cairo/src/cairo-unicode.c

   1 /* -*- Mode: c; c-basic-offset: 4; indent-tabs-mode: t; tab-width: 8; -*- */
   2 /* cairo - a vector graphics library with display and print output
   3  *
   4  * The code in this file is derived from GLib's gutf8.c and
   5  *   ultimately from libunicode. It is relicensed under the
   6  *   dual LGPL/MPL with permission of the original authors.
   7  *
   8  * Copyright © 1999 Tom Tromey
   9  * Copyright © 2005 Red Hat, Inc
  10  *
  11  * This library is free software; you can redistribute it and/or
  12  * modify it either under the terms of the GNU Lesser General Public
  13  * License version 2.1 as published by the Free Software Foundation
  14  * (the "LGPL") or, at your option, under the terms of the Mozilla
  15  * Public License Version 1.1 (the "MPL"). If you do not alter this
  16  * notice, a recipient may use your version of this file under either
  17  * the MPL or the LGPL.
  18  *
  19  * You should have received a copy of the LGPL along with this library
  20  * in the file COPYING-LGPL-2.1; if not, write to the Free Software
  21  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  22  * You should have received a copy of the MPL along with this library
  23  * in the file COPYING-MPL-1.1
  24  *
  25  * The contents of this file are subject to the Mozilla Public License
  26  * Version 1.1 (the "License"); you may not use this file except in
  27  * compliance with the License. You may obtain a copy of the License at
  28  * http://www.mozilla.org/MPL/
  29  *
  30  * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY
  31  * OF ANY KIND, either express or implied. See the LGPL or the MPL for
  32  * the specific language governing rights and limitations.
  33  *
  34  * The Original Code is the cairo graphics library.
  35  *
  36  * The Initial Developer of the Original Code is Tom Tromey.
  37  *  and Red Hat, Inc.
  38  *
  39  * Contributor(s):
  40  *      Owen Taylor <otaylor@redhat.com>
  41  */
  42
  43 #include "cairoint.h"
  44
  45 #define UTF8_COMPUTE(Char, Mask, Len)                                         \
  46   if (Char < 128)                                                             \
  47     {                                                                         \
  48       Len = 1;                                                                \
  49       Mask = 0x7f;                                                            \
  50     }                                                                         \
  51   else if ((Char & 0xe0) == 0xc0)                                             \
  52     {                                                                         \
  53       Len = 2;                                                                \
  54       Mask = 0x1f;                                                            \
  55     }                                                                         \
  56   else if ((Char & 0xf0) == 0xe0)                                             \
  57     {                                                                         \
  58       Len = 3;                                                                \
  59       Mask = 0x0f;                                                            \
  60     }                                                                         \
  61   else if ((Char & 0xf8) == 0xf0)                                             \
  62     {                                                                         \
  63       Len = 4;                                                                \
  64       Mask = 0x07;                                                            \
  65     }                                                                         \
  66   else if ((Char & 0xfc) == 0xf8)                                             \
  67     {                                                                         \
  68       Len = 5;                                                                \
  69       Mask = 0x03;                                                            \
  70     }                                                                         \
  71   else if ((Char & 0xfe) == 0xfc)                                             \
  72     {                                                                         \
  73       Len = 6;                                                                \
  74       Mask = 0x01;                                                            \
  75     }                                                                         \
  76   else                                                                        \
  77     Len = -1;
  78
  79 #define UTF8_LENGTH(Char)              \
  80   ((Char) < 0x80 ? 1 :                 \
  81    ((Char) < 0x800 ? 2 :               \
  82     ((Char) < 0x10000 ? 3 :            \
  83      ((Char) < 0x200000 ? 4 :          \
  84       ((Char) < 0x4000000 ? 5 : 6)))))
  85
  86 #define UTF8_GET(Result, Chars, Count, Mask, Len)                             \
  87   (Result) = (Chars)[0] & (Mask);                                             \
  88   for ((Count) = 1; (Count) < (Len); ++(Count))                               \
  89     {                                                                         \
  90       if (((Chars)[(Count)] & 0xc0) != 0x80)                                  \
  91         {                                                                     \
  92           (Result) = -1;                                                      \
  93           break;                                                              \
  94         }                                                                     \
  95       (Result) <<= 6;                                                         \
  96       (Result) |= ((Chars)[(Count)] & 0x3f);                                  \
  97     }
  98
  99 #define UNICODE_VALID(Char)                   \
 100     ((Char) < 0x110000 &&                     \
 101      (((Char) & 0xFFFFF800) != 0xD800) &&     \
 102      ((Char) < 0xFDD0 || (Char) > 0xFDEF) &&  \
 103      ((Char) & 0xFFFE) != 0xFFFE)
 104
 105 static const char utf8_skip_data[256] = {
 106     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,
 107     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,
 108     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,
 109     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,
 110     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,
 111     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,
 112     2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
 113     3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,4,4,4,4,4,4,4,4,5,5,5,5,6,6,1,1
 114 };
 115
 116 #define UTF8_NEXT_CHAR(p) ((p) + utf8_skip_data[*(unsigned char *)(p)])
 117
 118 /* Converts a sequence of bytes encoded as UTF-8 to a Unicode character.
 119  * If @p does not point to a valid UTF-8 encoded character, results are
 120  * undefined.
 121  **/
 122 static uint32_t
 123 _utf8_get_char (const unsigned char *p)
 124 {
 125     int i, mask = 0, len;
 126     uint32_t result;
 127     unsigned char c = (unsigned char) *p;
 128
 129     UTF8_COMPUTE (c, mask, len);
 130     if (len == -1)
 131         return (uint32_t)-1;
 132     UTF8_GET (result, p, i, mask, len);
 133
 134     return result;
 135 }
 136
 137 /* Like _utf8_get_char, but take a maximum length
 138  * and return (uint32_t)-2 on incomplete trailing character
 139  */
 140 static uint32_t
 141 _utf8_get_char_extended (const unsigned char *p,
 142                          long                 max_len)
 143 {
 144     int i, len;
 145     uint32_t wc = (unsigned char) *p;
 146
 147     if (wc < 0x80) {
 148         return wc;
 149     } else if (wc < 0xc0) {
 150         return (uint32_t)-1;
 151     } else if (wc < 0xe0) {
 152         len = 2;
 153         wc &= 0x1f;
 154     } else if (wc < 0xf0) {
 155         len = 3;
 156         wc &= 0x0f;
 157     } else if (wc < 0xf8) {
 158         len = 4;
 159         wc &= 0x07;
 160     } else if (wc < 0xfc) {
 161         len = 5;
 162         wc &= 0x03;
 163     } else if (wc < 0xfe) {
 164         len = 6;
 165         wc &= 0x01;
 166     } else {
 167         return (uint32_t)-1;
 168     }
 169
 170     if (max_len >= 0 && len > max_len) {
 171         for (i = 1; i < max_len; i++) {
 172             if ((((unsigned char *)p)[i] & 0xc0) != 0x80)
 173                 return (uint32_t)-1;
 174         }
 175         return (uint32_t)-2;
 176     }
 177
 178     for (i = 1; i < len; ++i) {
 179         uint32_t ch = ((unsigned char *)p)[i];
 180
 181         if ((ch & 0xc0) != 0x80) {
 182             if (ch)
 183                 return (uint32_t)-1;
 184             else
 185                 return (uint32_t)-2;
 186         }
 187
 188         wc <<= 6;
 189         wc |= (ch & 0x3f);
 190     }
 191
 192     if (UTF8_LENGTH(wc) != len)
 193         return (uint32_t)-1;
 194
 195     return wc;
 196 }
 197
 198 /**
 199  * _cairo_utf8_get_char_validated:
 200  * @p: a UTF-8 string
 201  * @unicode: location to store one Unicode character
 202  *
 203  * Decodes the first character of a valid UTF-8 string, and returns
 204  * the number of bytes consumed.
 205  *
 206  * Note that the string should be valid.  Do not use this without
 207  * validating the string first.
 208  *
 209  * Returns: the number of bytes forming the character returned.
 210  **/
 211 int
 212 _cairo_utf8_get_char_validated (const char *p,
 213                                 uint32_t   *unicode)
 214 {
 215     int i, mask = 0, len;
 216     uint32_t result;
 217     unsigned char c = (unsigned char) *p;
 218
 219     UTF8_COMPUTE (c, mask, len);
 220     if (len == -1) {
 221         if (unicode)
 222             *unicode = (uint32_t)-1;
 223         return 1;
 224     }
 225     UTF8_GET (result, p, i, mask, len);
 226
 227     if (unicode)
 228         *unicode = result;
 229     return len;
 230 }
 231
 232 /**
 233  * _cairo_utf8_to_ucs4:
 234  * @str: an UTF-8 string
 235  * @len: length of @str in bytes, or -1 if it is nul-terminated.
 236  *   If @len is supplied and the string has an embedded nul
 237  *   byte, only the portion before the nul byte is converted.
 238  * @result: location to store a pointer to a newly allocated UTF-32
 239  *   string (always native endian), or %NULL. Free with free(). A 0
 240  *   word will be written after the last character.
 241  * @items_written: location to store number of 32-bit words
 242  *   written. (Not including the trailing 0)
 243  *
 244  * Converts a UTF-8 string to UCS-4. UCS-4 is an encoding of Unicode
 245  * with 1 32-bit word per character. The string is validated to
 246  * consist entirely of valid Unicode characters.
 247  *
 248  * Return value: %CAIRO_STATUS_SUCCESS if the entire string was
 249  *   successfully converted. %CAIRO_STATUS_INVALID_STRING if an
 250  *   invalid sequence was found.
 251  **/
 252 cairo_status_t
 253 _cairo_utf8_to_ucs4 (const char *str,
 254                      int         len,
 255                      uint32_t  **result,
 256                      int        *items_written)
 257 {
 258     uint32_t *str32 = NULL;
 259     int n_chars, i;
 260     const unsigned char *in;
 261     const unsigned char * const ustr = (const unsigned char *) str;
 262
 263     in = ustr;
 264     n_chars = 0;
 265     while ((len < 0 || ustr + len - in > 0) && *in)
 266     {
 267         uint32_t wc = _utf8_get_char_extended (in, ustr + len - in);
 268         if (wc & 0x80000000 || !UNICODE_VALID (wc))
 269             return _cairo_error (CAIRO_STATUS_INVALID_STRING);
 270
 271         n_chars++;
 272         if (n_chars == INT_MAX)
 273             return _cairo_error (CAIRO_STATUS_INVALID_STRING);
 274
 275         in = UTF8_NEXT_CHAR (in);
 276     }
 277
 278     if (result) {
 279         str32 = _cairo_malloc_ab (n_chars + 1, sizeof (uint32_t));
 280         if (!str32)
 281             return _cairo_error (CAIRO_STATUS_NO_MEMORY);
 282
 283         in = ustr;
 284         for (i=0; i < n_chars; i++) {
 285             str32[i] = _utf8_get_char (in);
 286             in = UTF8_NEXT_CHAR (in);
 287         }
 288         str32[i] = 0;
 289
 290         *result = str32;
 291     }
 292
 293     if (items_written)
 294         *items_written = n_chars;
 295
 296     return CAIRO_STATUS_SUCCESS;
 297 }
 298
 299 /**
 300  * _cairo_ucs4_to_utf8:
 301  * @unicode: a UCS-4 character
 302  * @utf8: buffer to write utf8 string into. Must have at least 4 bytes
 303  * space available. Or %NULL.
 304  *
 305  * This space left intentionally blank.
 306  *
 307  * Return value: Number of bytes in the utf8 string or 0 if an invalid
 308  * unicode character
 309  **/
 310 int
 311 _cairo_ucs4_to_utf8 (uint32_t  unicode,
 312                      char     *utf8)
 313 {
 314     int bytes;
 315     char *p;
 316
 317     if (unicode < 0x80) {
 318         if (utf8)
 319             *utf8 = unicode;
 320         return 1;
 321     } else if (unicode < 0x800) {
 322         bytes = 2;
 323     } else if (unicode < 0x10000) {
 324         bytes = 3;
 325     } else if (unicode < 0x200000) {
 326         bytes = 4;
 327     } else {
 328         return 0;
 329     }
 330
 331     if (!utf8)
 332         return bytes;
 333
 334     p = utf8 + bytes;
 335     while (p > utf8) {
 336         *--p = 0x80 | (unicode & 0x3f);
 337         unicode >>= 6;
 338     }
 339     *p |= 0xf0 << (4 - bytes);
 340
 341     return bytes;
 342 }
 343
 344 #if CAIRO_HAS_UTF8_TO_UTF16
 345 /**
 346  * _cairo_utf8_to_utf16:
 347  * @str: an UTF-8 string
 348  * @len: length of @str in bytes, or -1 if it is nul-terminated.
 349  *   If @len is supplied and the string has an embedded nul
 350  *   byte, only the portion before the nul byte is converted.
 351  * @result: location to store a pointer to a newly allocated UTF-16
 352  *   string (always native endian). Free with free(). A 0
 353  *   word will be written after the last character.
 354  * @items_written: location to store number of 16-bit words
 355  *   written. (Not including the trailing 0)
 356  *
 357  * Converts a UTF-8 string to UTF-16. UTF-16 is an encoding of Unicode
 358  * where characters are represented either as a single 16-bit word, or
 359  * as a pair of 16-bit "surrogates". The string is validated to
 360  * consist entirely of valid Unicode characters.
 361  *
 362  * Return value: %CAIRO_STATUS_SUCCESS if the entire string was
 363  *   successfully converted. %CAIRO_STATUS_INVALID_STRING if an
 364  *   an invalid sequence was found.
 365  **/
 366 cairo_status_t
 367 _cairo_utf8_to_utf16 (const char *str,
 368                       int         len,
 369                       uint16_t **result,
 370                       int       *items_written)
 371 {
 372     uint16_t *str16 = NULL;
 373     int n16, i;
 374     const unsigned char *in;
 375     const unsigned char * const ustr = (const unsigned char *) str;
 376
 377     in = ustr;
 378     n16 = 0;
 379     while ((len < 0 || ustr + len - in > 0) && *in) {
 380         uint32_t wc = _utf8_get_char_extended (in, ustr + len - in);
 381         if (wc & 0x80000000 || !UNICODE_VALID (wc))
 382             return _cairo_error (CAIRO_STATUS_INVALID_STRING);
 383
 384         if (wc < 0x10000)
 385             n16 += 1;
 386         else
 387             n16 += 2;
 388
 389         if (n16 == INT_MAX - 1 || n16 == INT_MAX)
 390             return _cairo_error (CAIRO_STATUS_INVALID_STRING);
 391
 392         in = UTF8_NEXT_CHAR (in);
 393     }
 394
 395     str16 = _cairo_malloc_ab (n16 + 1, sizeof (uint16_t));
 396     if (!str16)
 397         return _cairo_error (CAIRO_STATUS_NO_MEMORY);
 398
 399     in = ustr;
 400     for (i = 0; i < n16;) {
 401         uint32_t wc = _utf8_get_char (in);
 402
 403         if (wc < 0x10000) {
 404             str16[i++] = wc;
 405         } else {
 406             str16[i++] = (wc - 0x10000) / 0x400 + 0xd800;
 407             str16[i++] = (wc - 0x10000) % 0x400 + 0xdc00;
 408         }
 409
 410         in = UTF8_NEXT_CHAR (in);
 411     }
 412
 413     str16[i] = 0;
 414
 415     *result = str16;
 416     if (items_written)
 417         *items_written = n16;
 418
 419     return CAIRO_STATUS_SUCCESS;
 420 }
 421 #endif