xref: /dragonfly/lib/libc/locale/collate.c (revision 6e285212) 
1 /*-
2  * Copyright (c) 1995 Alex Tatmanjants <alex@elvisti.kiev.ua>
3  *		at Electronni Visti IA, Kiev, Ukraine.
4  *			All rights reserved.
5  *
6  * Redistribution and use in source and binary forms, with or without
7  * modification, are permitted provided that the following conditions
8  * are met:
9  * 1. Redistributions of source code must retain the above copyright
10  *    notice, this list of conditions and the following disclaimer.
11  * 2. Redistributions in binary form must reproduce the above copyright
12  *    notice, this list of conditions and the following disclaimer in the
13  *    documentation and/or other materials provided with the distribution.
14  *
15  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND
16  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
17  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
18  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR BE LIABLE
19  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
20  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
21  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
22  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
23  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
24  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
25  * SUCH DAMAGE.
26  *
27  * $FreeBSD: src/lib/libc/locale/collate.c,v 1.21.2.4 2002/10/11 10:36:47 ache Exp $
28  * $DragonFly: src/lib/libc/locale/collate.c,v 1.2 2003/06/17 04:26:43 dillon Exp $
29  */
30 
31 #include <arpa/inet.h>
32 #include <rune.h>
33 #include <stdio.h>
34 #include <stdlib.h>
35 #include <string.h>
36 #include <errno.h>
37 #include <unistd.h>
38 #include <sysexits.h>
39 
40 #include "collate.h"
41 #include "setlocale.h"
42 #include "ldpart.h"
43 
44 int __collate_load_error = 1;
45 int __collate_substitute_nontrivial;
46 
47 u_char __collate_substitute_table[UCHAR_MAX + 1][STR_LEN];
48 struct __collate_st_char_pri __collate_char_pri_table[UCHAR_MAX + 1];
49 struct __collate_st_chain_pri *__collate_chain_pri_table;
50 
51 void __collate_err(int ex, const char *f) __dead2;
52 
53 int
54 __collate_load_tables(const char *encoding)
55 {
56 	FILE *fp;
57 	int i, saverr, chains;
58 	uint32_t u32;
59 	char strbuf[STR_LEN], buf[PATH_MAX];
60 	void *TMP_substitute_table, *TMP_char_pri_table, *TMP_chain_pri_table;
61 	static char collate_encoding[ENCODING_LEN + 1];
62 
63 	/* 'encoding' must be already checked. */
64 	if (strcmp(encoding, "C") == 0 || strcmp(encoding, "POSIX") == 0) {
65 		__collate_load_error = 1;
66 		return (_LDP_CACHE);
67 	}
68 
69 	/*
70 	 * If the locale name is the same as our cache, use the cache.
71 	 */
72 	if (strcmp(encoding, collate_encoding) == 0) {
73 		__collate_load_error = 0;
74 		return (_LDP_CACHE);
75 	}
76 
77 	/*
78 	 * Slurp the locale file into the cache.
79 	 */
80 
81 	/* 'PathLocale' must be already set & checked. */
82 	/* Range checking not needed, encoding has fixed size */
83 	(void)strcpy(buf, _PathLocale);
84 	(void)strcat(buf, "/");
85 	(void)strcat(buf, encoding);
86 	(void)strcat(buf, "/LC_COLLATE");
87 	if ((fp = fopen(buf, "r")) == NULL)
88 		return (_LDP_ERROR);
89 
90 	if (fread(strbuf, sizeof(strbuf), 1, fp) != 1) {
91 		saverr = errno;
92 		(void)fclose(fp);
93 		errno = saverr;
94 		return (_LDP_ERROR);
95 	}
96 	chains = -1;
97 	if (strcmp(strbuf, COLLATE_VERSION) == 0)
98 		chains = 0;
99 	else if (strcmp(strbuf, COLLATE_VERSION1_1) == 0)
100 		chains = 1;
101 	if (chains < 0) {
102 		(void)fclose(fp);
103 		errno = EFTYPE;
104 		return (_LDP_ERROR);
105 	}
106 	if (chains) {
107 		if (fread(&u32, sizeof(u32), 1, fp) != 1) {
108 			saverr = errno;
109 			(void)fclose(fp);
110 			errno = saverr;
111 			return (_LDP_ERROR);
112 		}
113 		if ((chains = (int)ntohl(u32)) < 1) {
114 			(void)fclose(fp);
115 			errno = EFTYPE;
116 			return (_LDP_ERROR);
117 		}
118 	} else
119 		chains = TABLE_SIZE;
120 
121 	if ((TMP_substitute_table =
122 	     malloc(sizeof(__collate_substitute_table))) == NULL) {
123 		saverr = errno;
124 		(void)fclose(fp);
125 		errno = saverr;
126 		return (_LDP_ERROR);
127 	}
128 	if ((TMP_char_pri_table =
129 	     malloc(sizeof(__collate_char_pri_table))) == NULL) {
130 		saverr = errno;
131 		free(TMP_substitute_table);
132 		(void)fclose(fp);
133 		errno = saverr;
134 		return (_LDP_ERROR);
135 	}
136 	if ((TMP_chain_pri_table =
137 	     malloc(sizeof(*__collate_chain_pri_table) * chains)) == NULL) {
138 		saverr = errno;
139 		free(TMP_substitute_table);
140 		free(TMP_char_pri_table);
141 		(void)fclose(fp);
142 		errno = saverr;
143 		return (_LDP_ERROR);
144 	}
145 
146 #define FREAD(a, b, c, d) \
147 { \
148 	if (fread(a, b, c, d) != c) { \
149 		saverr = errno; \
150 		free(TMP_substitute_table); \
151 		free(TMP_char_pri_table); \
152 		free(TMP_chain_pri_table); \
153 		(void)fclose(d); \
154 		errno = saverr; \
155 		return (_LDP_ERROR); \
156 	} \
157 }
158 
159 	FREAD(TMP_substitute_table, sizeof(__collate_substitute_table), 1, fp);
160 	FREAD(TMP_char_pri_table, sizeof(__collate_char_pri_table), 1, fp);
161 	FREAD(TMP_chain_pri_table,
162 	      sizeof(*__collate_chain_pri_table), chains, fp);
163 	(void)fclose(fp);
164 
165 	(void)strcpy(collate_encoding, encoding);
166 	if (__collate_substitute_table_ptr != NULL)
167 		free(__collate_substitute_table_ptr);
168 	__collate_substitute_table_ptr = TMP_substitute_table;
169 	if (__collate_char_pri_table_ptr != NULL)
170 		free(__collate_char_pri_table_ptr);
171 	__collate_char_pri_table_ptr = TMP_char_pri_table;
172 	if (__collate_chain_pri_table != NULL)
173 		free(__collate_chain_pri_table);
174 	__collate_chain_pri_table = TMP_chain_pri_table;
175 
176 	__collate_substitute_nontrivial = 0;
177 	for (i = 0; i < UCHAR_MAX + 1; i++) {
178 		if (__collate_substitute_table[i][0] != i ||
179 		    __collate_substitute_table[i][1] != 0) {
180 			__collate_substitute_nontrivial = 1;
181 			break;
182 		}
183 	}
184 	__collate_load_error = 0;
185 
186 	return (_LDP_LOADED);
187 }
188 
189 u_char *
190 __collate_substitute(s)
191 	const u_char *s;
192 {
193 	int dest_len, len, nlen;
194 	int delta = strlen(s);
195 	u_char *dest_str = NULL;
196 
197 	if (s == NULL || *s == '\0')
198 		return (__collate_strdup(""));
199 	delta += delta / 8;
200 	dest_str = malloc(dest_len = delta);
201 	if (dest_str == NULL)
202 		__collate_err(EX_OSERR, __FUNCTION__);
203 	len = 0;
204 	while (*s) {
205 		nlen = len + strlen(__collate_substitute_table[*s]);
206 		if (dest_len <= nlen) {
207 			dest_str = reallocf(dest_str, dest_len = nlen + delta);
208 			if (dest_str == NULL)
209 				__collate_err(EX_OSERR, __FUNCTION__);
210 		}
211 		(void)strcpy(dest_str + len, __collate_substitute_table[*s++]);
212 		len = nlen;
213 	}
214 	return (dest_str);
215 }
216 
217 void
218 __collate_lookup(t, len, prim, sec)
219 	const u_char *t;
220 	int *len, *prim, *sec;
221 {
222 	struct __collate_st_chain_pri *p2;
223 
224 	*len = 1;
225 	*prim = *sec = 0;
226 	for (p2 = __collate_chain_pri_table; p2->str[0] != '\0'; p2++) {
227 		if (*t == p2->str[0] &&
228 		    strncmp(t, p2->str, strlen(p2->str)) == 0) {
229 			*len = strlen(p2->str);
230 			*prim = p2->prim;
231 			*sec = p2->sec;
232 			return;
233 		}
234 	}
235 	*prim = __collate_char_pri_table[*t].prim;
236 	*sec = __collate_char_pri_table[*t].sec;
237 }
238 
239 u_char *
240 __collate_strdup(s)
241 	u_char *s;
242 {
243 	u_char *t = strdup(s);
244 
245 	if (t == NULL)
246 		__collate_err(EX_OSERR, __FUNCTION__);
247 	return (t);
248 }
249 
250 void
251 __collate_err(int ex, const char *f)
252 {
253 	extern char *__progname;                /* Program name, from crt0. */
254 	const char *s;
255 	int serrno = errno;
256 
257 	s = __progname;
258 	_write(STDERR_FILENO, s, strlen(s));
259 	_write(STDERR_FILENO, ": ", 2);
260 	s = f;
261 	_write(STDERR_FILENO, s, strlen(s));
262 	_write(STDERR_FILENO, ": ", 2);
263 	s = strerror(serrno);
264 	_write(STDERR_FILENO, s, strlen(s));
265 	_write(STDERR_FILENO, "\n", 1);
266 	exit(ex);
267 }
268 
269 #ifdef COLLATE_DEBUG
270 void
271 __collate_print_tables()
272 {
273 	int i;
274 	struct __collate_st_chain_pri *p2;
275 
276 	printf("Substitute table:\n");
277 	for (i = 0; i < UCHAR_MAX + 1; i++)
278 	    if (i != *__collate_substitute_table[i])
279 		printf("\t'%c' --> \"%s\"\n", i,
280 		       __collate_substitute_table[i]);
281 	printf("Chain priority table:\n");
282 	for (p2 = __collate_chain_pri_table; p2->str[0] != '\0'; p2++)
283 		printf("\t\"%s\" : %d %d\n", p2->str, p2->prim, p2->sec);
284 	printf("Char priority table:\n");
285 	for (i = 0; i < UCHAR_MAX + 1; i++)
286 		printf("\t'%c' : %d %d\n", i, __collate_char_pri_table[i].prim,
287 		       __collate_char_pri_table[i].sec);
288 }
289 #endif
290