compiler/clib/locale/collate.c

   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
  28 #include <sys/cdefs.h>
  29 __FBSDID("$FreeBSD: src/lib/libc/locale/collate.c,v 1.35 2005/02/27 20:31:13 ru Exp $");
  30
  31 #include "namespace.h"
  32 #ifdef __AROS__
  33 /* until we have arpa/inet.h in libc */
  34 #include <exec/types.h>
  35 #include <aros/macros.h>
  36 #define ntohl(x) AROS_BE2LONG(x)
  37 #else
  38 #include <arpa/inet.h>
  39 #endif
  40 #include <stdio.h>
  41 #include <stdlib.h>
  42 #include <string.h>
  43 #include <errno.h>
  44 #include <unistd.h>
  45 #include <sysexits.h>
  46 #include "un-namespace.h"
  47
  48 #include "collate.h"
  49 #include "setlocale.h"
  50 #include "ldpart.h"
  51
  52 #include "libc_private.h"
  53
  54 int __collate_load_error = 1;
  55 int __collate_substitute_nontrivial;
  56
  57 u_char __collate_substitute_table[UCHAR_MAX + 1][STR_LEN];
  58 struct __collate_st_char_pri __collate_char_pri_table[UCHAR_MAX + 1];
  59 struct __collate_st_chain_pri *__collate_chain_pri_table;
  60
  61 void __collate_err(int ex, const char *f) __dead2;
  62
  63 int
  64 __collate_load_tables(const char *encoding)
  65 {
  66         FILE *fp;
  67         int i, saverr, chains;
  68         uint32_t u32;
  69         char strbuf[STR_LEN], buf[PATH_MAX];
  70         void *TMP_substitute_table, *TMP_char_pri_table, *TMP_chain_pri_table;
  71         static char collate_encoding[ENCODING_LEN + 1];
  72
  73         /* 'encoding' must be already checked. */
  74         if (strcmp(encoding, "C") == 0 || strcmp(encoding, "POSIX") == 0) {
  75                 __collate_load_error = 1;
  76                 return (_LDP_CACHE);
  77         }
  78
  79         /*
  80          * If the locale name is the same as our cache, use the cache.
  81          */
  82         if (strcmp(encoding, collate_encoding) == 0) {
  83                 __collate_load_error = 0;
  84                 return (_LDP_CACHE);
  85         }
  86
  87         /*
  88          * Slurp the locale file into the cache.
  89          */
  90
  91         /* 'PathLocale' must be already set & checked. */
  92         /* Range checking not needed, encoding has fixed size */
  93         (void)strcpy(buf, _PathLocale);
  94         (void)strcat(buf, "/");
  95         (void)strcat(buf, encoding);
  96         (void)strcat(buf, "/LC_COLLATE");
  97         if ((fp = fopen(buf, "r")) == NULL)
  98                 return (_LDP_ERROR);
  99
 100         if (fread(strbuf, sizeof(strbuf), 1, fp) != 1) {
 101                 saverr = errno;
 102                 (void)fclose(fp);
 103                 errno = saverr;
 104                 return (_LDP_ERROR);
 105         }
 106         chains = -1;
 107         if (strcmp(strbuf, COLLATE_VERSION) == 0)
 108                 chains = 0;
 109         else if (strcmp(strbuf, COLLATE_VERSION1_2) == 0)
 110                 chains = 1;
 111         if (chains < 0) {
 112                 (void)fclose(fp);
 113                 errno = EFTYPE;
 114                 return (_LDP_ERROR);
 115         }
 116         if (chains) {
 117                 if (fread(&u32, sizeof(u32), 1, fp) != 1) {
 118                         saverr = errno;
 119                         (void)fclose(fp);
 120                         errno = saverr;
 121                         return (_LDP_ERROR);
 122                 }
 123                 if ((chains = (int)ntohl(u32)) < 1) {
 124                         (void)fclose(fp);
 125                         errno = EFTYPE;
 126                         return (_LDP_ERROR);
 127                 }
 128         } else
 129                 chains = TABLE_SIZE;
 130
 131         if ((TMP_substitute_table =
 132              malloc(sizeof(__collate_substitute_table))) == NULL) {
 133                 saverr = errno;
 134                 (void)fclose(fp);
 135                 errno = saverr;
 136                 return (_LDP_ERROR);
 137         }
 138         if ((TMP_char_pri_table =
 139              malloc(sizeof(__collate_char_pri_table))) == NULL) {
 140                 saverr = errno;
 141                 free(TMP_substitute_table);
 142                 (void)fclose(fp);
 143                 errno = saverr;
 144                 return (_LDP_ERROR);
 145         }
 146         if ((TMP_chain_pri_table =
 147              malloc(sizeof(*__collate_chain_pri_table) * chains)) == NULL) {
 148                 saverr = errno;
 149                 free(TMP_substitute_table);
 150                 free(TMP_char_pri_table);
 151                 (void)fclose(fp);
 152                 errno = saverr;
 153                 return (_LDP_ERROR);
 154         }
 155
 156 #define FREAD(a, b, c, d) \
 157 { \
 158         if (fread(a, b, c, d) != c) { \
 159                 saverr = errno; \
 160                 free(TMP_substitute_table); \
 161                 free(TMP_char_pri_table); \
 162                 free(TMP_chain_pri_table); \
 163                 (void)fclose(d); \
 164                 errno = saverr; \
 165                 return (_LDP_ERROR); \
 166         } \
 167 }
 168
 169         FREAD(TMP_substitute_table, sizeof(__collate_substitute_table), 1, fp);
 170         FREAD(TMP_char_pri_table, sizeof(__collate_char_pri_table), 1, fp);
 171         FREAD(TMP_chain_pri_table,
 172               sizeof(*__collate_chain_pri_table), chains, fp);
 173         (void)fclose(fp);
 174
 175         (void)strcpy(collate_encoding, encoding);
 176         if (__collate_substitute_table_ptr != NULL)
 177                 free(__collate_substitute_table_ptr);
 178         __collate_substitute_table_ptr = TMP_substitute_table;
 179         if (__collate_char_pri_table_ptr != NULL)
 180                 free(__collate_char_pri_table_ptr);
 181         __collate_char_pri_table_ptr = TMP_char_pri_table;
 182         for (i = 0; i < UCHAR_MAX + 1; i++) {
 183                 __collate_char_pri_table[i].prim =
 184                     ntohl(__collate_char_pri_table[i].prim);
 185                 __collate_char_pri_table[i].sec =
 186                     ntohl(__collate_char_pri_table[i].sec);
 187         }
 188         if (__collate_chain_pri_table != NULL)
 189                 free(__collate_chain_pri_table);
 190         __collate_chain_pri_table = TMP_chain_pri_table;
 191         for (i = 0; i < chains; i++) {
 192                 __collate_chain_pri_table[i].prim =
 193                     ntohl(__collate_chain_pri_table[i].prim);
 194                 __collate_chain_pri_table[i].sec =
 195                     ntohl(__collate_chain_pri_table[i].sec);
 196         }
 197         __collate_substitute_nontrivial = 0;
 198         for (i = 0; i < UCHAR_MAX + 1; i++) {
 199                 if (__collate_substitute_table[i][0] != i ||
 200                     __collate_substitute_table[i][1] != 0) {
 201                         __collate_substitute_nontrivial = 1;
 202                         break;
 203                 }
 204         }
 205         __collate_load_error = 0;
 206
 207         return (_LDP_LOADED);
 208 }
 209
 210 u_char *
 211 __collate_substitute(const u_char *s)
 212 {
 213         int dest_len, len, nlen;
 214         int delta = strlen(s);
 215         u_char *dest_str = NULL;
 216
 217         if (s == NULL || *s == '\0')
 218                 return (__collate_strdup(""));
 219         delta += delta / 8;
 220         dest_str = malloc(dest_len = delta);
 221         if (dest_str == NULL)
 222                 __collate_err(EX_OSERR, __func__);
 223         len = 0;
 224         while (*s) {
 225                 nlen = len + strlen(__collate_substitute_table[*s]);
 226                 if (dest_len <= nlen) {
 227                         dest_str = reallocf(dest_str, dest_len = nlen + delta);
 228                         if (dest_str == NULL)
 229                                 __collate_err(EX_OSERR, __func__);
 230                 }
 231                 (void)strcpy(dest_str + len, __collate_substitute_table[*s++]);
 232                 len = nlen;
 233         }
 234         return (dest_str);
 235 }
 236
 237 void
 238 __collate_lookup(const u_char *t, int *len, int *prim, int *sec)
 239 {
 240         struct __collate_st_chain_pri *p2;
 241
 242         *len = 1;
 243         *prim = *sec = 0;
 244         for (p2 = __collate_chain_pri_table; p2->str[0] != '\0'; p2++) {
 245                 if (*t == p2->str[0] &&
 246                     strncmp(t, p2->str, strlen(p2->str)) == 0) {
 247                         *len = strlen(p2->str);
 248                         *prim = p2->prim;
 249                         *sec = p2->sec;
 250                         return;
 251                 }
 252         }
 253         *prim = __collate_char_pri_table[*t].prim;
 254         *sec = __collate_char_pri_table[*t].sec;
 255 }
 256
 257 u_char *
 258 __collate_strdup(u_char *s)
 259 {
 260         u_char *t = strdup(s);
 261
 262         if (t == NULL)
 263                 __collate_err(EX_OSERR, __func__);
 264         return (t);
 265 }
 266
 267 void
 268 __collate_err(int ex, const char *f)
 269 {
 270         const char *s;
 271         int serrno = errno;
 272
 273 #ifndef __AROS__ /* fix this some day */
 274         s = _getprogname();
 275         _write(STDERR_FILENO, s, strlen(s));
 276 #endif
 277         _write(STDERR_FILENO, ": ", 2);
 278         s = f;
 279         _write(STDERR_FILENO, s, strlen(s));
 280         _write(STDERR_FILENO, ": ", 2);
 281         s = strerror(serrno);
 282         _write(STDERR_FILENO, s, strlen(s));
 283         _write(STDERR_FILENO, "\n", 1);
 284         exit(ex);
 285 }
 286
 287 #ifdef COLLATE_DEBUG
 288 void
 289 __collate_print_tables()
 290 {
 291         int i;
 292         struct __collate_st_chain_pri *p2;
 293
 294         printf("Substitute table:\n");
 295         for (i = 0; i < UCHAR_MAX + 1; i++)
 296             if (i != *__collate_substitute_table[i])
 297                 printf("\t'%c' --> \"%s\"\n", i,
 298                        __collate_substitute_table[i]);
 299         printf("Chain priority table:\n");
 300         for (p2 = __collate_chain_pri_table; p2->str[0] != '\0'; p2++)
 301                 printf("\t\"%s\" : %d %d\n", p2->str, p2->prim, p2->sec);
 302         printf("Char priority table:\n");
 303         for (i = 0; i < UCHAR_MAX + 1; i++)
 304                 printf("\t'%c' : %d %d\n", i, __collate_char_pri_table[i].prim,
 305                        __collate_char_pri_table[i].sec);
 306 }
 307 #endif