0x256d) {
282 s = 0xa27e;
283 } else if (c == 0x256e) {
284 s = 0xa2a1;
285 } else if (c == 0x256f) {
286 s = 0xa2a3;
287 } else if (c == 0x2570) {
288 s = 0xa2a2;
289 }
290 }
291
292 if (s <= 0) {
293 c1 = c & ~MBFL_WCSPLANE_MASK;
294 if (c1 == MBFL_WCSPLANE_BIG5) {
295 s = c & MBFL_WCSPLANE_MASK;
296 }
297 if (c == 0) {
298 s = 0;
299 } else if (s <= 0) {
300 s = -1;
301 }
302 }
303 if (s >= 0) {
304 if (s <= 0x80 || s == 0xff) { /* latin */
305 CK((*filter->output_function)(s, filter->data));
306 } else {
307 CK((*filter->output_function)((s >> 8) & 0xff, filter->data));
308 CK((*filter->output_function)(s & 0xff, filter->data));
309 }
310 } else {
311 if (filter->illegal_mode != MBFL_OUTPUTFILTER_ILLEGAL_MODE_NONE) {
312 CK(mbfl_filt_conv_illegal_output(c, filter));
313 }
314 }
315
316 return c;
317 }
318
mbfl_filt_ident_big5(int c,mbfl_identify_filter * filter)319 static int mbfl_filt_ident_big5(int c, mbfl_identify_filter *filter)
320 {
321 int c1;
322 if (filter->encoding->no_encoding == mbfl_no_encoding_cp950) {
323 c1 = 0x80;
324 } else {
325 c1 = 0xa0;
326 }
327
328 if (filter->status) { /* kanji second char */
329 if (c < 0x40 || (c > 0x7e && c < 0xa1) ||c > 0xfe) { /* bad */
330 filter->flag = 1;
331 }
332 filter->status = 0;
333 } else if (c >= 0 && c < 0x80) { /* latin ok */
334 ;
335 } else if (c > c1 && c < 0xff) { /* DBCS lead byte */
336 filter->status = 1;
337 } else { /* bad */
338 filter->flag = 1;
339 }
340
341 return c;
342 }
343
