xref: /freebsd/lib/libc/locale/gbk.c (revision 38f0b757fd84d17d0fc24739a7cda160c4516d81)
1 /*-
2  * Copyright (c) 2002-2004 Tim J. Robbins. All rights reserved.
3  * Copyright (c) 1993
4  *	The Regents of the University of California.  All rights reserved.
5  *
6  * This code is derived from software contributed to Berkeley by
7  * Paul Borman at Krystal Technologies.
8  *
9  * Copyright (c) 2011 The FreeBSD Foundation
10  * All rights reserved.
11  * Portions of this software were developed by David Chisnall
12  * under sponsorship from the FreeBSD Foundation.
13  *
14  * Redistribution and use in source and binary forms, with or without
15  * modification, are permitted provided that the following conditions
16  * are met:
17  * 1. Redistributions of source code must retain the above copyright
18  *    notice, this list of conditions and the following disclaimer.
19  * 2. Redistributions in binary form must reproduce the above copyright
20  *    notice, this list of conditions and the following disclaimer in the
21  *    documentation and/or other materials provided with the distribution.
22  * 4. Neither the name of the University nor the names of its contributors
23  *    may be used to endorse or promote products derived from this software
24  *    without specific prior written permission.
25  *
26  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
27  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
28  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
29  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
30  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
31  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
32  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
33  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
34  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
35  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
36  * SUCH DAMAGE.
37  */
38 
39 #include <sys/cdefs.h>
40 __FBSDID("$FreeBSD$");
41 
42 #include <sys/types.h>
43 #include <errno.h>
44 #include <runetype.h>
45 #include <stdlib.h>
46 #include <string.h>
47 #include <wchar.h>
48 #include "mblocal.h"
49 
50 extern int __mb_sb_limit;
51 
52 static size_t	_GBK_mbrtowc(wchar_t * __restrict, const char * __restrict,
53 		    size_t, mbstate_t * __restrict);
54 static int	_GBK_mbsinit(const mbstate_t *);
55 static size_t	_GBK_wcrtomb(char * __restrict, wchar_t,
56 		    mbstate_t * __restrict);
57 
58 typedef struct {
59 	wchar_t	ch;
60 } _GBKState;
61 
62 int
63 _GBK_init(struct xlocale_ctype *l, _RuneLocale *rl)
64 {
65 
66 	l->__mbrtowc = _GBK_mbrtowc;
67 	l->__wcrtomb = _GBK_wcrtomb;
68 	l->__mbsinit = _GBK_mbsinit;
69 	l->runes = rl;
70 	l->__mb_cur_max = 2;
71 	l->__mb_sb_limit = 128;
72 	return (0);
73 }
74 
75 static int
76 _GBK_mbsinit(const mbstate_t *ps)
77 {
78 
79 	return (ps == NULL || ((const _GBKState *)ps)->ch == 0);
80 }
81 
82 static __inline int
83 _gbk_check(u_int c)
84 {
85 
86 	c &= 0xff;
87 	return ((c >= 0x81 && c <= 0xfe) ? 2 : 1);
88 }
89 
90 static size_t
91 _GBK_mbrtowc(wchar_t * __restrict pwc, const char * __restrict s, size_t n,
92     mbstate_t * __restrict ps)
93 {
94 	_GBKState *gs;
95 	wchar_t wc;
96 	size_t len;
97 
98 	gs = (_GBKState *)ps;
99 
100 	if ((gs->ch & ~0xFF) != 0) {
101 		/* Bad conversion state. */
102 		errno = EINVAL;
103 		return ((size_t)-1);
104 	}
105 
106 	if (s == NULL) {
107 		s = "";
108 		n = 1;
109 		pwc = NULL;
110 	}
111 
112 	if (n == 0)
113 		/* Incomplete multibyte sequence */
114 		return ((size_t)-2);
115 
116 	if (gs->ch != 0) {
117 		if (*s == '\0') {
118 			errno = EILSEQ;
119 			return ((size_t)-1);
120 		}
121 		wc = (gs->ch << 8) | (*s & 0xFF);
122 		if (pwc != NULL)
123 			*pwc = wc;
124 		gs->ch = 0;
125 		return (1);
126 	}
127 
128 	len = (size_t)_gbk_check(*s);
129 	wc = *s++ & 0xff;
130 	if (len == 2) {
131 		if (n < 2) {
132 			/* Incomplete multibyte sequence */
133 			gs->ch = wc;
134 			return ((size_t)-2);
135 		}
136 		if (*s == '\0') {
137 			errno = EILSEQ;
138 			return ((size_t)-1);
139 		}
140 		wc = (wc << 8) | (*s++ & 0xff);
141 		if (pwc != NULL)
142 			*pwc = wc;
143                 return (2);
144 	} else {
145 		if (pwc != NULL)
146 			*pwc = wc;
147 		return (wc == L'\0' ? 0 : 1);
148 	}
149 }
150 
151 static size_t
152 _GBK_wcrtomb(char * __restrict s, wchar_t wc, mbstate_t * __restrict ps)
153 {
154 	_GBKState *gs;
155 
156 	gs = (_GBKState *)ps;
157 
158 	if (gs->ch != 0) {
159 		errno = EINVAL;
160 		return ((size_t)-1);
161 	}
162 
163 	if (s == NULL)
164 		/* Reset to initial shift state (no-op) */
165 		return (1);
166 	if (wc & 0x8000) {
167 		*s++ = (wc >> 8) & 0xff;
168 		*s = wc & 0xff;
169 		return (2);
170 	}
171 	*s = wc & 0xff;
172 	return (1);
173 }
174