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:
22 *
23 */
24 /*
25 * The source code included in this files was separated from mbfilter.c
26 * by moriyoshi koizumi <moriyoshi@php.net> on 4 dec 2002.
27 *
28 */
29
30 #include "mbfilter.h"
31 #include "mbfilter_ucs4.h"
32
33 static size_t mb_ucs4_to_wchar(unsigned char **in, size_t *in_len, uint32_t *buf, size_t bufsize, unsigned int *state);
34 static size_t mb_ucs4be_to_wchar(unsigned char **in, size_t *in_len, uint32_t *buf, size_t bufsize, unsigned int *state);
35 static void mb_wchar_to_ucs4be(uint32_t *in, size_t len, mb_convert_buf *buf, bool end);
36 static size_t mb_ucs4le_to_wchar(unsigned char **in, size_t *in_len, uint32_t *buf, size_t bufsize, unsigned int *state);
37 static void mb_wchar_to_ucs4le(uint32_t *in, size_t len, mb_convert_buf *buf, bool end);
38
39 static const char *mbfl_encoding_ucs4_aliases[] = {"ISO-10646-UCS-4", "UCS4", NULL};
40
41 /* This library historically had encodings called 'byte4be' and 'byte4le'
42 * which were almost identical to UCS-4
43 * Maintain minimal support by aliasing to UCS-4 */
44 static const char *mbfl_encoding_ucs4be_aliases[] = {"byte4be", NULL};
45 static const char *mbfl_encoding_ucs4le_aliases[] = {"byte4le", NULL};
46
47 static int mbfl_filt_conv_ucs4_wchar_flush(mbfl_convert_filter *filter);
48
49 const mbfl_encoding mbfl_encoding_ucs4 = {
50 mbfl_no_encoding_ucs4,
51 "UCS-4",
52 "UCS-4",
53 mbfl_encoding_ucs4_aliases,
54 NULL,
55 MBFL_ENCTYPE_WCS4,
56 &vtbl_ucs4_wchar,
57 &vtbl_wchar_ucs4,
58 mb_ucs4_to_wchar,
59 mb_wchar_to_ucs4be,
60 NULL
61 };
62
63 const mbfl_encoding mbfl_encoding_ucs4be = {
64 mbfl_no_encoding_ucs4be,
65 "UCS-4BE",
66 "UCS-4BE",
67 mbfl_encoding_ucs4be_aliases,
68 NULL,
69 MBFL_ENCTYPE_WCS4,
70 &vtbl_ucs4be_wchar,
71 &vtbl_wchar_ucs4be,
72 mb_ucs4be_to_wchar,
73 mb_wchar_to_ucs4be,
74 NULL
75 };
76
77 const mbfl_encoding mbfl_encoding_ucs4le = {
78 mbfl_no_encoding_ucs4le,
79 "UCS-4LE",
80 "UCS-4LE",
81 mbfl_encoding_ucs4le_aliases,
82 NULL,
83 MBFL_ENCTYPE_WCS4,
84 &vtbl_ucs4le_wchar,
85 &vtbl_wchar_ucs4le,
86 mb_ucs4le_to_wchar,
87 mb_wchar_to_ucs4le,
88 NULL
89 };
90
91 const struct mbfl_convert_vtbl vtbl_ucs4_wchar = {
92 mbfl_no_encoding_ucs4,
93 mbfl_no_encoding_wchar,
94 mbfl_filt_conv_common_ctor,
95 NULL,
96 mbfl_filt_conv_ucs4_wchar,
97 mbfl_filt_conv_ucs4_wchar_flush,
98 NULL,
99 };
100
101 const struct mbfl_convert_vtbl vtbl_wchar_ucs4 = {
102 mbfl_no_encoding_wchar,
103 mbfl_no_encoding_ucs4,
104 mbfl_filt_conv_common_ctor,
105 NULL,
106 mbfl_filt_conv_wchar_ucs4be,
107 mbfl_filt_conv_common_flush,
108 NULL,
109 };
110
111 const struct mbfl_convert_vtbl vtbl_ucs4be_wchar = {
112 mbfl_no_encoding_ucs4be,
113 mbfl_no_encoding_wchar,
114 mbfl_filt_conv_common_ctor,
115 NULL,
116 mbfl_filt_conv_ucs4be_wchar,
117 mbfl_filt_conv_ucs4_wchar_flush,
118 NULL,
119 };
120
121 const struct mbfl_convert_vtbl vtbl_wchar_ucs4be = {
122 mbfl_no_encoding_wchar,
123 mbfl_no_encoding_ucs4be,
124 mbfl_filt_conv_common_ctor,
125 NULL,
126 mbfl_filt_conv_wchar_ucs4be,
127 mbfl_filt_conv_common_flush,
128 NULL,
129 };
130
131 const struct mbfl_convert_vtbl vtbl_ucs4le_wchar = {
132 mbfl_no_encoding_ucs4le,
133 mbfl_no_encoding_wchar,
134 mbfl_filt_conv_common_ctor,
135 NULL,
136 mbfl_filt_conv_ucs4le_wchar,
137 mbfl_filt_conv_ucs4_wchar_flush,
138 NULL,
139 };
140
141 const struct mbfl_convert_vtbl vtbl_wchar_ucs4le = {
142 mbfl_no_encoding_wchar,
143 mbfl_no_encoding_ucs4le,
144 mbfl_filt_conv_common_ctor,
145 NULL,
146 mbfl_filt_conv_wchar_ucs4le,
147 mbfl_filt_conv_common_flush,
148 NULL,
149 };
150
151
152 #define CK(statement) do { if ((statement) < 0) return (-1); } while (0)
153
154 /*
155 * UCS-4 => wchar
156 */
mbfl_filt_conv_ucs4_wchar(int c,mbfl_convert_filter * filter)157 int mbfl_filt_conv_ucs4_wchar(int c, mbfl_convert_filter *filter)
158 {
159 int n, endian;
160
161 endian = filter->status & 0xff00;
162 switch (filter->status & 0xff) {
163 case 0:
164 if (endian) {
165 n = c & 0xff;
166 } else {
167 n = (c & 0xffu) << 24;
168 }
169 filter->cache = n;
170 filter->status++;
171 break;
172 case 1:
173 if (endian) {
174 n = (c & 0xff) << 8;
175 } else {
176 n = (c & 0xff) << 16;
177 }
178 filter->cache |= n;
179 filter->status++;
180 break;
181 case 2:
182 if (endian) {
183 n = (c & 0xff) << 16;
184 } else {
185 n = (c & 0xff) << 8;
186 }
187 filter->cache |= n;
188 filter->status++;
189 break;
190 default:
191 if (endian) {
192 n = (c & 0xffu) << 24;
193 } else {
194 n = c & 0xff;
195 }
196 n |= filter->cache;
197 filter->status &= ~0xff;
198 if ((n & 0xffff) == 0 && ((n >> 16) & 0xffff) == 0xfffe) {
199 if (endian) {
200 filter->status = 0; /* big-endian */
201 } else {
202 filter->status = 0x100; /* little-endian */
203 }
204 } else if (n != 0xfeff) {
205 CK((*filter->output_function)(n, filter->data));
206 }
207 break;
208 }
209
210 return 0;
211 }
212
213 /*
214 * UCS-4BE => wchar
215 */
mbfl_filt_conv_ucs4be_wchar(int c,mbfl_convert_filter * filter)216 int mbfl_filt_conv_ucs4be_wchar(int c, mbfl_convert_filter *filter)
217 {
218 int n;
219
220 if (filter->status == 0) {
221 filter->status = 1;
222 n = (c & 0xffu) << 24;
223 filter->cache = n;
224 } else if (filter->status == 1) {
225 filter->status = 2;
226 n = (c & 0xff) << 16;
227 filter->cache |= n;
228 } else if (filter->status == 2) {
229 filter->status = 3;
230 n = (c & 0xff) << 8;
231 filter->cache |= n;
232 } else {
233 filter->status = 0;
234 n = (c & 0xff) | filter->cache;
235 CK((*filter->output_function)(n, filter->data));
236 }
237 return 0;
238 }
239
240 /*
241 * wchar => UCS-4BE
242 */
mbfl_filt_conv_wchar_ucs4be(int c,mbfl_convert_filter * filter)243 int mbfl_filt_conv_wchar_ucs4be(int c, mbfl_convert_filter *filter)
244 {
245 if (c != MBFL_BAD_INPUT) {
246 CK((*filter->output_function)((c >> 24) & 0xff, filter->data));
247 CK((*filter->output_function)((c >> 16) & 0xff, filter->data));
248 CK((*filter->output_function)((c >> 8) & 0xff, filter->data));
249 CK((*filter->output_function)(c & 0xff, filter->data));
250 } else {
251 CK(mbfl_filt_conv_illegal_output(c, filter));
252 }
253
254 return 0;
255 }
256
257 /*
258 * UCS-4LE => wchar
259 */
mbfl_filt_conv_ucs4le_wchar(int c,mbfl_convert_filter * filter)260 int mbfl_filt_conv_ucs4le_wchar(int c, mbfl_convert_filter *filter)
261 {
262 int n;
263
264 if (filter->status == 0) {
265 filter->status = 1;
266 n = (c & 0xff);
267 filter->cache = n;
268 } else if (filter->status == 1) {
269 filter->status = 2;
270 n = (c & 0xff) << 8;
271 filter->cache |= n;
272 } else if (filter->status == 2) {
273 filter->status = 3;
274 n = (c & 0xff) << 16;
275 filter->cache |= n;
276 } else {
277 filter->status = 0;
278 n = ((c & 0xffu) << 24) | filter->cache;
279 CK((*filter->output_function)(n, filter->data));
280 }
281 return 0;
282 }
283
284 /*
285 * wchar => UCS-4LE
286 */
mbfl_filt_conv_wchar_ucs4le(int c,mbfl_convert_filter * filter)287 int mbfl_filt_conv_wchar_ucs4le(int c, mbfl_convert_filter *filter)
288 {
289 if (c != MBFL_BAD_INPUT) {
290 CK((*filter->output_function)(c & 0xff, filter->data));
291 CK((*filter->output_function)((c >> 8) & 0xff, filter->data));
292 CK((*filter->output_function)((c >> 16) & 0xff, filter->data));
293 CK((*filter->output_function)((c >> 24) & 0xff, filter->data));
294 } else {
295 CK(mbfl_filt_conv_illegal_output(c, filter));
296 }
297
298 return 0;
299 }
300
mbfl_filt_conv_ucs4_wchar_flush(mbfl_convert_filter * filter)301 static int mbfl_filt_conv_ucs4_wchar_flush(mbfl_convert_filter *filter)
302 {
303 if (filter->status & 0xF) {
304 /* Input string was truncated */
305 CK((*filter->output_function)(MBFL_BAD_INPUT, filter->data));
306 }
307 filter->status = 0;
308
309 if (filter->flush_function) {
310 (*filter->flush_function)(filter->data);
311 }
312
313 return 0;
314 }
315
316 #define DETECTED_BE 1
317 #define DETECTED_LE 2
318
mb_ucs4_to_wchar(unsigned char ** in,size_t * in_len,uint32_t * buf,size_t bufsize,unsigned int * state)319 static size_t mb_ucs4_to_wchar(unsigned char **in, size_t *in_len, uint32_t *buf, size_t bufsize, unsigned int *state)
320 {
321 if (*state == DETECTED_BE) {
322 return mb_ucs4be_to_wchar(in, in_len, buf, bufsize, NULL);
323 } else if (*state == DETECTED_LE) {
324 return mb_ucs4le_to_wchar(in, in_len, buf, bufsize, NULL);
325 } else if (*in_len >= 4) {
326 unsigned char *p = *in;
327 uint32_t c1 = *p++;
328 uint32_t c2 = *p++;
329 uint32_t c3 = *p++;
330 uint32_t c4 = *p++;
331 uint32_t w = (c1 << 24) | (c2 << 16) | (c3 << 8) | c4;
332
333 if (w == 0xFFFE0000) {
334 /* Little-endian BOM */
335 *in = p;
336 *in_len -= 4;
337 *state = DETECTED_LE;
338 return mb_ucs4le_to_wchar(in, in_len, buf, bufsize, NULL);
339 } else if (w == 0xFEFF) {
340 /* Big-endian BOM; don't send it to output */
341 *in = p;
342 *in_len -= 4;
343 }
344 }
345
346 *state = DETECTED_BE;
347 return mb_ucs4be_to_wchar(in, in_len, buf, bufsize, NULL);
348 }
349
mb_ucs4be_to_wchar(unsigned char ** in,size_t * in_len,uint32_t * buf,size_t bufsize,unsigned int * state)350 static size_t mb_ucs4be_to_wchar(unsigned char **in, size_t *in_len, uint32_t *buf, size_t bufsize, unsigned int *state)
351 {
352 unsigned char *p = *in, *e = p + (*in_len & ~3);
353 uint32_t *out = buf, *limit = buf + bufsize;
354
355 while (p < e && out < limit) {
356 uint32_t c1 = *p++;
357 uint32_t c2 = *p++;
358 uint32_t c3 = *p++;
359 uint32_t c4 = *p++;
360 uint32_t w = (c1 << 24) | (c2 << 16) | (c3 << 8) | c4;
361 *out++ = w;
362 }
363
364 if (p == e && (*in_len & 0x3) && out < limit) {
365 /* There are 1-3 trailing bytes, which shouldn't be there */
366 *out++ = MBFL_BAD_INPUT;
367 p = *in + *in_len;
368 }
369
370 *in_len -= (p - *in);
371 *in = p;
372 return out - buf;
373 }
374
mb_wchar_to_ucs4be(uint32_t * in,size_t len,mb_convert_buf * buf,bool end)375 static void mb_wchar_to_ucs4be(uint32_t *in, size_t len, mb_convert_buf *buf, bool end)
376 {
377 unsigned char *out, *limit;
378 MB_CONVERT_BUF_LOAD(buf, out, limit);
379 MB_CONVERT_BUF_ENSURE(buf, out, limit, len * 4);
380
381 while (len--) {
382 uint32_t w = *in++;
383 if (w != MBFL_BAD_INPUT) {
384 out = mb_convert_buf_add4(out, (w >> 24) & 0xFF, (w >> 16) & 0xFF, (w >> 8) & 0xFF, w & 0xFF);
385 } else {
386 MB_CONVERT_ERROR(buf, out, limit, w, mb_wchar_to_ucs4be);
387 MB_CONVERT_BUF_ENSURE(buf, out, limit, len * 4);
388 }
389 }
390
391 MB_CONVERT_BUF_STORE(buf, out, limit);
392 }
393
mb_ucs4le_to_wchar(unsigned char ** in,size_t * in_len,uint32_t * buf,size_t bufsize,unsigned int * state)394 static size_t mb_ucs4le_to_wchar(unsigned char **in, size_t *in_len, uint32_t *buf, size_t bufsize, unsigned int *state)
395 {
396 unsigned char *p = *in, *e = p + (*in_len & ~3);
397 uint32_t *out = buf, *limit = buf + bufsize;
398
399 while (p < e && out < limit) {
400 uint32_t c1 = *p++;
401 uint32_t c2 = *p++;
402 uint32_t c3 = *p++;
403 uint32_t c4 = *p++;
404 uint32_t w = (c4 << 24) | (c3 << 16) | (c2 << 8) | c1;
405 *out++ = w;
406 }
407
408 if (p == e && (*in_len & 0x3) && out < limit) {
409 /* There are 1-3 trailing bytes, which shouldn't be there */
410 *out++ = MBFL_BAD_INPUT;
411 p = *in + *in_len;
412 }
413
414 *in_len -= (p - *in);
415 *in = p;
416 return out - buf;
417 }
418
mb_wchar_to_ucs4le(uint32_t * in,size_t len,mb_convert_buf * buf,bool end)419 static void mb_wchar_to_ucs4le(uint32_t *in, size_t len, mb_convert_buf *buf, bool end)
420 {
421 unsigned char *out, *limit;
422 MB_CONVERT_BUF_LOAD(buf, out, limit);
423 MB_CONVERT_BUF_ENSURE(buf, out, limit, len * 4);
424
425 while (len--) {
426 uint32_t w = *in++;
427 if (w != MBFL_BAD_INPUT) {
428 out = mb_convert_buf_add4(out, w & 0xFF, (w >> 8) & 0xFF, (w >> 16) & 0xFF, (w >> 24) & 0xFF);
429 } else {
430 MB_CONVERT_ERROR(buf, out, limit, w, mb_wchar_to_ucs4le);
431 MB_CONVERT_BUF_ENSURE(buf, out, limit, len * 4);
432 }
433 }
434
435 MB_CONVERT_BUF_STORE(buf, out, limit);
436 }
437
