1 /*
2 * Copyright (c) 2003-2004, Artem B. Bityuckiy
3 * Copyright (c) 1999,2000, Konstantin Chuguev. All rights reserved.
4 *
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
7 * are met:
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
13 *
14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
15 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
16 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
17 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
18 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
19 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
20 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
21 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
22 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
23 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
24 * SUCH DAMAGE.
25 */
26 #include "cesbi.h"
27
28 #if defined (ICONV_TO_UCS_CES_EUC) \
29 || defined (ICONV_FROM_UCS_CES_EUC)
30
31 #include <_ansi.h>
32 #include <newlib.h>
33 #include <string.h>
34 #include <stdlib.h>
35 #include <limits.h>
36 #include <sys/types.h>
37 #include "../lib/local.h"
38 #include "../lib/ucsconv.h"
39 #include "../lib/encnames.h"
40 #include "../ccs/ccsnames.h"
41
42 #define TYPE_EUC_JP 0
43 #define TYPE_EUC_KR 1
44 #define TYPE_EUC_TW 2
45
46 #define MAX_CS_NUM 3
47
48 /* CS description structure */
49 typedef struct
50 {
51 char *csname;
52 char *prefix;
53 int bytes;
54 int prefixbytes;
55 int touchmsb; /* If 1, msb will be set by euc converter */
56 } euc_cs_desc_t;
57
58 typedef struct
59 {
60 int type;
61 int mb_cur_max;
62 euc_cs_desc_t *desc;
63
64 void *data[MAX_CS_NUM];
65 } euc_data_t;
66
67 #if defined (_ICONV_TO_ENCODING_EUC_JP) \
68 || defined (_ICONV_FROM_ENCODING_EUC_JP) \
69 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
70 static euc_cs_desc_t euc_jp_cs_desc[] =
71 {
72 {ICONV_CCS_JIS_X0208_1990, "", 2, 0, 1},
73 {ICONV_CCS_JIS_X0201_1976, "\x8e", 1, 1, 0},
74 {ICONV_CCS_JIS_X0212_1990, "\x8f", 2, 1, 1},
75 {0}
76 };
77 #endif
78
79 #if defined (_ICONV_TO_ENCODING_EUC_TW) \
80 || defined (_ICONV_FROM_ENCODING_EUC_TW) \
81 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
82 static euc_cs_desc_t euc_tw_cs_desc [] =
83 {
84 {ICONV_CCS_CNS11643_PLANE1, "", 2, 0, 1},
85 {ICONV_CCS_CNS11643_PLANE2, "\x8e\xa2", 2, 2, 1},
86 {ICONV_CCS_CNS11643_PLANE14, "\x8e\xae", 2, 2, 1},
87 {0}
88 };
89 #endif
90
91 #if defined (_ICONV_TO_ENCODING_EUC_KR) \
92 || defined (_ICONV_FROM_ENCODING_EUC_KR) \
93 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
94 static euc_cs_desc_t euc_kr_cs_desc [] =
95 {
96 {ICONV_CCS_KSX1001, "", 2, 0, 1},
97 {0}
98 };
99 #endif
100
101 #if defined (ICONV_FROM_UCS_CES_EUC)
102 static void *
euc_from_ucs_init(const char * encoding)103 euc_from_ucs_init (
104 const char *encoding)
105 {
106 int i;
107 euc_data_t *data;
108
109 if ((data = (euc_data_t *)calloc (1, sizeof (euc_data_t))) == NULL)
110 return 0;
111
112 #if defined (_ICONV_TO_ENCODING_EUC_JP) \
113 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
114 if (strcmp (encoding, ICONV_ENCODING_EUC_JP) == 0)
115 {
116 data->type = TYPE_EUC_JP;
117 data->mb_cur_max = 3;
118 data->desc = &euc_jp_cs_desc[0];
119 goto ok;
120 }
121 #endif
122 #if defined (_ICONV_TO_ENCODING_EUC_KR) \
123 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
124 if (strcmp (encoding, ICONV_ENCODING_EUC_KR) == 0)
125 {
126 data->type = TYPE_EUC_KR;
127 data->mb_cur_max = 2;
128 data->desc = &euc_kr_cs_desc[0];
129 goto ok;
130 }
131 #endif
132 #if defined (_ICONV_TO_ENCODING_EUC_TW) \
133 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
134 if (strcmp (encoding, ICONV_ENCODING_EUC_TW) == 0)
135 {
136 data->type = TYPE_EUC_TW;
137 data->mb_cur_max = 4;
138 data->desc = &euc_tw_cs_desc[0];
139 goto ok;
140 }
141 #endif
142
143 goto error1;
144
145 ok:
146 for (i = 0; data->desc[i].csname != NULL; i++)
147 {
148 data->data[i] = _iconv_from_ucs_ces_handlers_table.init (
149 data->desc[i].csname);
150 if (data->data[i] == NULL)
151 goto error;
152 }
153
154 return data;
155
156 error:
157 _iconv_from_ucs_ces_handlers_table.close (data);
158 return NULL;
159 error1:
160 free ((void *)data);
161 return NULL;
162 }
163
164 static size_t
euc_from_ucs_close(void * data)165 euc_from_ucs_close (
166 void *data)
167 {
168 int i;
169 size_t res = 0;
170
171 for (i = 0; i < MAX_CS_NUM; i++)
172 {
173 if (((euc_data_t *)data)->data[i] != NULL)
174 res |= _iconv_from_ucs_ces_handlers_table.close (
175 ((euc_data_t *)data)->data[i]);
176 }
177 free(data);
178
179 return res;
180 }
181
182 static size_t
euc_convert_from_ucs(void * data,register ucs4_t in,unsigned char ** outbuf,size_t * outbytesleft)183 euc_convert_from_ucs (void *data,
184 register ucs4_t in,
185 unsigned char **outbuf,
186 size_t *outbytesleft)
187 {
188 int i;
189 int j;
190 int res;
191 unsigned char *outbuf1;
192 size_t outbytesleft1;
193 euc_data_t *d = (euc_data_t *)data;
194
195 if (in < 0x80) /* CS0 ASCII */
196 return _iconv_from_ucs_ces_handlers_us_ascii.convert_from_ucs (
197 NULL,
198 in,
199 outbuf,
200 outbytesleft);
201
202 /* Try other CS */
203 for (i = 0; d->desc[i].csname != NULL; i++)
204 {
205
206 if (((int)*outbytesleft - d->desc[i].prefixbytes - d->desc[i].bytes) < 0)
207 {
208 unsigned char buf[ICONV_MB_LEN_MAX];
209 outbytesleft1 = ICONV_MB_LEN_MAX;
210 outbuf1 = &buf[0];
211 /* See wether this is right sequence */
212 res =
213 (int)_iconv_from_ucs_ces_handlers_table.convert_from_ucs (
214 d->data[i],
215 in,
216 &outbuf1,
217 &outbytesleft1);
218 if (res > 0)
219 return (size_t)ICONV_CES_NOSPACE;
220
221 continue;
222 }
223
224 outbuf1 = *outbuf + d->desc[i].prefixbytes;
225 outbytesleft1 = *outbytesleft - d->desc[i].prefixbytes;
226
227 res = (int)_iconv_from_ucs_ces_handlers_table.convert_from_ucs (
228 d->data[i],
229 in,
230 &outbuf1,
231 &outbytesleft1);
232 if (res == d->desc[i].bytes)
233 {
234 for (j = 0; j < d->desc[i].prefixbytes; j++)
235 (*outbuf)[j] = d->desc[i].prefix[j];
236
237 if (d->desc[i].touchmsb)
238 for (j = 0; j < d->desc[i].bytes; j++)
239 {
240 if ((*outbuf)[j + d->desc[i].prefixbytes] & 0x80)
241 return (size_t)ICONV_CES_INVALID_CHARACTER;
242 (*outbuf)[j + d->desc[i].prefixbytes] |= 0x80;
243 }
244
245 *outbuf = outbuf1;
246 *outbytesleft = outbytesleft1;
247
248 return (size_t)(res + d->desc[i].bytes);
249 }
250 }
251
252 return (size_t)ICONV_CES_INVALID_CHARACTER;
253 }
254 #endif /* ICONV_FROM_UCS_CES_EUC */
255
256 #if defined (ICONV_TO_UCS_CES_EUC)
257 static void *
euc_to_ucs_init(const char * encoding)258 euc_to_ucs_init (
259 const char *encoding)
260 {
261 int i;
262 euc_data_t *data;
263
264 if ((data = (euc_data_t *)calloc (1, sizeof (euc_data_t))) == NULL)
265 return 0;
266
267 #if defined (_ICONV_TO_ENCODING_EUC_JP) \
268 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
269 if (strcmp (encoding, ICONV_ENCODING_EUC_JP) == 0)
270 {
271 data->type = TYPE_EUC_JP;
272 data->mb_cur_max = 3;
273 data->desc = &euc_jp_cs_desc[0];
274 goto ok;
275 }
276 #endif
277 #if defined (_ICONV_TO_ENCODING_EUC_KR) \
278 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
279 if (strcmp (encoding, ICONV_ENCODING_EUC_KR) == 0)
280 {
281 data->type = TYPE_EUC_KR;
282 data->mb_cur_max = 2;
283 data->desc = &euc_kr_cs_desc[0];
284 goto ok;
285 }
286 #endif
287 #if defined (_ICONV_TO_ENCODING_EUC_TW) \
288 || defined (_ICONV_ENABLE_EXTERNAL_CCS)
289 if (strcmp (encoding, ICONV_ENCODING_EUC_TW) == 0)
290 {
291 data->type = TYPE_EUC_TW;
292 data->mb_cur_max = 4;
293 data->desc = &euc_tw_cs_desc[0];
294 goto ok;
295 }
296 #endif
297
298 goto error1;
299
300 ok:
301 for (i = 0; data->desc[i].csname != NULL; i++)
302 {
303 data->data[i] = _iconv_to_ucs_ces_handlers_table.init (
304 data->desc[i].csname);
305 if (data->data[i] == NULL)
306 goto error;
307 }
308
309 return data;
310
311 error:
312 _iconv_to_ucs_ces_handlers_table.close (data);
313 return NULL;
314 error1:
315 free ((void *)data);
316 return NULL;
317 }
318
319 static size_t
euc_to_ucs_close(void * data)320 euc_to_ucs_close (
321 void *data)
322 {
323 int i;
324 size_t res = 0;
325
326 for (i = 0; i < MAX_CS_NUM; i++)
327 {
328 if (((euc_data_t *)data)->data[i] != NULL)
329 res |= _iconv_to_ucs_ces_handlers_table.close (
330 ((euc_data_t *)data)->data[i]);
331 }
332 free(data);
333
334 return res;
335 }
336
337 static ucs4_t
euc_convert_to_ucs(void * data,const unsigned char ** inbuf,size_t * inbytesleft)338 euc_convert_to_ucs (void *data,
339 const unsigned char **inbuf,
340 size_t *inbytesleft)
341 {
342 int i;
343 int j;
344 ucs4_t res;
345 unsigned char buf[ICONV_MB_LEN_MAX];
346 size_t inbytesleft1;
347 euc_data_t *d = (euc_data_t *)data;
348 unsigned char *inbuf1 = &buf[0];
349
350 if (**inbuf < 0x80) /* CS0 is always ASCII */
351 return _iconv_to_ucs_ces_handlers_us_ascii.convert_to_ucs (
352 NULL,
353 inbuf,
354 inbytesleft);
355
356 for (i = 1; d->desc[i].csname != NULL; i++)
357 {
358 if (memcmp((const void *)(*inbuf),
359 (const void *)d->desc[i].prefix,
360 d->desc[i].prefixbytes) == 0)
361 {
362 if (((int)*inbytesleft - d->desc[i].prefixbytes - d->desc[i].bytes) < 0)
363 return (ucs4_t)ICONV_CES_BAD_SEQUENCE;
364
365 if (d->desc[i].touchmsb)
366 for (j = 0; j < d->desc[i].bytes; j++)
367 {
368 if (!((*inbuf)[j + d->desc[i].prefixbytes] & 0x80))
369 return (ucs4_t)ICONV_CES_INVALID_CHARACTER;
370 inbuf1[j] = (*inbuf)[j + d->desc[i].prefixbytes] & 0x7F;
371 }
372 else
373 for (j = 0; j < d->desc[i].bytes; j++)
374 inbuf1[j] = (*inbuf)[j + d->desc[i].prefixbytes];
375
376 inbytesleft1 = d->desc[i].bytes;
377
378 res = _iconv_to_ucs_ces_handlers_table.convert_to_ucs (
379 d->data[i],
380 (const unsigned char **)&inbuf1,
381 &inbytesleft1);
382 if (((__int32_t)res) > 0)
383 {
384 *inbuf += d->desc[i].bytes + d->desc[i].prefixbytes;
385 *inbytesleft -= d->desc[i].bytes + d->desc[i].prefixbytes;
386 }
387
388 return res;
389 }
390 }
391
392 /* Process CS1 */
393 if (((int)(*inbytesleft - d->desc[0].prefixbytes - d->desc[0].bytes)) < 0)
394 return (ucs4_t)ICONV_CES_BAD_SEQUENCE;
395
396 if (d->desc[0].touchmsb)
397 for (j = 0; j < d->desc[0].bytes; j++)
398 {
399 if (!((*inbuf)[j + d->desc[0].prefixbytes] & 0x80))
400 return (ucs4_t)ICONV_CES_INVALID_CHARACTER;
401 inbuf1[j] = (*inbuf)[j] & 0x7F;
402 }
403 else
404 for (j = 0; j < d->desc[0].bytes; j++)
405 inbuf1[j] = (*inbuf)[j];
406
407 inbytesleft1 = d->desc[0].bytes;
408
409 res = _iconv_to_ucs_ces_handlers_table.convert_to_ucs (
410 d->data[0],
411 (const unsigned char **)&inbuf1,
412 &inbytesleft1);
413 if (((__int32_t)res) > 0)
414 {
415 *inbuf += d->desc[0].bytes;
416 *inbytesleft -= d->desc[0].bytes;
417 }
418
419 return res;
420 }
421 #endif /* ICONV_TO_UCS_CES_EUC */
422
423 static int
euc_get_mb_cur_max(void * data)424 euc_get_mb_cur_max (void *data)
425 {
426 return ((euc_data_t *)data)->mb_cur_max;
427 }
428
429 #if defined (ICONV_FROM_UCS_CES_EUC)
430 const iconv_from_ucs_ces_handlers_t
431 _iconv_from_ucs_ces_handlers_euc =
432 {
433 euc_from_ucs_init,
434 euc_from_ucs_close,
435 euc_get_mb_cur_max,
436 NULL,
437 NULL,
438 NULL,
439 euc_convert_from_ucs
440 };
441 #endif
442
443 #if defined (ICONV_TO_UCS_CES_EUC)
444 const iconv_to_ucs_ces_handlers_t
445 _iconv_to_ucs_ces_handlers_euc =
446 {
447 euc_to_ucs_init,
448 euc_to_ucs_close,
449 euc_get_mb_cur_max,
450 NULL,
451 NULL,
452 NULL,
453 euc_convert_to_ucs
454 };
455 #endif
456
457 #endif /* ICONV_TO_UCS_CES_EUC || ICONV_FROM_UCS_CES_EUC */
458
459
460
