Home | History | Annotate | Line # | Download | only in modules
citrus_big5.c revision 1.2
      1 /*	$NetBSD: citrus_big5.c,v 1.2 2002/03/27 15:01:17 yamt Exp $	*/
      2 
      3 /*-
      4  * Copyright (c)2002 Citrus Project,
      5  * All rights reserved.
      6  *
      7  * Redistribution and use in source and binary forms, with or without
      8  * modification, are permitted provided that the following conditions
      9  * are met:
     10  * 1. Redistributions of source code must retain the above copyright
     11  *    notice, this list of conditions and the following disclaimer.
     12  * 2. Redistributions in binary form must reproduce the above copyright
     13  *    notice, this list of conditions and the following disclaimer in the
     14  *    documentation and/or other materials provided with the distribution.
     15  *
     16  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
     17  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     18  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     19  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
     20  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     21  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     22  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     23  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     24  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     25  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     26  * SUCH DAMAGE.
     27  */
     28 
     29 /*-
     30  * Copyright (c) 1993
     31  *	The Regents of the University of California.  All rights reserved.
     32  *
     33  * This code is derived from software contributed to Berkeley by
     34  * Paul Borman at Krystal Technologies.
     35  *
     36  * Redistribution and use in source and binary forms, with or without
     37  * modification, are permitted provided that the following conditions
     38  * are met:
     39  * 1. Redistributions of source code must retain the above copyright
     40  *    notice, this list of conditions and the following disclaimer.
     41  * 2. Redistributions in binary form must reproduce the above copyright
     42  *    notice, this list of conditions and the following disclaimer in the
     43  *    documentation and/or other materials provided with the distribution.
     44  * 3. All advertising materials mentioning features or use of this software
     45  *    must display the following acknowledgement:
     46  *	This product includes software developed by the University of
     47  *	California, Berkeley and its contributors.
     48  * 4. Neither the name of the University nor the names of its contributors
     49  *    may be used to endorse or promote products derived from this software
     50  *    without specific prior written permission.
     51  *
     52  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
     53  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
     54  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
     55  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
     56  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
     57  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
     58  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
     59  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
     60  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
     61  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
     62  * SUCH DAMAGE.
     63  */
     64 
     65 #include <sys/cdefs.h>
     66 #if defined(LIBC_SCCS) && !defined(lint)
     67 __RCSID("$NetBSD: citrus_big5.c,v 1.2 2002/03/27 15:01:17 yamt Exp $");
     68 #endif /* LIBC_SCCS and not lint */
     69 
     70 #include <assert.h>
     71 #include <errno.h>
     72 #include <string.h>
     73 #include <stdio.h>
     74 #include <stdlib.h>
     75 #include <stddef.h>
     76 #include <locale.h>
     77 #include <wchar.h>
     78 #include <sys/types.h>
     79 #include <limits.h>
     80 #include "citrus_module.h"
     81 #include "citrus_ctype.h"
     82 #include "citrus_big5.h"
     83 
     84 /* ----------------------------------------------------------------------
     85  * private stuffs used by templates
     86  */
     87 
     88 typedef struct {
     89 	char ch[2];
     90 	int chlen;
     91 } _BIG5State;
     92 
     93 typedef struct {
     94 	int dummy;
     95 } _BIG5EncodingInfo;
     96 
     97 typedef struct {
     98 	_BIG5EncodingInfo	ei;
     99 	struct {
    100 		/* for future multi-locale facility */
    101 		_BIG5State	s_mblen;
    102 		_BIG5State	s_mbrlen;
    103 		_BIG5State	s_mbrtowc;
    104 		_BIG5State	s_mbtowc;
    105 		_BIG5State	s_mbsrtowcs;
    106 		_BIG5State	s_wcrtomb;
    107 		_BIG5State	s_wcsrtombs;
    108 		_BIG5State	s_wcstombs;
    109 		_BIG5State	s_wctomb;
    110 	} states;
    111 } _BIG5CTypeInfo;
    112 
    113 #define _TO_EI(_cl_)			((_BIG5EncodingInfo *)(_cl_))
    114 #define	_TO_CEI(_cl_)			((_BIG5CTypeInfo *)(_cl_))
    115 #define _TO_STATE(_ps_)			((_BIG5State *)(_ps_))
    116 #define _CEI_TO_EI(_cei_)		(&(_cei_)->ei)
    117 #define _CEI_TO_STATE(_cei_, _func_)	(_cei_)->states.s_##_func_
    118 
    119 #define _FUNCNAME(m)			_citrus_BIG5_##m
    120 #define _ENCODING_INFO			_BIG5EncodingInfo
    121 #define _CTYPE_INFO			_BIG5CTypeInfo
    122 #define _ENCODING_STATE			_BIG5State
    123 #define _ENCODING_MB_CUR_MAX(_ei_)	2
    124 #define _ENCODING_IS_STATE_DEPENDENT	0
    125 
    126 
    127 static __inline void
    128 /*ARGSUSED*/
    129 _citrus_BIG5_init_state(_BIG5EncodingInfo * __restrict ei,
    130 			_BIG5State * __restrict s)
    131 {
    132 	memset(s, 0, sizeof(*s));
    133 }
    134 
    135 static __inline void
    136 /*ARGSUSED*/
    137 _citrus_BIG5_pack_state(_BIG5EncodingInfo * __restrict ei,
    138 			void * __restrict pspriv,
    139 			const _BIG5State * __restrict s)
    140 {
    141 	memcpy(pspriv, (const void *)s, sizeof(*s));
    142 }
    143 
    144 static __inline void
    145 /*ARGSUSED*/
    146 _citrus_BIG5_unpack_state(_BIG5EncodingInfo * __restrict ei,
    147 			  _BIG5State * __restrict s,
    148 			  const void * __restrict pspriv)
    149 {
    150 	memcpy((void *)s, pspriv, sizeof(*s));
    151 }
    152 
    153 static __inline int
    154 _citrus_BIG5_check(u_int c)
    155 {
    156 	c &= 0xff;
    157 	return ((c >= 0xa1 && c <= 0xfe) ? 2 : 1);
    158 }
    159 
    160 static __inline int
    161 _citrus_BIG5_check2(u_int c)
    162 {
    163 	c &= 0xff;
    164 	if ((c >= 0x40 && c <= 0x7f) || (c >= 0xa1 && c <= 0xfe))
    165 		return 1;
    166 	else
    167 		return 0;
    168 }
    169 
    170 static int
    171 /*ARGSUSED*/
    172 _citrus_BIG5_stdencoding_init(_BIG5EncodingInfo * __restrict ei,
    173 			      const void * __restrict var, size_t lenvar)
    174 {
    175 	_DIAGASSERT(ei != NULL);
    176 
    177 	memset((void *)ei, 0, sizeof(*ei));
    178 
    179 	return (0);
    180 }
    181 
    182 static void
    183 /*ARGSUSED*/
    184 _citrus_BIG5_stdencoding_uninit(_BIG5EncodingInfo *ei)
    185 {
    186 }
    187 
    188 static int
    189 /*ARGSUSED*/
    190 _citrus_BIG5_mbrtowc_priv(_BIG5EncodingInfo * __restrict ei,
    191 			  wchar_t * __restrict pwc,
    192 			  const char ** __restrict s, size_t n,
    193 			  _BIG5State * __restrict psenc,
    194 			  size_t * __restrict nresult)
    195 {
    196 	wchar_t wchar;
    197 	int c;
    198 	int chlenbak;
    199 	const char *s0;
    200 
    201 	_DIAGASSERT(nresult != 0);
    202 	_DIAGASSERT(ei != NULL);
    203 	_DIAGASSERT(psenc != NULL);
    204 	_DIAGASSERT(s != NULL && *s != NULL);
    205 
    206 	s0 = *s;
    207 
    208 	if (s0 == NULL) {
    209 		_citrus_BIG5_init_state(ei, psenc);
    210 		*nresult = 0;
    211 		return (0);
    212 	}
    213 
    214 	chlenbak = psenc->chlen;
    215 
    216 	/* make sure we have the first byte in the buffer */
    217 	switch (psenc->chlen) {
    218 	case 0:
    219 		if (n < 1)
    220 			goto restart;
    221 		psenc->ch[0] = *s0++;
    222 		psenc->chlen = 1;
    223 		n--;
    224 		break;
    225 	case 1:
    226 		break;
    227 	default:
    228 		/* illegal state */
    229 		goto ilseq;
    230 	}
    231 
    232 	c = _citrus_BIG5_check(psenc->ch[0] & 0xff);
    233 	if (c == 0)
    234 		goto ilseq;
    235 	while (psenc->chlen < c) {
    236 		if (n < 1) {
    237 			goto restart;
    238 		}
    239 		psenc->ch[psenc->chlen] = *s0++;
    240 		psenc->chlen++;
    241 		n--;
    242 	}
    243 
    244 	switch (c) {
    245 	case 1:
    246 		wchar = psenc->ch[0] & 0xff;
    247 		break;
    248 	case 2:
    249 		if (!_citrus_BIG5_check2(psenc->ch[1] & 0xff))
    250 			goto ilseq;
    251 		wchar = ((psenc->ch[0] & 0xff) << 8) | (psenc->ch[1] & 0xff);
    252 		break;
    253 	default:
    254 		/* illegal state */
    255 		goto ilseq;
    256 	}
    257 
    258 	*s = s0;
    259 	psenc->chlen = 0;
    260 	if (pwc)
    261 		*pwc = wchar;
    262 	if (!wchar)
    263 		*nresult = 0;
    264 	else
    265 		*nresult = c - chlenbak;
    266 
    267 	return (0);
    268 
    269 ilseq:
    270 	psenc->chlen = 0;
    271 	*nresult = (size_t)-1;
    272 	return (EILSEQ);
    273 
    274 restart:
    275 	*s = s0;
    276 	*nresult = (size_t)-2;
    277 	return (0);
    278 }
    279 
    280 static int
    281 /*ARGSUSED*/
    282 _citrus_BIG5_wcrtomb_priv(_BIG5EncodingInfo * __restrict ei,
    283 			  char * __restrict s,
    284 			  size_t n, wchar_t wc, _BIG5State * __restrict psenc,
    285 			  size_t * __restrict nresult)
    286 {
    287 	int l;
    288 
    289 	_DIAGASSERT(ei != NULL);
    290 	_DIAGASSERT(nresult != 0);
    291 	_DIAGASSERT(s != NULL);
    292 
    293 	/* check invalid sequence */
    294 	if (wc & ~0xffff)
    295 		goto ilseq;
    296 
    297 	if (wc & 0x8000) {
    298 		if (_citrus_BIG5_check((wc >> 8) & 0xff) != 2 ||
    299 		    !_citrus_BIG5_check2(wc & 0xff))
    300 			goto ilseq;
    301 		l = 2;
    302 	} else {
    303 		if (wc & ~0xff || !_citrus_BIG5_check(wc & 0xff))
    304 			goto ilseq;
    305 		l = 1;
    306 	}
    307 
    308 	if (n < l) {
    309 		/* bound check failure */
    310 		goto ilseq;
    311 	}
    312 
    313 	if (l == 2) {
    314 		s[0] = (wc >> 8) & 0xff;
    315 		s[1] = wc & 0xff;
    316 	} else
    317 		s[0] = wc & 0xff;
    318 
    319 	*nresult = l;
    320 
    321 	return (0);
    322 
    323 ilseq:
    324 	*nresult = (size_t)-1;
    325 	return (EILSEQ);
    326 }
    327 
    328 
    329 /* ----------------------------------------------------------------------
    330  * public interface for ctype
    331  */
    332 
    333 _CITRUS_CTYPE_DECLS(BIG5);
    334 _CITRUS_CTYPE_DEF_OPS(BIG5);
    335 
    336 #include "citrus_ctype_template.h"
    337