sys/libiconv/iconv_xlat16.c

   1 /*-
   2  * Copyright (c) 2003, 2005 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: head/sys/libkern/iconv_xlat16.c 194638 2009-06-22 17:09:46Z delphij $
  27  */
  28
  29 #include <sys/param.h>
  30 #include <sys/kernel.h>
  31 #include <sys/systm.h>
  32 #include <sys/malloc.h>
  33 #include <sys/iconv.h>
  34
  35 #include "iconv_converter_if.h"
  36
  37 /*
  38  * "XLAT16" converter
  39  */
  40
  41 #ifdef MODULE_DEPEND
  42 MODULE_DEPEND(iconv_xlat16, libiconv, 2, 2, 2);
  43 #endif
  44
  45 #define C2I1(c) ((c) & 0x8000 ? ((c) & 0xff) | 0x100 : (c) & 0xff)
  46 #define C2I2(c) ((c) & 0x8000 ? ((c) >> 8) & 0x7f : ((c) >> 8) & 0xff)
  47
  48 /*
  49  * XLAT16 converter instance
  50  */
  51 struct iconv_xlat16 {
  52         KOBJ_FIELDS;
  53         uint32_t *              d_table[0x200];
  54         void *                  f_ctp;
  55         void *                  t_ctp;
  56         struct iconv_cspair *   d_csp;
  57 };
  58
  59 static int
  60 iconv_xlat16_open(struct iconv_converter_class *dcp,
  61         struct iconv_cspair *csp, struct iconv_cspair *cspf, void **dpp)
  62 {
  63         struct iconv_xlat16 *dp;
  64         uint32_t *headp, **idxp;
  65         int i;
  66
  67         dp = (struct iconv_xlat16 *)kobj_create((struct kobj_class*)dcp, M_ICONV, M_WAITOK);
  68         headp = (uint32_t *)((caddr_t)csp->cp_data + sizeof(dp->d_table));
  69         idxp = (uint32_t **)csp->cp_data;
  70         for (i = 0 ; i < 0x200 ; i++) {
  71                 if (*idxp) {
  72                         dp->d_table[i] = headp;
  73                         headp += 0x80;
  74                 } else {
  75                         dp->d_table[i] = NULL;
  76                 }
  77                 idxp++;
  78         }
  79
  80         if (strcmp(csp->cp_to, KICONV_WCTYPE_NAME) != 0) {
  81                 if (iconv_open(KICONV_WCTYPE_NAME, csp->cp_from, &dp->f_ctp) != 0)
  82                         dp->f_ctp = NULL;
  83                 if (iconv_open(KICONV_WCTYPE_NAME, csp->cp_to, &dp->t_ctp) != 0)
  84                         dp->t_ctp = NULL;
  85         } else {
  86                 dp->f_ctp = dp->t_ctp = dp;
  87         }
  88
  89         dp->d_csp = csp;
  90         csp->cp_refcount++;
  91         *dpp = (void*)dp;
  92         return (0);
  93 }
  94
  95 static int
  96 iconv_xlat16_close(void *data)
  97 {
  98         struct iconv_xlat16 *dp = data;
  99
 100         if (dp->f_ctp && dp->f_ctp != data)
 101                 iconv_close(dp->f_ctp);
 102         if (dp->t_ctp && dp->t_ctp != data)
 103                 iconv_close(dp->t_ctp);
 104         dp->d_csp->cp_refcount--;
 105         kobj_delete((struct kobj*)data, M_ICONV);
 106         return (0);
 107 }
 108
 109 static int
 110 iconv_xlat16_conv(void *d2p, const char **inbuf,
 111         size_t *inbytesleft, char **outbuf, size_t *outbytesleft,
 112         int convchar, int casetype)
 113 {
 114         struct iconv_xlat16 *dp = (struct iconv_xlat16*)d2p;
 115         const char *src;
 116         char *dst;
 117         int nullin, ret = 0;
 118         size_t in, on, ir, or, inlen;
 119         uint32_t code;
 120         u_char u, l;
 121         uint16_t c1, c2, ctmp;
 122
 123         if (inbuf == NULL || *inbuf == NULL || outbuf == NULL || *outbuf == NULL)
 124                 return (0);
 125         ir = in = *inbytesleft;
 126         or = on = *outbytesleft;
 127         src = *inbuf;
 128         dst = *outbuf;
 129
 130         while(ir > 0 && or > 0) {
 131
 132                 inlen = 0;
 133                 code = 0;
 134
 135                 c1 = ir > 1 ? *(src+1) & 0xff : 0;
 136                 c2 = *src & 0xff;
 137                 ctmp = 0;
 138
 139                 c1 = c2 & 0x80 ? c1 | 0x100 : c1;
 140                 c2 = c2 & 0x80 ? c2 & 0x7f : c2;
 141
 142                 if (ir > 1 && dp->d_table[c1] && dp->d_table[c1][c2]) {
 143                         /*
 144                          * inbuf char is a double byte char
 145                          */
 146                         inlen = 2;
 147
 148                         /* toupper,tolower */
 149                         if (casetype == KICONV_FROM_LOWER && dp->f_ctp)
 150                                 ctmp = towlower(((u_char)*src << 8) | (u_char)*(src + 1),
 151                                     dp->f_ctp);
 152                         else if (casetype == KICONV_FROM_UPPER && dp->f_ctp)
 153                                 ctmp = towupper(((u_char)*src << 8) | (u_char)*(src + 1),
 154                                     dp->f_ctp);
 155                         if (ctmp) {
 156                                 c1 = C2I1(ctmp);
 157                                 c2 = C2I2(ctmp);
 158                         }
 159                 }
 160
 161                 if (inlen == 0) {
 162                         c1 &= 0xff00;
 163                         if (!dp->d_table[c1]) {
 164                                 ret = -1;
 165                                 break;
 166                         }
 167                         /*
 168                          * inbuf char is a single byte char
 169                          */
 170                         inlen = 1;
 171
 172                         if (casetype & (KICONV_FROM_LOWER|KICONV_FROM_UPPER))
 173                                 code = dp->d_table[c1][c2];
 174
 175                         if (casetype == KICONV_FROM_LOWER) {
 176                                 if (dp->f_ctp)
 177                                         ctmp = towlower((u_char)*src, dp->f_ctp);
 178                                 else if (code & XLAT16_HAS_FROM_LOWER_CASE)
 179                                         ctmp = (u_char)(code >> 16);
 180                         } else if (casetype == KICONV_FROM_UPPER) {
 181                                 if (dp->f_ctp)
 182                                         ctmp = towupper((u_char)*src, dp->f_ctp);
 183                                 else if (code & XLAT16_HAS_FROM_UPPER_CASE)
 184                                         ctmp = (u_char)(code >> 16);
 185                         }
 186                         if (ctmp) {
 187                                 c1 = C2I1(ctmp << 8);
 188                                 c2 = C2I2(ctmp << 8);
 189                         }
 190                 }
 191
 192                 code = dp->d_table[c1][c2];
 193                 if (!code) {
 194                         ret = -1;
 195                         break;
 196                 }
 197
 198                 nullin = (code & XLAT16_ACCEPT_NULL_IN) ? 1 : 0;
 199                 if (inlen == 1 && nullin) {
 200                         /*
 201                          * XLAT16_ACCEPT_NULL_IN requires inbuf has 2byte
 202                          */
 203                         ret = -1;
 204                         break;
 205                 }
 206
 207                 /*
 208                  * now start translation
 209                  */
 210                 u = (u_char)(code >> 8);
 211                 l = (u_char)code;
 212
 213 #ifdef XLAT16_ACCEPT_3BYTE_CHR
 214                 if (code & XLAT16_IS_3BYTE_CHR) {
 215                         if (or < 3) {
 216                                 ret = -1;
 217                                 break;
 218                         }
 219                         *dst++ = u;
 220                         *dst++ = l;
 221                         *dst++ = (u_char)(code >> 16);
 222                         or -= 3;
 223                 } else
 224 #endif
 225                 if (u || code & XLAT16_ACCEPT_NULL_OUT) {
 226                         if (or < 2) {
 227                                 ret = -1;
 228                                 break;
 229                         }
 230
 231                         /* toupper,tolower */
 232                         if (casetype == KICONV_LOWER && dp->t_ctp) {
 233                                 code = towlower((uint16_t)code, dp->t_ctp);
 234                                 u = (u_char)(code >> 8);
 235                                 l = (u_char)code;
 236                         }
 237                         if (casetype == KICONV_UPPER && dp->t_ctp) {
 238                                 code = towupper((uint16_t)code, dp->t_ctp);
 239                                 u = (u_char)(code >> 8);
 240                                 l = (u_char)code;
 241                         }
 242
 243                         *dst++ = u;
 244                         *dst++ = l;
 245                         or -= 2;
 246                 } else {
 247                         /* toupper,tolower */
 248                         if (casetype == KICONV_LOWER) {
 249                                 if (dp->t_ctp)
 250                                         l = (u_char)towlower(l, dp->t_ctp);
 251                                 else if (code & XLAT16_HAS_LOWER_CASE)
 252                                         l = (u_char)(code >> 16);
 253                         }
 254                         if (casetype == KICONV_UPPER) {
 255                                 if (dp->t_ctp)
 256                                         l = (u_char)towupper(l, dp->t_ctp);
 257                                 else if (code & XLAT16_HAS_UPPER_CASE)
 258                                         l = (u_char)(code >> 16);
 259                         }
 260
 261                         *dst++ = l;
 262                         or--;
 263                 }
 264
 265                 if (inlen == 2) {
 266                         /*
 267                          * there is a case that inbuf char is a single
 268                          * byte char while inlen == 2
 269                          */
 270                         if ((u_char)*(src+1) == 0 && !nullin ) {
 271                                 src++;
 272                                 ir--;
 273                         } else {
 274                                 src += 2;
 275                                 ir -= 2;
 276                         }
 277                 } else {
 278                         src++;
 279                         ir--;
 280                 }
 281
 282                 if (convchar == 1)
 283                         break;
 284         }
 285
 286         *inbuf += in - ir;
 287         *outbuf += on - or;
 288         *inbytesleft -= in - ir;
 289         *outbytesleft -= on - or;
 290         return (ret);
 291 }
 292
 293 static const char *
 294 iconv_xlat16_name(struct iconv_converter_class *dcp)
 295 {
 296         return ("xlat16");
 297 }
 298
 299 static int
 300 iconv_xlat16_tolower(void *d2p, int c)
 301 {
 302         struct iconv_xlat16 *dp = (struct iconv_xlat16*)d2p;
 303         int c1, c2, out;
 304
 305         if (c < 0x100) {
 306                 c1 = C2I1(c << 8);
 307                 c2 = C2I2(c << 8);
 308         } else if (c < 0x10000) {
 309                 c1 = C2I1(c);
 310                 c2 = C2I2(c);
 311         } else
 312                 return (c);
 313
 314         if (dp->d_table[c1] && dp->d_table[c1][c2] & XLAT16_HAS_LOWER_CASE) {
 315                 /*return (int)(dp->d_table[c1][c2] & 0xffff);*/
 316                 out = dp->d_table[c1][c2] & 0xffff;
 317                 if ((out & 0xff) == 0)
 318                         out = (out >> 8) & 0xff;
 319                 return (out);
 320         } else
 321                 return (c);
 322 }
 323
 324 static int
 325 iconv_xlat16_toupper(void *d2p, int c)
 326 {
 327         struct iconv_xlat16 *dp = (struct iconv_xlat16*)d2p;
 328         int c1, c2, out;
 329
 330         if (c < 0x100) {
 331                 c1 = C2I1(c << 8);
 332                 c2 = C2I2(c << 8);
 333         } else if (c < 0x10000) {
 334                 c1 = C2I1(c);
 335                 c2 = C2I2(c);
 336         } else
 337                 return (c);
 338
 339         if (dp->d_table[c1] && dp->d_table[c1][c2] & XLAT16_HAS_UPPER_CASE) {
 340                 out = dp->d_table[c1][c2] & 0xffff;
 341                 if ((out & 0xff) == 0)
 342                         out = (out >> 8) & 0xff;
 343                 return (out);
 344         } else
 345                 return (c);
 346 }
 347
 348 static kobj_method_t iconv_xlat16_methods[] = {
 349         KOBJMETHOD(iconv_converter_open,        iconv_xlat16_open),
 350         KOBJMETHOD(iconv_converter_close,       iconv_xlat16_close),
 351         KOBJMETHOD(iconv_converter_conv,        iconv_xlat16_conv),
 352 #if 0
 353         KOBJMETHOD(iconv_converter_init,        iconv_xlat16_init),
 354         KOBJMETHOD(iconv_converter_done,        iconv_xlat16_done),
 355 #endif
 356         KOBJMETHOD(iconv_converter_name,        iconv_xlat16_name),
 357         KOBJMETHOD(iconv_converter_tolower,     iconv_xlat16_tolower),
 358         KOBJMETHOD(iconv_converter_toupper,     iconv_xlat16_toupper),
 359         KOBJMETHOD_END
 360 };
 361
 362 KICONV_CONVERTER(xlat16, sizeof(struct iconv_xlat16));