Sync Citrus iconv support with NetBSD.
[dragonfly/netmp.git] / lib / libc / citrus / modules / citrus_zw.c
blob2edc314c43b258a21f2bd77d244f0f5927cf2e6b
1 /* $NetBSD: citrus_zw.c,v 1.3 2006/11/24 17:27:52 tnozaki Exp $ */
2 /* $DragonFly: src/lib/libc/citrus/modules/citrus_zw.c,v 1.1 2008/04/10 10:21:02 hasso Exp $ */
4 /*-
5 * Copyright (c)2004, 2006 Citrus Project,
6 * All rights reserved.
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 * notice, this list of conditions and the following disclaimer in the
15 * documentation and/or other materials provided with the distribution.
17 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
18 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
19 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
20 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
21 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
22 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
23 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
24 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
25 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
26 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
27 * SUCH DAMAGE.
31 #include <sys/types.h>
32 #include <assert.h>
33 #include <errno.h>
34 #include <string.h>
35 #include <stdio.h>
36 #include <stdint.h>
37 #include <stdlib.h>
38 #include <stddef.h>
39 #include <locale.h>
40 #include <wchar.h>
41 #include <limits.h>
43 #include "citrus_namespace.h"
44 #include "citrus_types.h"
45 #include "citrus_module.h"
46 #include "citrus_ctype.h"
47 #include "citrus_stdenc.h"
48 #include "citrus_zw.h"
50 /* ----------------------------------------------------------------------
51 * private stuffs used by templates
54 typedef struct {
55 int dummy;
56 } _ZWEncodingInfo;
58 typedef enum {
59 NONE, AMBIGIOUS, ASCII, GB2312
60 } _ZWCharset;
62 typedef struct {
63 int chlen;
64 char ch[4];
65 _ZWCharset charset;
66 } _ZWState;
68 typedef struct {
69 _ZWEncodingInfo ei;
70 struct {
71 /* for future multi-locale facility */
72 _ZWState s_mblen;
73 _ZWState s_mbrlen;
74 _ZWState s_mbrtowc;
75 _ZWState s_mbtowc;
76 _ZWState s_mbsrtowcs;
77 _ZWState s_wcrtomb;
78 _ZWState s_wcsrtombs;
79 _ZWState s_wctomb;
80 } states;
81 } _ZWCTypeInfo;
83 #define _CEI_TO_EI(_cei_) (&(_cei_)->ei)
84 #define _CEI_TO_STATE(_cei_, _func_) (_cei_)->states.s_##_func_
86 #define _FUNCNAME(m) _citrus_ZW_##m
87 #define _ENCODING_INFO _ZWEncodingInfo
88 #define _CTYPE_INFO _ZWCTypeInfo
89 #define _ENCODING_STATE _ZWState
90 #define _ENCODING_MB_CUR_MAX(_ei_) MB_LEN_MAX
91 #define _ENCODING_IS_STATE_DEPENDENT 1
92 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_) ((_ps_)->charset != NONE)
94 static __inline void
95 /*ARGSUSED*/
96 _citrus_ZW_init_state(_ZWEncodingInfo * __restrict ei,
97 _ZWState * __restrict psenc)
99 /* ei my be unused */
100 _DIAGASSERT(psenc != NULL);
102 psenc->chlen = 0;
103 psenc->charset = NONE;
106 static __inline void
107 /*ARGSUSED*/
108 _citrus_ZW_pack_state(_ZWEncodingInfo * __restrict ei,
109 void *__restrict pspriv, const _ZWState * __restrict psenc)
111 /* ei may be unused */
112 _DIAGASSERT(pspriv != NULL);
113 _DIAGASSERT(psenc != NULL);
115 memcpy(pspriv, (const void *)psenc, sizeof(*psenc));
118 static __inline void
119 /*ARGSUSED*/
120 _citrus_ZW_unpack_state(_ZWEncodingInfo * __restrict ei,
121 _ZWState * __restrict psenc, const void * __restrict pspriv)
123 /* ei may be unused */
124 _DIAGASSERT(psenc != NULL);
125 _DIAGASSERT(pspriv != NULL);
127 memcpy((void *)psenc, pspriv, sizeof(*psenc));
130 static int
131 _citrus_ZW_mbrtowc_priv(_ZWEncodingInfo * __restrict ei,
132 wchar_t * __restrict pwc, const char **__restrict s, size_t n,
133 _ZWState * __restrict psenc, size_t * __restrict nresult)
135 const char *s0;
136 int ch, len;
137 wchar_t wc;
139 /* ei may be unused */
140 /* pwc may be null */
141 _DIAGASSERT(s != NULL);
142 _DIAGASSERT(psenc != NULL);
143 _DIAGASSERT(nresult != NULL);
145 if (*s == NULL) {
146 _citrus_ZW_init_state(ei, psenc);
147 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT;
148 return 0;
150 s0 = *s;
151 len = 0;
153 #define STORE \
154 do { \
155 if (n-- < 1) { \
156 *nresult = (size_t)-2; \
157 *s = s0; \
158 return 0; \
160 ch = (unsigned char)*s0++; \
161 if (len++ > MB_LEN_MAX || ch > 0x7F)\
162 goto ilseq; \
163 psenc->ch[psenc->chlen++] = ch; \
164 } while (/*CONSTCOND*/0)
166 loop:
167 switch (psenc->charset) {
168 case ASCII:
169 switch (psenc->chlen) {
170 case 0:
171 STORE;
172 switch (psenc->ch[0]) {
173 case '\0': case '\n':
174 psenc->charset = NONE;
176 /*FALLTHROUGH*/
177 case 1:
178 break;
179 default:
180 return EINVAL;
182 ch = (unsigned char)psenc->ch[0];
183 if (ch > 0x7F)
184 goto ilseq;
185 wc = (wchar_t)ch;
186 psenc->chlen = 0;
187 break;
188 case NONE:
189 if (psenc->chlen != 0)
190 return EINVAL;
191 STORE;
192 ch = (unsigned char)psenc->ch[0];
193 if (ch != 'z') {
194 if (ch != '\n' && ch != '\0')
195 psenc->charset = ASCII;
196 wc = (wchar_t)ch;
197 psenc->chlen = 0;
198 break;
200 psenc->charset = AMBIGIOUS;
201 psenc->chlen = 0;
202 /* FALLTHROUGH */
203 case AMBIGIOUS:
204 if (psenc->chlen != 0)
205 return EINVAL;
206 STORE;
207 if (psenc->ch[0] != 'W') {
208 psenc->charset = ASCII;
209 wc = L'z';
210 break;
212 psenc->charset = GB2312;
213 psenc->chlen = 0;
214 /* FALLTHROUGH */
215 case GB2312:
216 switch (psenc->chlen) {
217 case 0:
218 STORE;
219 ch = (unsigned char)psenc->ch[0];
220 if (ch == '\0') {
221 psenc->charset = NONE;
222 wc = (wchar_t)ch;
223 psenc->chlen = 0;
224 break;
225 } else if (ch == '\n') {
226 psenc->charset = NONE;
227 psenc->chlen = 0;
228 goto loop;
230 /*FALLTHROUGH*/
231 case 1:
232 STORE;
233 if (psenc->ch[0] == ' ') {
234 ch = (unsigned char)psenc->ch[1];
235 wc = (wchar_t)ch;
236 psenc->chlen = 0;
237 break;
238 } else if (psenc->ch[0] == '#') {
239 ch = (unsigned char)psenc->ch[1];
240 if (ch == '\n') {
241 psenc->charset = NONE;
242 wc = (wchar_t)ch;
243 psenc->chlen = 0;
244 break;
245 } else if (ch == ' ') {
246 wc = (wchar_t)ch;
247 psenc->chlen = 0;
248 break;
251 ch = (unsigned char)psenc->ch[0];
252 if (ch < 0x21 || ch > 0x7E)
253 goto ilseq;
254 wc = (wchar_t)(ch << 8);
255 ch = (unsigned char)psenc->ch[1];
256 if (ch < 0x21 || ch > 0x7E) {
257 ilseq:
258 *nresult = (size_t)-1;
259 return EILSEQ;
261 wc |= (wchar_t)ch;
262 psenc->chlen = 0;
263 break;
264 default:
265 return EINVAL;
267 break;
268 default:
269 return EINVAL;
271 if (pwc != NULL)
272 *pwc = wc;
274 *nresult = (size_t)(wc == 0 ? 0 : len);
275 *s = s0;
277 return 0;
280 static int
281 /*ARGSUSED*/
282 _citrus_ZW_wcrtomb_priv(_ZWEncodingInfo * __restrict ei,
283 char *__restrict s, size_t n, wchar_t wc,
284 _ZWState * __restrict psenc, size_t * __restrict nresult)
286 int ch;
288 /* ei may be null */
289 _DIAGASSERT(s != NULL);
290 _DIAGASSERT(psenc != NULL);
291 _DIAGASSERT(nresult != NULL);
293 if (psenc->chlen != 0)
294 return EINVAL;
295 if ((uint32_t)wc <= 0x7F) {
296 ch = (unsigned char)wc;
297 switch (psenc->charset) {
298 case NONE:
299 if (ch == '\0' || ch == '\n') {
300 psenc->ch[psenc->chlen++] = ch;
301 } else {
302 if (n < 4)
303 return E2BIG;
304 n -= 4;
305 psenc->ch[psenc->chlen++] = 'z';
306 psenc->ch[psenc->chlen++] = 'W';
307 psenc->ch[psenc->chlen++] = ' ';
308 psenc->ch[psenc->chlen++] = ch;
309 psenc->charset = GB2312;
311 break;
312 case GB2312:
313 if (n < 2)
314 return E2BIG;
315 n -= 2;
316 if (ch == '\0') {
317 psenc->ch[psenc->chlen++] = '\n';
318 psenc->ch[psenc->chlen++] = '\0';
319 psenc->charset = NONE;
320 } else if (ch == '\n') {
321 psenc->ch[psenc->chlen++] = '#';
322 psenc->ch[psenc->chlen++] = '\n';
323 psenc->charset = NONE;
324 } else {
325 psenc->ch[psenc->chlen++] = ' ';
326 psenc->ch[psenc->chlen++] = ch;
328 break;
329 default:
330 return EINVAL;
332 } else if ((uint32_t)wc <= 0x7E7E) {
333 switch (psenc->charset) {
334 case NONE:
335 if (n < 2)
336 return E2BIG;
337 n -= 2;
338 psenc->ch[psenc->chlen++] = 'z';
339 psenc->ch[psenc->chlen++] = 'W';
340 psenc->charset = GB2312;
341 /* FALLTHROUGH*/
342 case GB2312:
343 if (n < 2)
344 return E2BIG;
345 n -= 2;
346 ch = (wc >> 8) & 0xFF;
347 if (ch < 0x21 || ch > 0x7E)
348 goto ilseq;
349 psenc->ch[psenc->chlen++] = ch;
350 ch = wc & 0xFF;
351 if (ch < 0x21 || ch > 0x7E)
352 goto ilseq;
353 psenc->ch[psenc->chlen++] = ch;
354 break;
355 default:
356 return EINVAL;
358 } else {
359 ilseq:
360 *nresult = (size_t)-1;
361 return EILSEQ;
363 memcpy(s, psenc->ch, psenc->chlen);
364 *nresult = psenc->chlen;
365 psenc->chlen = 0;
367 return 0;
370 static int
371 /*ARGSUSED*/
372 _citrus_ZW_put_state_reset(_ZWEncodingInfo * __restrict ei,
373 char * __restrict s, size_t n,
374 _ZWState * __restrict psenc, size_t * __restrict nresult)
376 /* ei may be unused */
377 _DIAGASSERT(s != NULL);
378 _DIAGASSERT(psenc != NULL);
379 _DIAGASSERT(nresult != NULL);
381 if (psenc->chlen != 0)
382 return EINVAL;
383 switch (psenc->charset) {
384 case GB2312:
385 if (n-- < 1)
386 return E2BIG;
387 psenc->ch[psenc->chlen++] = '\n';
388 psenc->charset = NONE;
389 /*FALLTHROUGH*/
390 case NONE:
391 *nresult = psenc->chlen;
392 if (psenc->chlen > 0) {
393 memcpy(s, psenc->ch, psenc->chlen);
394 psenc->chlen = 0;
396 break;
397 default:
398 return EINVAL;
401 return 0;
404 static __inline int
405 /*ARGSUSED*/
406 _citrus_ZW_stdenc_get_state_desc_generic(_ZWEncodingInfo * __restrict ei,
407 _ZWState * __restrict psenc, int * __restrict rstate)
409 /* ei may be unused */
410 _DIAGASSERT(psenc != NULL);
411 _DIAGASSERT(rstate != NULL);
413 switch (psenc->charset) {
414 case NONE:
415 if (psenc->chlen != 0)
416 return EINVAL;
417 *rstate = _STDENC_SDGEN_INITIAL;
418 break;
419 case AMBIGIOUS:
420 if (psenc->chlen != 0)
421 return EINVAL;
422 *rstate = _STDENC_SDGEN_INCOMPLETE_SHIFT;
423 break;
424 case ASCII:
425 case GB2312:
426 switch (psenc->chlen) {
427 case 0:
428 *rstate = _STDENC_SDGEN_STABLE;
429 break;
430 case 1:
431 *rstate = (psenc->ch[0] == '#')
432 ? _STDENC_SDGEN_INCOMPLETE_SHIFT
433 : _STDENC_SDGEN_INCOMPLETE_CHAR;
434 break;
435 default:
436 return EINVAL;
438 break;
439 default:
440 return EINVAL;
442 return 0;
445 static __inline int
446 /*ARGSUSED*/
447 _citrus_ZW_stdenc_wctocs(_ZWEncodingInfo * __restrict ei,
448 _csid_t * __restrict csid, _index_t * __restrict idx, wchar_t wc)
450 /* ei seems to be unused */
451 _DIAGASSERT(csid != NULL);
452 _DIAGASSERT(idx != NULL);
454 *csid = (_csid_t)(wc <= 0x7FU) ? 0 : 1;
455 *idx = (_index_t)wc;
457 return 0;
460 static __inline int
461 /*ARGSUSED*/
462 _citrus_ZW_stdenc_cstowc(_ZWEncodingInfo * __restrict ei,
463 wchar_t * __restrict wc, _csid_t csid, _index_t idx)
465 /* ei seems to be unused */
466 _DIAGASSERT(wc != NULL);
468 switch (csid) {
469 case 0: case 1:
470 break;
471 default:
472 return EINVAL;
474 *wc = (wchar_t)idx;
476 return 0;
479 static void
480 /*ARGSUSED*/
481 _citrus_ZW_encoding_module_uninit(_ZWEncodingInfo *ei)
485 static int
486 /*ARGSUSED*/
487 _citrus_ZW_encoding_module_init(_ZWEncodingInfo * __restrict ei,
488 const void *__restrict var, size_t lenvar)
490 return 0;
493 /* ----------------------------------------------------------------------
494 * public interface for ctype
497 _CITRUS_CTYPE_DECLS(ZW);
498 _CITRUS_CTYPE_DEF_OPS(ZW);
500 #include "citrus_ctype_template.h"
502 /* ----------------------------------------------------------------------
503 * public interface for stdenc
506 _CITRUS_STDENC_DECLS(ZW);
507 _CITRUS_STDENC_DEF_OPS(ZW);
509 #include "citrus_stdenc_template.h"