xref: /illumos-gate/usr/src/lib/libc/port/locale/gb2312.c (revision 1fa2a66491e7d8ae0be84e7da4da8e812480c710)
1 /*
2  * Copyright 2013 Garrett D'Amore <garrett@damore.org>
3  * Copyright 2010 Nexenta Systems, Inc.  All rights reserved.
4  * Copyright (c) 2004 Tim J. Robbins. All rights reserved.
5  * Copyright (c) 2003 David Xu <davidxu@freebsd.org>
6  * All rights reserved.
7  *
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.
16  *
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.
28  */
29 
30 #include "lint.h"
31 #include <sys/types.h>
32 #include <errno.h>
33 #include "runetype.h"
34 #include <stdlib.h>
35 #include <string.h>
36 #include <wchar.h>
37 #include "mblocal.h"
38 #include "lctype.h"
39 
40 static size_t	_GB2312_mbrtowc(wchar_t *_RESTRICT_KYWD,
41 		    const char *_RESTRICT_KYWD,
42 		    size_t, mbstate_t *_RESTRICT_KYWD, boolean_t);
43 static int	_GB2312_mbsinit(const mbstate_t *);
44 static size_t	_GB2312_wcrtomb(char *_RESTRICT_KYWD, wchar_t,
45 		    mbstate_t *_RESTRICT_KYWD);
46 static size_t	_GB2312_mbsnrtowcs(wchar_t *_RESTRICT_KYWD,
47 		    const char **_RESTRICT_KYWD, size_t, size_t,
48 		    mbstate_t *_RESTRICT_KYWD);
49 static size_t	_GB2312_wcsnrtombs(char *_RESTRICT_KYWD,
50 		    const wchar_t **_RESTRICT_KYWD, size_t, size_t,
51 		    mbstate_t *_RESTRICT_KYWD);
52 
53 void
54 _GB2312_init(struct lc_ctype *lct)
55 {
56 
57 	lct->lc_mbrtowc = _GB2312_mbrtowc;
58 	lct->lc_wcrtomb = _GB2312_wcrtomb;
59 	lct->lc_mbsinit = _GB2312_mbsinit;
60 	lct->lc_mbsnrtowcs = _GB2312_mbsnrtowcs;
61 	lct->lc_wcsnrtombs = _GB2312_wcsnrtombs;
62 	lct->lc_max_mblen = 2;
63 	lct->lc_is_ascii = 0;
64 }
65 
66 static int
67 _GB2312_mbsinit(const mbstate_t *ps)
68 {
69 
70 	return (ps == NULL || ((const _GB2312State *)ps)->count == 0);
71 }
72 
73 static int
74 _GB2312_check(const char *str, size_t n)
75 {
76 	const uchar_t *s = (const uchar_t *)str;
77 
78 	if (n == 0)
79 		/* Incomplete multibyte sequence */
80 		return (-2);
81 	if (s[0] >= 0xa1 && s[0] <= 0xfe) {
82 		if (n < 2)
83 			/* Incomplete multibyte sequence */
84 			return (-2);
85 		if (s[1] < 0xa1 || s[1] > 0xfe)
86 			/* Invalid multibyte sequence */
87 			return (-1);
88 		return (2);
89 	} else if (s[0] & 0x80) {
90 		/* Invalid multibyte sequence */
91 		return (-1);
92 	}
93 	return (1);
94 }
95 
96 static size_t
97 _GB2312_mbrtowc(wchar_t *_RESTRICT_KYWD pwc, const char *_RESTRICT_KYWD s,
98     size_t n, mbstate_t *_RESTRICT_KYWD ps, boolean_t zero)
99 {
100 	_GB2312State *gs;
101 	wchar_t wc;
102 	int i, len, ocount;
103 	size_t ncopy;
104 
105 	gs = (_GB2312State *)ps;
106 
107 	if (gs->count < 0 || gs->count > sizeof (gs->bytes)) {
108 		errno = EINVAL;
109 		return ((size_t)-1);
110 	}
111 
112 	if (s == NULL) {
113 		s = "";
114 		n = 1;
115 		pwc = NULL;
116 	}
117 
118 	ncopy = MIN(MIN(n, MB_CUR_MAX), sizeof (gs->bytes) - gs->count);
119 	(void) memcpy(gs->bytes + gs->count, s, ncopy);
120 	ocount = gs->count;
121 	gs->count += ncopy;
122 	s = (char *)gs->bytes;
123 	n = gs->count;
124 
125 	if ((len = _GB2312_check(s, n)) < 0)
126 		return ((size_t)len);
127 	wc = 0;
128 	i = len;
129 	while (i-- > 0)
130 		wc = (wc << 8) | (unsigned char)*s++;
131 	if (pwc != NULL)
132 		*pwc = wc;
133 	gs->count = 0;
134 	if (zero || wc != L'\0') {
135 		return (len - ocount);
136 	} else {
137 		return (0);
138 	}
139 }
140 
141 static size_t
142 _GB2312_wcrtomb(char *_RESTRICT_KYWD s, wchar_t wc,
143     mbstate_t *_RESTRICT_KYWD ps)
144 {
145 	_GB2312State *gs;
146 
147 	gs = (_GB2312State *)ps;
148 
149 	if (gs->count != 0) {
150 		errno = EINVAL;
151 		return ((size_t)-1);
152 	}
153 
154 	if (s == NULL)
155 		/* Reset to initial shift state (no-op) */
156 		return (1);
157 	if (wc & 0x8000) {
158 		*s++ = (wc >> 8) & 0xff;
159 		*s = wc & 0xff;
160 		return (2);
161 	}
162 	*s = wc & 0xff;
163 	return (1);
164 }
165 
166 static size_t
167 _GB2312_mbsnrtowcs(wchar_t *_RESTRICT_KYWD dst,
168     const char **_RESTRICT_KYWD src, size_t nms, size_t len,
169     mbstate_t *_RESTRICT_KYWD ps)
170 {
171 	return (__mbsnrtowcs_std(dst, src, nms, len, ps, _GB2312_mbrtowc));
172 }
173 
174 static size_t
175 _GB2312_wcsnrtombs(char *_RESTRICT_KYWD dst,
176     const wchar_t **_RESTRICT_KYWD src, size_t nwc, size_t len,
177     mbstate_t *_RESTRICT_KYWD ps)
178 {
179 	return (__wcsnrtombs_std(dst, src, nwc, len, ps, _GB2312_wcrtomb));
180 }
181