xref: /freebsd/lib/libc/locale/euc.c (revision 74bf4e164ba5851606a27d4feff27717452583e5)
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  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. All advertising materials mentioning features or use of this software
18  *    must display the following acknowledgement:
19  *	This product includes software developed by the University of
20  *	California, Berkeley and its contributors.
21  * 4. Neither the name of the University nor the names of its contributors
22  *    may be used to endorse or promote products derived from this software
23  *    without specific prior written permission.
24  *
25  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35  * SUCH DAMAGE.
36  */
37 
38 #if defined(LIBC_SCCS) && !defined(lint)
39 static char sccsid[] = "@(#)euc.c	8.1 (Berkeley) 6/4/93";
40 #endif /* LIBC_SCCS and not lint */
41 #include <sys/param.h>
42 __FBSDID("$FreeBSD$");
43 
44 #include <errno.h>
45 #include <limits.h>
46 #include <runetype.h>
47 #include <stdlib.h>
48 #include <string.h>
49 #include <wchar.h>
50 #include "mblocal.h"
51 
52 int	_EUC_init(_RuneLocale *);
53 size_t	_EUC_mbrtowc(wchar_t * __restrict, const char * __restrict, size_t,
54 	    mbstate_t * __restrict);
55 int	_EUC_mbsinit(const mbstate_t *);
56 size_t	_EUC_wcrtomb(char * __restrict, wchar_t, mbstate_t * __restrict);
57 
58 typedef struct {
59 	int	count[4];
60 	wchar_t	bits[4];
61 	wchar_t	mask;
62 } _EucInfo;
63 
64 typedef struct {
65 	wchar_t	ch;
66 	int	set;
67 	int	want;
68 } _EucState;
69 
70 int
71 _EUC_init(_RuneLocale *rl)
72 {
73 	_EucInfo *ei;
74 	int x, new__mb_cur_max;
75 	char *v, *e;
76 
77 	if (rl->__variable == NULL)
78 		return (EFTYPE);
79 
80 	v = (char *)rl->__variable;
81 
82 	while (*v == ' ' || *v == '\t')
83 		++v;
84 
85 	if ((ei = malloc(sizeof(_EucInfo))) == NULL)
86 		return (errno == 0 ? ENOMEM : errno);
87 
88 	new__mb_cur_max = 0;
89 	for (x = 0; x < 4; ++x) {
90 		ei->count[x] = (int)strtol(v, &e, 0);
91 		if (v == e || !(v = e)) {
92 			free(ei);
93 			return (EFTYPE);
94 		}
95 		if (new__mb_cur_max < ei->count[x])
96 			new__mb_cur_max = ei->count[x];
97 		while (*v == ' ' || *v == '\t')
98 			++v;
99 		ei->bits[x] = (int)strtol(v, &e, 0);
100 		if (v == e || !(v = e)) {
101 			free(ei);
102 			return (EFTYPE);
103 		}
104 		while (*v == ' ' || *v == '\t')
105 			++v;
106 	}
107 	ei->mask = (int)strtol(v, &e, 0);
108 	if (v == e || !(v = e)) {
109 		free(ei);
110 		return (EFTYPE);
111 	}
112 	rl->__variable = ei;
113 	rl->__variable_len = sizeof(_EucInfo);
114 	_CurrentRuneLocale = rl;
115 	__mb_cur_max = new__mb_cur_max;
116 	__mbrtowc = _EUC_mbrtowc;
117 	__wcrtomb = _EUC_wcrtomb;
118 	__mbsinit = _EUC_mbsinit;
119 	return (0);
120 }
121 
122 int
123 _EUC_mbsinit(const mbstate_t *ps)
124 {
125 
126 	return (ps == NULL || ((const _EucState *)ps)->want == 0);
127 }
128 
129 #define	CEI	((_EucInfo *)(_CurrentRuneLocale->__variable))
130 
131 #define	_SS2	0x008e
132 #define	_SS3	0x008f
133 
134 #define	GR_BITS	0x80808080 /* XXX: to be fixed */
135 
136 static __inline int
137 _euc_set(u_int c)
138 {
139 	c &= 0xff;
140 	return ((c & 0x80) ? c == _SS3 ? 3 : c == _SS2 ? 2 : 1 : 0);
141 }
142 
143 size_t
144 _EUC_mbrtowc(wchar_t * __restrict pwc, const char * __restrict s, size_t n,
145     mbstate_t * __restrict ps)
146 {
147 	_EucState *es;
148 	int i, set, want;
149 	wchar_t wc;
150 	const char *os;
151 
152 	es = (_EucState *)ps;
153 
154 	if (es->want < 0 || es->want > MB_CUR_MAX || es->set < 0 ||
155 	    es->set > 3) {
156 		errno = EINVAL;
157 		return ((size_t)-1);
158 	}
159 
160 	if (s == NULL) {
161 		s = "";
162 		n = 1;
163 		pwc = NULL;
164 	}
165 
166 	if (n == 0)
167 		/* Incomplete multibyte sequence */
168 		return ((size_t)-2);
169 
170 	os = s;
171 
172 	if (es->want == 0) {
173 		want = CEI->count[set = _euc_set(*s)];
174 		if (set == 2 || set == 3) {
175 			--want;
176 			if (--n == 0) {
177 				/* Incomplete multibyte sequence */
178 				es->set = set;
179 				es->want = want;
180 				es->ch = 0;
181 				return ((size_t)-2);
182 			}
183 			++s;
184 			if (*s == '\0') {
185 				errno = EILSEQ;
186 				return ((size_t)-1);
187 			}
188 		}
189 		wc = (unsigned char)*s++;
190 	} else {
191 		set = es->set;
192 		want = es->want;
193 		wc = es->ch;
194 	}
195 	for (i = (es->want == 0) ? 1 : 0; i < MIN(want, n); i++) {
196 		if (*s == '\0') {
197 			errno = EILSEQ;
198 			return ((size_t)-1);
199 		}
200 		wc = (wc << 8) | (unsigned char)*s++;
201 	}
202 	if (i < want) {
203 		/* Incomplete multibyte sequence */
204 		es->set = set;
205 		es->want = want - i;
206 		es->ch = wc;
207 		return ((size_t)-2);
208 	}
209 	wc = (wc & ~CEI->mask) | CEI->bits[set];
210 	if (pwc != NULL)
211 		*pwc = wc;
212 	es->want = 0;
213 	return (wc == L'\0' ? 0 : s - os);
214 }
215 
216 size_t
217 _EUC_wcrtomb(char * __restrict s, wchar_t wc, mbstate_t * __restrict ps)
218 {
219 	_EucState *es;
220 	wchar_t m, nm;
221 	int i, len;
222 
223 	es = (_EucState *)ps;
224 
225 	if (es->want != 0) {
226 		errno = EINVAL;
227 		return ((size_t)-1);
228 	}
229 
230 	if (s == NULL)
231 		/* Reset to initial shift state (no-op) */
232 		return (1);
233 
234 	m = wc & CEI->mask;
235 	nm = wc & ~m;
236 
237 	if (m == CEI->bits[1]) {
238 CodeSet1:
239 		/* Codeset 1: The first byte must have 0x80 in it. */
240 		i = len = CEI->count[1];
241 		while (i-- > 0)
242 			*s++ = (nm >> (i << 3)) | 0x80;
243 	} else {
244 		if (m == CEI->bits[0])
245 			i = len = CEI->count[0];
246 		else if (m == CEI->bits[2]) {
247 			i = len = CEI->count[2];
248 			*s++ = _SS2;
249 			--i;
250 			/* SS2 designates G2 into GR */
251 			nm |= GR_BITS;
252 		} else if (m == CEI->bits[3]) {
253 			i = len = CEI->count[3];
254 			*s++ = _SS3;
255 			--i;
256 			/* SS3 designates G3 into GR */
257 			nm |= GR_BITS;
258 		} else
259 			goto CodeSet1;	/* Bletch */
260 		while (i-- > 0)
261 			*s++ = (nm >> (i << 3)) & 0xff;
262 	}
263 	return (len);
264 }
265