lib/libkiconv/xlat16_iconv.c

   1 /*-
   2  * Copyright (c) 2003 Ryuichiro Imura
   3  * All rights reserved.
   4  *
   5  * Redistribution and use in source and binary forms, with or without
   6  * modification, are permitted provided that the following conditions
   7  * are met:
   8  * 1. Redistributions of source code must retain the above copyright
   9  *    notice, this list of conditions and the following disclaimer.
  10  * 2. Redistributions in binary form must reproduce the above copyright
  11  *    notice, this list of conditions and the following disclaimer in the
  12  *    documentation and/or other materials provided with the distribution.
  13  *
  14  * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
  15  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  16  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  17  * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
  18  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  19  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  20  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  21  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  22  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  23  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  24  * SUCH DAMAGE.
  25  *
  26  * $FreeBSD: src/lib/libkiconv/xlat16_iconv.c,v 1.4.8.1 2009/04/15 03:14:26 kensmith Exp $
  27  */
  28
  29 /*
  30  * kiconv(3) requires shared linked, and reduce module size
  31  * when statically linked.
  32  */
  33 #include <sys/types.h>
  34 #include <sys/iconv.h>
  35 #include <sys/sysctl.h>
  36
  37 #include <ctype.h>
  38 #include <dlfcn.h>
  39 #include <err.h>
  40 #include <errno.h>
  41 #include <stdio.h>
  42 #include <stdlib.h>
  43 #include <string.h>
  44 #include <iconv.h>
  45
  46 #include "quirks.h"
  47
  48 struct xlat16_table {
  49         uint32_t *      idx[0x200];
  50         void *          data;
  51         size_t          size;
  52 };
  53
  54 static struct xlat16_table kiconv_xlat16_open(const char *, const char *, int);
  55
  56 static int my_iconv_init(void);
  57 static iconv_t (*my_iconv_open)(const char *, const char *);
  58 static size_t (*my_iconv)(iconv_t, const char **, size_t *, char **, size_t *);
  59 static int (*my_iconv_close)(iconv_t);
  60 static size_t my_iconv_char(iconv_t, const u_char **, size_t *, u_char **, size_t *);
  61
  62 int
  63 kiconv_add_xlat16_cspair(const char *tocode, const char *fromcode, int flag)
  64 {
  65         int error;
  66         size_t i, size, idxsize;
  67         struct iconv_cspair_info *csi;
  68         struct xlat16_table xt;
  69         void *data;
  70         char *p;
  71         if (sysctlbyname("kern.iconv.cslist", NULL, &size, NULL, 0) == -1)
  72                 return (-1);
  73         if (size > 0) {
  74                 csi = malloc(size);
  75                 if (csi == NULL)
  76                         return (-1);
  77                 if (sysctlbyname("kern.iconv.cslist", csi, &size, NULL, 0) == -1) {
  78                         free(csi);
  79                         return (-1);
  80                 }
  81                 for (i = 0; i < (size/sizeof(*csi)); i++, csi++){
  82                         if (strcmp(csi->cs_to, tocode) == 0 &&
  83                             strcmp(csi->cs_from, fromcode) == 0)
  84                                 return (0);
  85                 }
  86         }
  87
  88         xt = kiconv_xlat16_open(tocode, fromcode, flag);
  89         if (xt.size == 0)
  90                 return (-1);
  91
  92         idxsize = sizeof(xt.idx);
  93
  94         if ((idxsize + xt.size) > ICONV_CSMAXDATALEN) {
  95                 errno = E2BIG;
  96                 return (-1);
  97         }
  98
  99         if ((data = malloc(idxsize + xt.size)) != NULL) {
 100                 p = data;
 101                 memcpy(p, xt.idx, idxsize);
 102                 p += idxsize;
 103                 memcpy(p, xt.data, xt.size);
 104                 error = kiconv_add_xlat16_table(tocode, fromcode, data,
 105                     (int)(idxsize + xt.size));
 106                 return (error);
 107         }
 108         return (-1);
 109 }
 110
 111 int
 112 kiconv_add_xlat16_cspairs(const char *foreigncode, const char *localcode)
 113 {
 114         int error;
 115
 116         error = kiconv_add_xlat16_cspair(foreigncode, localcode,
 117             KICONV_FROM_LOWER | KICONV_FROM_UPPER);
 118         if (error)
 119                 return (error);
 120         error = kiconv_add_xlat16_cspair(localcode, foreigncode,
 121             KICONV_LOWER | KICONV_UPPER);
 122         if (error)
 123                 return (error);
 124         return (0);
 125 }
 126
 127 static struct xlat16_table
 128 kiconv_xlat16_open(const char *tocode, const char *fromcode, int lcase)
 129 {
 130         u_char src[3], dst[4], *srcp, *dstp, ud, ld;
 131         int us, ls, ret;
 132         uint16_t c;
 133         uint32_t table[0x80];
 134         size_t inbytesleft, outbytesleft, pre_q_size, post_q_size;
 135         struct xlat16_table xt;
 136         struct quirk_replace_list *pre_q_list, *post_q_list;
 137         iconv_t cd;
 138         char *p;
 139
 140         xt.data = NULL;
 141         xt.size = 0;
 142
 143         src[2] = '\0';
 144         dst[3] = '\0';
 145         ret = my_iconv_init();
 146         if (ret)
 147                 return (xt);
 148         cd = my_iconv_open(search_quirk(tocode, fromcode, &pre_q_list, &pre_q_size),
 149             search_quirk(fromcode, tocode, &post_q_list, &post_q_size));
 150         if (cd == (iconv_t) (-1))
 151                 return (xt);
 152
 153         if ((xt.data = malloc(0x200 * 0x80 * sizeof(uint32_t))) == NULL)
 154                 return (xt);
 155
 156         p = xt.data;
 157
 158         for (ls = 0 ; ls < 0x200 ; ls++) {
 159                 xt.idx[ls] = NULL;
 160                 for (us = 0 ; us < 0x80 ; us++) {
 161                         srcp = src;
 162                         dstp = dst;
 163
 164                         inbytesleft = 2;
 165                         outbytesleft = 3;
 166                         bzero(dst, outbytesleft);
 167
 168                         c = ((ls & 0x100 ? us | 0x80 : us) << 8) | (u_char)ls;
 169                         c = quirk_vendor2unix(c, pre_q_list, pre_q_size);
 170                         src[0] = (u_char)(c >> 8);
 171                         src[1] = (u_char)c;
 172
 173                         ret = my_iconv_char(cd, (const u_char **)&srcp,
 174                             &inbytesleft, &dstp, &outbytesleft);
 175                         if (ret == -1) {
 176                                 table[us] = 0;
 177                                 continue;
 178                         }
 179
 180                         ud = (u_char)dst[0];
 181                         ld = (u_char)dst[1];
 182
 183                         switch(outbytesleft) {
 184                         case 0:
 185 #ifdef XLAT16_ACCEPT_3BYTE_CHR
 186                                 table[us] = (ud << 8) | ld;
 187                                 table[us] |= (u_char)dst[2] << 16;
 188                                 table[us] |= XLAT16_IS_3BYTE_CHR;
 189 #else
 190                                 table[us] = 0;
 191                                 continue;
 192 #endif
 193                                 break;
 194                         case 1:
 195                                 table[us] = quirk_unix2vendor((ud << 8) | ld,
 196                                     post_q_list, post_q_size);
 197                                 if ((table[us] >> 8) == 0)
 198                                         table[us] |= XLAT16_ACCEPT_NULL_OUT;
 199                                 break;
 200                         case 2:
 201                                 table[us] = ud;
 202                                 if (lcase & KICONV_LOWER && ud != tolower(ud)) {
 203                                         table[us] |= (u_char)tolower(ud) << 16;
 204                                         table[us] |= XLAT16_HAS_LOWER_CASE;
 205                                 }
 206                                 if (lcase & KICONV_UPPER && ud != toupper(ud)) {
 207                                         table[us] |= (u_char)toupper(ud) << 16;
 208                                         table[us] |= XLAT16_HAS_UPPER_CASE;
 209                                 }
 210                                 break;
 211                         }
 212
 213                         switch(inbytesleft) {
 214                         case 0:
 215                                 if ((ls & 0xff) == 0)
 216                                         table[us] |= XLAT16_ACCEPT_NULL_IN;
 217                                 break;
 218                         case 1:
 219                                 c = ls > 0xff ? us | 0x80 : us;
 220                                 if (lcase & KICONV_FROM_LOWER && c != tolower(c)) {
 221                                         table[us] |= (u_char)tolower(c) << 16;
 222                                         table[us] |= XLAT16_HAS_FROM_LOWER_CASE;
 223                                 }
 224                                 if (lcase & KICONV_FROM_UPPER && c != toupper(c)) {
 225                                         table[us] |= (u_char)toupper(c) << 16;
 226                                         table[us] |= XLAT16_HAS_FROM_UPPER_CASE;
 227                                 }
 228                                 break;
 229                         }
 230
 231                         if (table[us] == 0)
 232                                 continue;
 233
 234                         /*
 235                          * store not NULL
 236                          */
 237                         xt.idx[ls] = table;
 238                 }
 239                 if (xt.idx[ls]) {
 240                         memcpy(p, table, sizeof(table));
 241                         p += sizeof(table);
 242                 }
 243         }
 244         my_iconv_close(cd);
 245
 246         xt.size = p - (char *)xt.data;
 247         xt.data = realloc(xt.data, xt.size);
 248         return (xt);
 249 }
 250
 251 static int
 252 my_iconv_init(void)
 253 {
 254         void *iconv_lib;
 255 #ifdef __PIC__
 256         iconv_lib = dlopen("libc.so", RTLD_LAZY | RTLD_GLOBAL);
 257         if (iconv_lib == NULL) {
 258                 warn("Unable to load iconv library: %s\n", dlerror());
 259                 errno = ENOENT;
 260                 return (-1);
 261         }
 262         my_iconv_open = dlsym(iconv_lib, "iconv_open");
 263         my_iconv = dlsym(iconv_lib, "iconv");
 264         my_iconv_close = dlsym(iconv_lib, "iconv_close");
 265 #else
 266         my_iconv_open = iconv_open;
 267         my_iconv = iconv;
 268         my_iconv_close = iconv_close;
 269 #endif
 270
 271         return (0);
 272 }
 273
 274 static size_t
 275 my_iconv_char(iconv_t cd, const u_char **ibuf, size_t * ilen, u_char **obuf,
 276         size_t * olen)
 277 {
 278         const u_char *sp;
 279         u_char *dp, ilocal[3], olocal[3];
 280         u_char c1, c2;
 281         int ret;
 282         size_t ir, or;
 283
 284         sp = *ibuf;
 285         dp = *obuf;
 286         ir = *ilen;
 287
 288         bzero(*obuf, *olen);
 289         ret = my_iconv(cd, (const char **)&sp, ilen, (char **)&dp, olen);
 290         c1 = (*obuf)[0];
 291         c2 = (*obuf)[1];
 292
 293         if (ret == -1) {
 294                 if (*ilen == ir - 1 && (*ibuf)[1] == '\0' && (c1 || c2))
 295                         return (0);
 296                 else
 297                         return (-1);
 298         }
 299
 300         /*
 301          * We must judge if inbuf is a single byte char or double byte char.
 302          * Here, to judge, try first byte(*sp) conversion and compare.
 303          */
 304         ir = 1;
 305         or = 3;
 306
 307         bzero(olocal, or);
 308         memcpy(ilocal, *ibuf, sizeof(ilocal));
 309         sp = ilocal;
 310         dp = olocal;
 311
 312         if ((my_iconv(cd,(const char **)&sp, &ir, (char **)&dp, &or)) != -1) {
 313                 if (olocal[0] != c1)
 314                         return (ret);
 315
 316                 if (olocal[1] == c2 && (*ibuf)[1] == '\0') {
 317                         /*
 318                          * inbuf is a single byte char
 319                          */
 320                         *ilen = 1;
 321                         *olen = or;
 322                         return (ret);
 323                 }
 324
 325                 switch(or) {
 326                 case 0:
 327                 case 1:
 328                         if (olocal[1] == c2) {
 329                                 /*
 330                                  * inbuf is a single byte char,
 331                                  * so return false here.
 332                                  */
 333                                 return (-1);
 334                         } else {
 335                                 /*
 336                                  * inbuf is a double byte char
 337                                  */
 338                                 return (ret);
 339                         }
 340                         break;
 341                 case 2:
 342                         /*
 343                          * should compare second byte of inbuf
 344                          */
 345                         break;
 346                 }
 347         } else {
 348                 /*
 349                  * inbuf clould not be splitted, so inbuf is
 350                  * a double byte char.
 351                  */
 352                 return (ret);
 353         }
 354
 355         /*
 356          * try second byte(*(sp+1)) conversion, and compare
 357          */
 358         ir = 1;
 359         or = 3;
 360
 361         bzero(olocal, or);
 362
 363         sp = ilocal + 1;
 364         dp = olocal;
 365
 366         if ((my_iconv(cd,(const char **)&sp, &ir, (char **)&dp, &or)) != -1) {
 367                 if (olocal[0] == c2)
 368                         /*
 369                          * inbuf is a single byte char
 370                          */
 371                         return (-1);
 372         }
 373
 374         return (ret);
 375 }