eglib/src/gunicode.c

   1 /*
   2  * gunicode.c: Some Unicode routines
   3  *
   4  * Author:
   5  *   Miguel de Icaza (miguel@novell.com)
   6  *
   7  * (C) 2006 Novell, Inc.
   8  *
   9  * utf8 validation code came from:
  10  *      libxml2-2.6.26 licensed under the MIT X11 license
  11  *
  12  * Authors credit in libxml's string.c:
  13  *   William Brack <wbrack@mmm.com.hk>
  14  *   daniel@veillard.com
  15  *
  16  * Permission is hereby granted, free of charge, to any person obtaining
  17  * a copy of this software and associated documentation files (the
  18  * "Software"), to deal in the Software without restriction, including
  19  * without limitation the rights to use, copy, modify, merge, publish,
  20  * distribute, sublicense, and/or sell copies of the Software, and to
  21  * permit persons to whom the Software is furnished to do so, subject to
  22  * the following conditions:
  23  *
  24  * The above copyright notice and this permission notice shall be
  25  * included in all copies or substantial portions of the Software.
  26  *
  27  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
  28  * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
  29  * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
  30  * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
  31  * LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
  32  * OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
  33  * WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
  34  *
  35  */
  36 #include <config.h>
  37 #include <stdio.h>
  38 #include <glib.h>
  39 #include <unicode-data.h>
  40 #include <errno.h>
  41 #ifdef _MSC_VER
  42 /* FIXME */
  43 #define CODESET 1
  44 #include <Windows.h>
  45
  46 typedef int iconv_t;
  47 #else
  48 #ifdef HAVE_LANGINFO_H
  49 #include <langinfo.h>
  50 #endif
  51 #ifdef HAVE_ICONV_H
  52 #include <iconv.h>
  53 #endif
  54 #endif
  55
  56 static char *my_charset;
  57 static gboolean is_utf8;
  58
  59 /*
  60  * Character set conversion
  61  */
  62 /*
  63 * Index into the table below with the first byte of a UTF-8 sequence to
  64 * get the number of trailing bytes that are supposed to follow it.
  65 * Note that *legal* UTF-8 values can't have 4 or 5-bytes. The table is
  66 * left as-is for anyone who may want to do such conversion, which was
  67 * allowed in earlier algorithms.
  68 */
  69 const gchar g_trailingBytesForUTF8 [256] = {
  70         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
  71         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
  72         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
  73         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
  74         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
  75         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
  76         1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
  77         2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,0,0
  78 };
  79
  80 /*
  81 * Magic values subtracted from a buffer value during UTF8 conversion.
  82 * This table contains as many values as there might be trailing bytes
  83 * in a UTF-8 sequence.
  84 */
  85 static const gulong offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL,
  86 0x03C82080UL, 0xFA082080UL, 0x82082080UL };
  87
  88 GUnicodeType
  89 g_unichar_type (gunichar c)
  90 {
  91 int i;
  92
  93         guint16 cp = (guint16) c;
  94         for (i = 0; i < unicode_category_ranges_count; i++) {
  95                 if (cp < unicode_category_ranges [i].start)
  96                         continue;
  97                 if (unicode_category_ranges [i].end <= cp)
  98                         continue;
  99                 return unicode_category [i] [cp - unicode_category_ranges [i].start];
 100         }
 101
 102         /*
 103         // 3400-4DB5: OtherLetter
 104         // 4E00-9FC3: OtherLetter
 105         // AC00-D7A3: OtherLetter
 106         // D800-DFFF: OtherSurrogate
 107         // E000-F8FF: OtherPrivateUse
 108         // 20000-2A6D6 OtherLetter
 109         // F0000-FFFFD OtherPrivateUse
 110         // 100000-10FFFD OtherPrivateUse
 111         */
 112         if (0x3400 <= cp && cp < 0x4DB5)
 113                 return G_UNICODE_OTHER_LETTER;
 114         if (0x4E00 <= cp && cp < 0x9FC3)
 115                 return G_UNICODE_OTHER_LETTER;
 116         if (0xAC00<= cp && cp < 0xD7A3)
 117                 return G_UNICODE_OTHER_LETTER;
 118         if (0xD800 <= cp && cp < 0xDFFF)
 119                 return G_UNICODE_SURROGATE;
 120         if (0xE000 <= cp && cp < 0xF8FF)
 121                 return G_UNICODE_PRIVATE_USE;
 122         /* since the argument is UTF-16, we cannot check beyond FFFF */
 123
 124         /* It should match any of above */
 125         return 0;
 126 }
 127
 128 gunichar
 129 g_unichar_case (gunichar c, gboolean upper)
 130 {
 131         gint8 i, i2;
 132         guint32 cp = (guint32) c, v;
 133
 134         for (i = 0; i < simple_case_map_ranges_count; i++) {
 135                 if (cp < simple_case_map_ranges [i].start)
 136                         return c;
 137                 if (simple_case_map_ranges [i].end <= cp)
 138                         continue;
 139                 if (c < 0x10000) {
 140                         const guint16 *tab = upper ? simple_upper_case_mapping_lowarea [i] : simple_lower_case_mapping_lowarea [i];
 141                         v = tab [cp - simple_case_map_ranges [i].start];
 142                 } else {
 143                         const guint32 *tab;
 144                         i2 = (gint8)(i - (upper ? simple_upper_case_mapping_lowarea_table_count : simple_lower_case_mapping_lowarea_table_count));
 145                         tab = upper ? simple_upper_case_mapping_higharea [i2] : simple_lower_case_mapping_higharea [i2];
 146                         v = tab [cp - simple_case_map_ranges [i].start];
 147                 }
 148                 return v != 0 ? (gunichar) v : c;
 149         }
 150         return c;
 151 }
 152
 153 gunichar
 154 g_unichar_toupper (gunichar c)
 155 {
 156         return g_unichar_case (c, TRUE);
 157 }
 158
 159 gunichar
 160 g_unichar_tolower (gunichar c)
 161 {
 162         return g_unichar_case (c, FALSE);
 163 }
 164
 165 gunichar
 166 g_unichar_totitle (gunichar c)
 167 {
 168         guint8 i;
 169         guint32 cp;
 170
 171         cp = (guint32) c;
 172         for (i = 0; i < simple_titlecase_mapping_count; i++) {
 173                 if (simple_titlecase_mapping [i].codepoint == cp)
 174                         return simple_titlecase_mapping [i].title;
 175                 if (simple_titlecase_mapping [i].codepoint > cp)
 176                         /* it is ordered, hence no more match */
 177                         break;
 178         }
 179         return g_unichar_toupper (c);
 180 }
 181
 182 gboolean
 183 g_unichar_isxdigit (gunichar c)
 184 {
 185         return (g_unichar_xdigit_value (c) != -1);
 186
 187 }
 188
 189 gint
 190 g_unichar_xdigit_value (gunichar c)
 191 {
 192         if (c >= 0x30 && c <= 0x39) /*0-9*/
 193                 return (c - 0x30);
 194         if (c >= 0x41 && c <= 0x46) /*A-F*/
 195                 return (c - 0x37);
 196         if (c >= 0x61 && c <= 0x66) /*a-f*/
 197                 return (c - 0x57);
 198         return -1;
 199 }
 200
 201 gchar *
 202 g_convert (const gchar *str, gssize len,
 203            const gchar *to_codeset, const gchar *from_codeset,
 204            gsize *bytes_read, gsize *bytes_written, GError **error)
 205 {
 206         char *result = NULL;
 207 #ifdef G_OS_WIN32
 208 #elif HAVE_ICONV_H
 209         iconv_t convertor;
 210         char *buffer, *output;
 211         const char *strptr = (const char *) str;
 212         size_t str_len = len == -1 ? strlen (str) : len;
 213         size_t buffer_size;
 214         size_t left, out_left;
 215
 216         convertor = iconv_open (to_codeset, from_codeset);
 217         if (convertor == (iconv_t) -1){
 218                 *bytes_written = 0;
 219                 *bytes_read = 0;
 220                 return NULL;
 221         }
 222
 223         buffer_size = str_len + 1 + 8;
 224         buffer = g_malloc (buffer_size);
 225         out_left = str_len;
 226         output = buffer;
 227         left = str_len;
 228         while (left > 0){
 229                 int res = iconv (convertor, (char **) &strptr, &left, &output, &out_left);
 230                 if (res == (size_t) -1){
 231                         if (errno == E2BIG){
 232                                 char *n;
 233                                 size_t extra_space = 8 + left;
 234                                 size_t output_used = output - buffer;
 235
 236                                 buffer_size += extra_space;
 237
 238                                 n = g_realloc (buffer, buffer_size);
 239
 240                                 if (n == NULL){
 241                                         if (error != NULL)
 242                                                 *error = g_error_new (NULL, G_CONVERT_ERROR_FAILED, "No memory left");
 243                                         g_free (buffer);
 244                                         result = NULL;
 245                                         goto leave;
 246                                 }
 247                                 buffer = n;
 248                                 out_left += extra_space;
 249                                 output = buffer + output_used;
 250                         } else if (errno == EILSEQ){
 251                                 if (error != NULL)
 252                                         *error = g_error_new (NULL, G_CONVERT_ERROR_ILLEGAL_SEQUENCE, "Invalid multi-byte sequence on input");
 253                                 result = NULL;
 254                                 g_free (buffer);
 255                                 goto leave;
 256                         } else if (errno == EINVAL){
 257                                 if (error != NULL)
 258                                         *error = g_error_new (NULL, G_CONVERT_ERROR_PARTIAL_INPUT, "Partial character sequence");
 259                                 result = NULL;
 260                                 g_free (buffer);
 261                                 goto leave;
 262                         }
 263                 }
 264         }
 265         if (bytes_read != NULL)
 266                 *bytes_read = strptr - str;
 267         if (bytes_written != NULL)
 268                 *bytes_written = output - buffer;
 269         *output = 0;
 270         result = buffer;
 271  leave:
 272         iconv_close (convertor);
 273 #endif
 274         return result;
 275 }
 276
 277 /*
 278  * This is broken, and assumes an UTF8 system, but will do for eglib's first user
 279  */
 280 gchar *
 281 g_filename_from_utf8 (const gchar *utf8string, gssize len, gsize *bytes_read, gsize *bytes_written, GError **error)
 282 {
 283         char *res;
 284
 285         if (len == -1)
 286                 len = strlen (utf8string);
 287
 288         res = g_malloc (len + 1);
 289         g_strlcpy (res, utf8string, len + 1);
 290         return res;
 291 }
 292
 293 gboolean
 294 g_get_charset (G_CONST_RETURN char **charset)
 295 {
 296 #ifdef G_OS_WIN32
 297         static char buf[14];
 298         sprintf (buf, "CP%u", GetACP ());
 299         *charset = buf;
 300         is_utf8 = FALSE;
 301 #else
 302         if (my_charset == NULL){
 303 #if HAVE_LANGINFO_H
 304                 my_charset = g_strdup (nl_langinfo (CODESET));
 305 #else
 306                 my_charset = g_strdup ("UTF-8");
 307 #endif
 308                 is_utf8 = strcmp (my_charset, "UTF-8") == 0;
 309         }
 310
 311         if (charset != NULL)
 312                 *charset = my_charset;
 313
 314 #endif
 315         return is_utf8;
 316 }
 317
 318 gchar *
 319 g_locale_to_utf8 (const gchar *opsysstring, gssize len, gsize *bytes_read, gsize *bytes_written, GError **error)
 320 {
 321         g_get_charset (NULL);
 322
 323         return g_convert (opsysstring, len, "UTF-8", my_charset, bytes_read, bytes_written, error);
 324 }
 325
 326 gchar *
 327 g_locale_from_utf8 (const gchar *utf8string, gssize len, gsize *bytes_read, gsize *bytes_written, GError **error)
 328 {
 329         g_get_charset (NULL);
 330
 331         return g_convert (utf8string, len, my_charset, "UTF-8", bytes_read, bytes_written, error);
 332 }
 333 /**
 334  * g_utf8_validate
 335  * @utf: Pointer to putative UTF-8 encoded string.
 336  *
 337  * Checks @utf for being valid UTF-8. @utf is assumed to be
 338  * null-terminated. This function is not super-strict, as it will
 339  * allow longer UTF-8 sequences than necessary. Note that Java is
 340  * capable of producing these sequences if provoked. Also note, this
 341  * routine checks for the 4-byte maximum size, but does not check for
 342  * 0x10ffff maximum value.
 343  *
 344  * Return value: true if @utf is valid.
 345  **/
 346 gboolean
 347 g_utf8_validate (const gchar *str, gssize max_len, const gchar **end)
 348 {
 349         gssize byteCount = 0;
 350         gboolean retVal = TRUE;
 351         gboolean lastRet = TRUE;
 352         guchar* ptr = (guchar*) str;
 353         guint length;
 354         guchar a;
 355         guchar* srcPtr;
 356         if (max_len == 0)
 357                 return 0;
 358         else if (max_len < 0)
 359                 byteCount = max_len;
 360         while (*ptr != 0 && byteCount <= max_len) {
 361                 length = g_trailingBytesForUTF8 [*ptr] + 1;
 362                 srcPtr = (guchar*) ptr + length;
 363                 switch (length) {
 364                 default: retVal = FALSE;
 365                 /* Everything else falls through when "TRUE"... */
 366                 case 4: if ((a = (*--srcPtr)) < (guchar) 0x80 || a > (guchar) 0xBF) retVal = FALSE;
 367                                 if ((a == (guchar) 0xBF || a == (guchar) 0xBE) && *(srcPtr-1) == (guchar) 0xBF) {
 368                                 if (*(srcPtr-2) == (guchar) 0x8F || *(srcPtr-2) == (guchar) 0x9F ||
 369                                         *(srcPtr-2) == (guchar) 0xAF || *(srcPtr-2) == (guchar) 0xBF)
 370                                         retVal = FALSE;
 371                                 }
 372                 case 3: if ((a = (*--srcPtr)) < (guchar) 0x80 || a > (guchar) 0xBF) retVal = FALSE;
 373                 case 2: if ((a = (*--srcPtr)) < (guchar) 0x80 || a > (guchar) 0xBF) retVal = FALSE;
 374
 375                 switch (*ptr) {
 376                 /* no fall-through in this inner switch */
 377                 case 0xE0: if (a < (guchar) 0xA0) retVal = FALSE; break;
 378                 case 0xED: if (a > (guchar) 0x9F) retVal = FALSE; break;
 379                 case 0xEF: if (a == (guchar)0xB7 && (*(srcPtr+1) > (guchar) 0x8F && *(srcPtr+1) < 0xB0)) retVal = FALSE;
 380                                    if (a == (guchar)0xBF && (*(srcPtr+1) == (guchar) 0xBE || *(srcPtr+1) == 0xBF)) retVal = FALSE; break;
 381                 case 0xF0: if (a < (guchar) 0x90) retVal = FALSE; break;
 382                 case 0xF4: if (a > (guchar) 0x8F) retVal = FALSE; break;
 383                 default:   if (a < (guchar) 0x80) retVal = FALSE;
 384                 }
 385
 386                 case 1: if (*ptr >= (guchar ) 0x80 && *ptr < (guchar) 0xC2) retVal = FALSE;
 387                 }
 388                 if (*ptr > (guchar) 0xF4)
 389                         retVal = FALSE;
 390                 //If the string is invalid, set the end to the invalid byte.
 391                 if (!retVal && lastRet) {
 392                         if (end != NULL)
 393                                 *end = (gchar*) ptr;
 394                         lastRet = FALSE;
 395                 }
 396                 ptr += length;
 397                 if(max_len > 0)
 398                         byteCount += length;
 399         }
 400         if (retVal && end != NULL)
 401                 *end = (gchar*) ptr;
 402         return retVal;
 403 }
 404 /**
 405  * g_utf8_get_char
 406  * @src: Pointer to UTF-8 encoded character.
 407  *
 408  * Return value: UTF-16 value of @src
 409  **/
 410 gunichar
 411 g_utf8_get_char (const gchar *src)
 412 {
 413         gunichar ch = 0;
 414         guchar* ptr = (guchar*) src;
 415         gushort extraBytesToRead = g_trailingBytesForUTF8 [*ptr];
 416
 417         switch (extraBytesToRead) {
 418         case 5: ch += *ptr++; ch <<= 6; // remember, illegal UTF-8
 419         case 4: ch += *ptr++; ch <<= 6; // remember, illegal UTF-8
 420         case 3: ch += *ptr++; ch <<= 6;
 421         case 2: ch += *ptr++; ch <<= 6;
 422         case 1: ch += *ptr++; ch <<= 6;
 423         case 0: ch += *ptr;
 424         }
 425         ch -= offsetsFromUTF8 [extraBytesToRead];
 426         return ch;
 427 }
 428 glong
 429 g_utf8_strlen (const gchar *str, gssize max)
 430 {
 431         gssize byteCount = 0;
 432         guchar* ptr = (guchar*) str;
 433         glong length = 0;
 434         if (max == 0)
 435                 return 0;
 436         else if (max < 0)
 437                 byteCount = max;
 438         while (*ptr != 0 && byteCount <= max) {
 439                 gssize cLen = g_trailingBytesForUTF8 [*ptr] + 1;
 440                 if (max > 0 && (byteCount + cLen) > max)
 441                         return length;
 442                 ptr += cLen;
 443                 length++;
 444                 if (max > 0)
 445                         byteCount += cLen;
 446         }
 447         return length;
 448 }