big5.c revision 128004
138333Sphk/*-
2128004Stjr * Copyright (c) 2002-2004 Tim J. Robbins. All rights reserved.
338333Sphk * Copyright (c) 1993
438333Sphk *	The Regents of the University of California.  All rights reserved.
538333Sphk *
638333Sphk * This code is derived from software contributed to Berkeley by
738333Sphk * Paul Borman at Krystal Technologies.
838333Sphk *
938333Sphk * Redistribution and use in source and binary forms, with or without
1038333Sphk * modification, are permitted provided that the following conditions
1138333Sphk * are met:
1238333Sphk * 1. Redistributions of source code must retain the above copyright
1338333Sphk *    notice, this list of conditions and the following disclaimer.
1438333Sphk * 2. Redistributions in binary form must reproduce the above copyright
1538333Sphk *    notice, this list of conditions and the following disclaimer in the
1638333Sphk *    documentation and/or other materials provided with the distribution.
1738333Sphk * 3. All advertising materials mentioning features or use of this software
1838333Sphk *    must display the following acknowledgement:
1938333Sphk *	This product includes software developed by the University of
2038333Sphk *	California, Berkeley and its contributors.
2138333Sphk * 4. Neither the name of the University nor the names of its contributors
2238333Sphk *    may be used to endorse or promote products derived from this software
2338333Sphk *    without specific prior written permission.
2438333Sphk *
2538333Sphk * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
2638333Sphk * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
2738333Sphk * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
2838333Sphk * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
2938333Sphk * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
3038333Sphk * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
3138333Sphk * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
3238333Sphk * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
3338333Sphk * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
3438333Sphk * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
3538333Sphk * SUCH DAMAGE.
3638333Sphk */
3738333Sphk
3838333Sphk#if defined(LIBC_SCCS) && !defined(lint)
3938333Sphkstatic char sccsid[] = "@(#)big5.c	8.1 (Berkeley) 6/4/93";
4038333Sphk#endif /* LIBC_SCCS and not lint */
41128004Stjr#include <sys/param.h>
4292986Sobrien__FBSDID("$FreeBSD: head/lib/libc/locale/big5.c 128004 2004-04-07 10:48:19Z tjr $");
4338333Sphk
44121893Stjr#include <runetype.h>
4538333Sphk#include <stdlib.h>
46128004Stjr#include <string.h>
47121893Stjr#include <wchar.h>
4838333Sphk
49121893Stjrextern size_t (*__mbrtowc)(wchar_t * __restrict, const char * __restrict,
50121893Stjr    size_t, mbstate_t * __restrict);
51128004Stjrextern int (*__mbsinit)(const mbstate_t *);
52121893Stjrextern size_t (*__wcrtomb)(char * __restrict, wchar_t, mbstate_t * __restrict);
5338333Sphk
54121893Stjrint	_BIG5_init(_RuneLocale *);
55121893Stjrsize_t	_BIG5_mbrtowc(wchar_t * __restrict, const char * __restrict, size_t,
56121893Stjr	    mbstate_t * __restrict);
57128004Stjrint	_BIG5_mbsinit(const mbstate_t *);
58121893Stjrsize_t	_BIG5_wcrtomb(char * __restrict, wchar_t, mbstate_t * __restrict);
59121893Stjr
60128004Stjrtypedef struct {
61128004Stjr	int	count;
62128004Stjr	u_char	bytes[2];
63128004Stjr} _BIG5State;
64128004Stjr
6538333Sphkint
66121893Stjr_BIG5_init(_RuneLocale *rl)
6738333Sphk{
68121893Stjr
69121893Stjr	__mbrtowc = _BIG5_mbrtowc;
70121893Stjr	__wcrtomb = _BIG5_wcrtomb;
71128004Stjr	__mbsinit = _BIG5_mbsinit;
7238333Sphk	_CurrentRuneLocale = rl;
7338333Sphk	__mb_cur_max = 2;
7438333Sphk	return (0);
7538333Sphk}
7638333Sphk
77128004Stjrint
78128004Stjr_BIG5_mbsinit(const mbstate_t *ps)
79128004Stjr{
80128004Stjr
81128004Stjr	return (ps == NULL || ((_BIG5State *)ps)->count == 0);
82128004Stjr}
83128004Stjr
84121893Stjrstatic __inline int
85121893Stjr_big5_check(u_int c)
8638333Sphk{
87121893Stjr
8838333Sphk	c &= 0xff;
8938333Sphk	return ((c >= 0xa1 && c <= 0xfe) ? 2 : 1);
9038333Sphk}
9138333Sphk
92121893Stjrsize_t
93121893Stjr_BIG5_mbrtowc(wchar_t * __restrict pwc, const char * __restrict s, size_t n,
94128004Stjr    mbstate_t * __restrict ps)
9538333Sphk{
96128004Stjr	_BIG5State *bs;
97121893Stjr	wchar_t wc;
98128004Stjr	int i, len, ocount;
99128004Stjr	size_t ncopy;
10038333Sphk
101128004Stjr	bs = (_BIG5State *)ps;
102128004Stjr
103128004Stjr	if (s == NULL) {
104128004Stjr		s = "";
105128004Stjr		n = 1;
106128004Stjr		pwc = NULL;
107128004Stjr	}
108128004Stjr
109128004Stjr	ncopy = MIN(MIN(n, MB_CUR_MAX), sizeof(bs->bytes) - bs->count);
110128004Stjr	memcpy(bs->bytes + bs->count, s, ncopy);
111128004Stjr	ocount = bs->count;
112128004Stjr	bs->count += ncopy;
113128004Stjr	s = (char *)bs->bytes;
114128004Stjr	n = bs->count;
115128004Stjr
116121893Stjr	if (n == 0 || (size_t)(len = _big5_check(*s)) > n)
117121893Stjr		/* Incomplete multibyte sequence */
118121893Stjr		return ((size_t)-2);
119121893Stjr	wc = 0;
120121893Stjr	i = len;
121121893Stjr	while (i-- > 0)
122121893Stjr		wc = (wc << 8) | (unsigned char)*s++;
123121893Stjr	if (pwc != NULL)
124121893Stjr		*pwc = wc;
125128004Stjr	bs->count = 0;
126128004Stjr	return (wc == L'\0' ? 0 : len - ocount);
12738333Sphk}
12838333Sphk
129121893Stjrsize_t
130121893Stjr_BIG5_wcrtomb(char * __restrict s, wchar_t wc,
131121893Stjr    mbstate_t * __restrict ps __unused)
13238333Sphk{
133121893Stjr
134121893Stjr	if (s == NULL)
135121893Stjr		/* Reset to initial shift state (no-op) */
136121893Stjr		return (1);
137121893Stjr	if (wc & 0x8000) {
138121893Stjr		*s++ = (wc >> 8) & 0xff;
139121893Stjr		*s = wc & 0xff;
140121893Stjr		return (2);
14138333Sphk	}
142121893Stjr	*s = wc & 0xff;
143121893Stjr	return (1);
14438333Sphk}
145