big5.c revision 128155
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 128155 2004-04-12 13:09:18Z tjr $");
4338333Sphk
44128155Stjr#include <errno.h>
45121893Stjr#include <runetype.h>
4638333Sphk#include <stdlib.h>
47128004Stjr#include <string.h>
48121893Stjr#include <wchar.h>
4938333Sphk
50121893Stjrextern size_t (*__mbrtowc)(wchar_t * __restrict, const char * __restrict,
51121893Stjr    size_t, mbstate_t * __restrict);
52128004Stjrextern int (*__mbsinit)(const mbstate_t *);
53121893Stjrextern size_t (*__wcrtomb)(char * __restrict, wchar_t, mbstate_t * __restrict);
5438333Sphk
55121893Stjrint	_BIG5_init(_RuneLocale *);
56121893Stjrsize_t	_BIG5_mbrtowc(wchar_t * __restrict, const char * __restrict, size_t,
57121893Stjr	    mbstate_t * __restrict);
58128004Stjrint	_BIG5_mbsinit(const mbstate_t *);
59121893Stjrsize_t	_BIG5_wcrtomb(char * __restrict, wchar_t, mbstate_t * __restrict);
60121893Stjr
61128004Stjrtypedef struct {
62128004Stjr	int	count;
63128004Stjr	u_char	bytes[2];
64128004Stjr} _BIG5State;
65128004Stjr
6638333Sphkint
67121893Stjr_BIG5_init(_RuneLocale *rl)
6838333Sphk{
69121893Stjr
70121893Stjr	__mbrtowc = _BIG5_mbrtowc;
71121893Stjr	__wcrtomb = _BIG5_wcrtomb;
72128004Stjr	__mbsinit = _BIG5_mbsinit;
7338333Sphk	_CurrentRuneLocale = rl;
7438333Sphk	__mb_cur_max = 2;
7538333Sphk	return (0);
7638333Sphk}
7738333Sphk
78128004Stjrint
79128004Stjr_BIG5_mbsinit(const mbstate_t *ps)
80128004Stjr{
81128004Stjr
82128081Stjr	return (ps == NULL || ((const _BIG5State *)ps)->count == 0);
83128004Stjr}
84128004Stjr
85121893Stjrstatic __inline int
86121893Stjr_big5_check(u_int c)
8738333Sphk{
88121893Stjr
8938333Sphk	c &= 0xff;
9038333Sphk	return ((c >= 0xa1 && c <= 0xfe) ? 2 : 1);
9138333Sphk}
9238333Sphk
93121893Stjrsize_t
94121893Stjr_BIG5_mbrtowc(wchar_t * __restrict pwc, const char * __restrict s, size_t n,
95128004Stjr    mbstate_t * __restrict ps)
9638333Sphk{
97128004Stjr	_BIG5State *bs;
98121893Stjr	wchar_t wc;
99128004Stjr	int i, len, ocount;
100128004Stjr	size_t ncopy;
10138333Sphk
102128004Stjr	bs = (_BIG5State *)ps;
103128004Stjr
104128155Stjr	if (bs->count < 0 || bs->count > sizeof(bs->bytes)) {
105128155Stjr		errno = EINVAL;
106128155Stjr		return ((size_t)-1);
107128155Stjr	}
108128155Stjr
109128004Stjr	if (s == NULL) {
110128004Stjr		s = "";
111128004Stjr		n = 1;
112128004Stjr		pwc = NULL;
113128004Stjr	}
114128004Stjr
115128004Stjr	ncopy = MIN(MIN(n, MB_CUR_MAX), sizeof(bs->bytes) - bs->count);
116128004Stjr	memcpy(bs->bytes + bs->count, s, ncopy);
117128004Stjr	ocount = bs->count;
118128004Stjr	bs->count += ncopy;
119128004Stjr	s = (char *)bs->bytes;
120128004Stjr	n = bs->count;
121128004Stjr
122121893Stjr	if (n == 0 || (size_t)(len = _big5_check(*s)) > n)
123121893Stjr		/* Incomplete multibyte sequence */
124121893Stjr		return ((size_t)-2);
125121893Stjr	wc = 0;
126121893Stjr	i = len;
127121893Stjr	while (i-- > 0)
128121893Stjr		wc = (wc << 8) | (unsigned char)*s++;
129121893Stjr	if (pwc != NULL)
130121893Stjr		*pwc = wc;
131128004Stjr	bs->count = 0;
132128004Stjr	return (wc == L'\0' ? 0 : len - ocount);
13338333Sphk}
13438333Sphk
135121893Stjrsize_t
136128155Stjr_BIG5_wcrtomb(char * __restrict s, wchar_t wc, mbstate_t * __restrict ps)
13738333Sphk{
138128155Stjr	_BIG5State *bs;
139121893Stjr
140128155Stjr	bs = (_BIG5State *)ps;
141128155Stjr
142128155Stjr	if (bs->count != 0) {
143128155Stjr		errno = EINVAL;
144128155Stjr		return ((size_t)-1);
145128155Stjr	}
146128155Stjr
147121893Stjr	if (s == NULL)
148121893Stjr		/* Reset to initial shift state (no-op) */
149121893Stjr		return (1);
150121893Stjr	if (wc & 0x8000) {
151121893Stjr		*s++ = (wc >> 8) & 0xff;
152121893Stjr		*s = wc & 0xff;
153121893Stjr		return (2);
15438333Sphk	}
155121893Stjr	*s = wc & 0xff;
156121893Stjr	return (1);
15738333Sphk}
158