camel/camel-utf8.c

   1 /* -*- Mode: C; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
   2 /*
   3  *  Authors: Michael Zucchi <notzed@ximian.com>
   4  *
   5  *  Copyright 2003 Ximian, Inc. (www.ximian.com)
   6  *
   7  *  This program is free software; you can redistribute it and/or modify
   8  *  it under the terms of the GNU Lesser General Public License as published by
   9  *  the Free Software Foundation; either version 2 of the License, or
  10  *  (at your option) any later version.
  11  *
  12  *  This program is distributed in the hope that it will be useful,
  13  *  but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15  *  GNU Lesser General Public License for more details.
  16  *
  17  *  You should have received a copy of the GNU Lesser General Public License
  18  *  along with this program; if not, write to the Free Software
  19  *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  *
  21  */
  22
  23
  24 #ifdef HAVE_CONFIG_H
  25 #include <config.h>
  26 #endif
  27
  28 #include <string.h>
  29 #include <sys/types.h>
  30
  31 #include <glib.h>
  32
  33 #include "camel-utf8.h"
  34
  35 /**
  36  * camel_utf8_putc:
  37  * @ptr:
  38  * @c:
  39  *
  40  * Output a 32 bit unicode character as utf8 octets.  At most 4 octets will
  41  * be written to @ptr.  @ptr will be advanced to the next character position.
  42  **/
  43 void
  44 camel_utf8_putc(unsigned char **ptr, guint32 c)
  45 {
  46         register unsigned char *p = *ptr;
  47
  48         if (c <= 0x7f)
  49                 *p++ = c;
  50         else if (c <= 0x7ff) {
  51                 *p++ = 0xc0 | c >> 6;
  52                 *p++ = 0x80 | (c & 0x3f);
  53         } else if (c <= 0xffff) {
  54                 *p++ = 0xe0 | c >> 12;
  55                 *p++ = 0x80 | ((c >> 6) & 0x3f);
  56                 *p++ = 0x80 | (c & 0x3f);
  57         } else {
  58                 /* see unicode standard 3.0, S 3.8, max 4 octets */
  59                 *p++ = 0xf0 | c >> 18;
  60                 *p++ = 0x80 | ((c >> 12) & 0x3f);
  61                 *p++ = 0x80 | ((c >> 6) & 0x3f);
  62                 *p++ = 0x80 | (c & 0x3f);
  63         }
  64
  65         *ptr = p;
  66 }
  67
  68 /**
  69  * camel_utf8_getc:
  70  * @ptr:
  71  *
  72  * Get a Unicode character from a utf8 stream.  @ptr will be advanced
  73  * to the next character position.  Invalid utf8 characters will be
  74  * silently skipped.  @ptr should point to a NUL terminated array.
  75  *
  76  * Return value: The next Unicode character.  @ptr will be advanced to
  77  * the next character always.
  78  **/
  79 guint32
  80 camel_utf8_getc(const unsigned char **ptr)
  81 {
  82         register unsigned char *p = (unsigned char *)*ptr;
  83         register unsigned char c, r;
  84         register guint32 v, m;
  85
  86 again:
  87         r = *p++;
  88 loop:
  89         if (r < 0x80) {
  90                 *ptr = p;
  91                 v = r;
  92         } else if (r < 0xf8) { /* valid start char? (max 4 octets) */
  93                 v = r;
  94                 m = 0x7f80;     /* used to mask out the length bits */
  95                 do {
  96                         c = *p++;
  97                         if ((c & 0xc0) != 0x80) {
  98                                 r = c;
  99                                 goto loop;
 100                         }
 101                         v = (v<<6) | (c & 0x3f);
 102                         r<<=1;
 103                         m<<=5;
 104                 } while (r & 0x40);
 105
 106                 *ptr = p;
 107
 108                 v &= ~m;
 109         } else {
 110                 goto again;
 111         }
 112
 113         return v;
 114 }
 115
 116 /**
 117  * camel_utf8_getc_limit:
 118  * @ptr:
 119  * @end: must not be NULL.
 120  *
 121  * Get the next utf8 char at @ptr, and return it, advancing @ptr to
 122  * the next character.  If @end is reached before a full utf8
 123  * character can be read, then the invalid Unicode char 0xffff is
 124  * returned as a sentinel (Unicode 3.1, section 2.7), and @ptr is not
 125  * advanced.
 126  *
 127  * Return value: The next utf8 char, or 0xffff.
 128  **/
 129 guint32
 130 camel_utf8_getc_limit(const unsigned char **ptr, const unsigned char *end)
 131 {
 132         register unsigned char *p = (unsigned char *)*ptr;
 133         register unsigned char c, r;
 134         register guint32 v = 0xffff, m;
 135
 136 again:
 137         while (p < end) {
 138                 r = *p++;
 139 loop:
 140                 if (r < 0x80) {
 141                         *ptr = p;
 142                         return r;
 143                 } else if (r < 0xf8) { /* valid start char? (max 4 octets) */
 144                         v = r;
 145                         m = 0x7f80;     /* used to mask out the length bits */
 146                         do {
 147                                 if (p >= end)
 148                                         return 0xffff;
 149
 150                                 c = *p++;
 151                                 if ((c & 0xc0) != 0x80) {
 152                                         r = c;
 153                                         goto loop;
 154                                 }
 155                                 v = (v<<6) | (c & 0x3f);
 156                                 r<<=1;
 157                                 m<<=5;
 158                         } while (r & 0x40);
 159
 160                         *ptr = p;
 161
 162                         v &= ~m;
 163                         return v;
 164                 } else {
 165                         goto again;
 166                 }
 167         }
 168
 169         return 0xffff;
 170 }
 171
 172 void
 173 g_string_append_u(GString *out, guint32 c)
 174 {
 175         unsigned char buffer[8];
 176         unsigned char *p = buffer;
 177
 178         camel_utf8_putc(&p, c);
 179         *p = 0;
 180         g_string_append(out, (gchar *) buffer);
 181 }
 182
 183 static const char utf7_alphabet[] =
 184         "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+,";
 185
 186 static const unsigned char utf7_rank[256] = {
 187         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 188         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 189         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0x3e,0x3f,0xff,0xff,0xff,
 190         0x34,0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0xff,0xff,0xff,0xff,0xff,0xff,
 191         0xff,0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,
 192         0x0f,0x10,0x11,0x12,0x13,0x14,0x15,0x16,0x17,0x18,0x19,0xff,0xff,0xff,0xff,0xff,
 193         0xff,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,0x20,0x21,0x22,0x23,0x24,0x25,0x26,0x27,0x28,
 194         0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,0x30,0x31,0x32,0x33,0xff,0xff,0xff,0xff,0xff,
 195         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 196         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 197         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 198         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 199         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 200         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 201         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 202         0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,0xff,
 203 };
 204
 205 /**
 206  * camel_utf7_utf8:
 207  * @ptr:
 208  *
 209  * Convert a modified utf7 string to utf8.  If the utf7 string
 210  * contains 8 bit characters, they are treated as iso-8859-1.
 211  *
 212  * The IMAP rules [rfc2060] are used in the utf7 encoding.
 213  *
 214  * Return value: The converted string.
 215  **/
 216 char *
 217 camel_utf7_utf8(const char *ptr)
 218 {
 219         const unsigned char *p = (unsigned char *)ptr;
 220         unsigned int c;
 221         guint32 v=0, x;
 222         GString *out;
 223         int i=0;
 224         int state = 0;
 225         char *ret;
 226
 227         out = g_string_new("");
 228         do {
 229                 c = *p++;
 230                 switch(state) {
 231                 case 0:
 232                         if (c == '&')
 233                                 state = 1;
 234                         else
 235                                 g_string_append_u(out, c);
 236                         break;
 237                 case 1:
 238                         if (c == '-') {
 239                                 g_string_append_c(out, '&');
 240                                 state = 0;
 241                         } else if (utf7_rank[c] != 0xff) {
 242                                 v = utf7_rank[c];
 243                                 i = 6;
 244                                 state = 2;
 245                         } else {
 246                                 /* invalid */
 247                                 g_string_append(out, "&-");
 248                                 state = 0;
 249                         }
 250                         break;
 251                 case 2:
 252                         if (c == '-') {
 253                                 state = 0;
 254                         } else if (utf7_rank[c] != 0xff) {
 255                                 v = (v<<6) | utf7_rank[c];
 256                                 i+=6;
 257                                 if (i >= 16) {
 258                                         x = (v >> (i-16)) & 0xffff;
 259                                         g_string_append_u(out, x);
 260                                         i-=16;
 261                                 }
 262                         } else {
 263                                 g_string_append_u(out, c);
 264                                 state = 0;
 265                         }
 266                         break;
 267                 }
 268         } while (c);
 269
 270         ret = g_strdup(out->str);
 271         g_string_free(out, TRUE);
 272
 273         return ret;
 274 }
 275
 276 static void utf7_closeb64(GString *out, guint32 v, guint32 i)
 277 {
 278         guint32 x;
 279
 280         if (i>0) {
 281                 x = (v << (6-i)) & 0x3f;
 282                 g_string_append_c(out, utf7_alphabet[x]);
 283         }
 284         g_string_append_c(out, '-');
 285 }
 286
 287 /**
 288  * camel_utf8_utf7:
 289  * @ptr:
 290  *
 291  * Convert a utf8 string to a modified utf7 format.
 292  *
 293  * The IMAP rules [rfc2060] are used in the utf7 encoding.
 294  *
 295  * Return value:
 296  **/
 297 char *
 298 camel_utf8_utf7(const char *ptr)
 299 {
 300         const unsigned char *p = (unsigned char *)ptr;
 301         unsigned int c;
 302         guint32 x, v = 0;
 303         int state = 0;
 304         GString *out;
 305         int i = 0;
 306         char *ret;
 307
 308         out = g_string_new("");
 309
 310         while ( (c = camel_utf8_getc(&p)) ) {
 311                 if (c >= 0x20 && c <= 0x7e) {
 312                         if (state == 1) {
 313                                 utf7_closeb64(out, v, i);
 314                                 state = 0;
 315                                 i = 0;
 316                         }
 317                         if (c == '&')
 318                                 g_string_append(out, "&-");
 319                         else
 320                                 g_string_append_c(out, c);
 321                 } else {
 322                         if (state == 0) {
 323                                 g_string_append_c(out, '&');
 324                                 state = 1;
 325                         }
 326                         v = (v << 16) | c;
 327                         i += 16;
 328                         while (i >= 6) {
 329                                 x = (v >> (i-6)) & 0x3f;
 330                                 g_string_append_c(out, utf7_alphabet[x]);
 331                                 i -= 6;
 332                         }
 333                 }
 334         }
 335
 336         if (state == 1)
 337                 utf7_closeb64(out, v, i);
 338
 339         ret = g_strdup(out->str);
 340         g_string_free(out, TRUE);
 341
 342         return ret;
 343 }
 344
 345 /**
 346  * camel_utf8_ucs2:
 347  * @ptr:
 348  *
 349  * Convert a utf8 string into a ucs2 one.  The ucs string will be in
 350  * network byte order, and terminated with a 16 bit NULL.
 351  *
 352  * Return value:
 353  **/
 354 char *
 355 camel_utf8_ucs2(const char *ptr)
 356 {
 357         GByteArray *work = g_byte_array_new();
 358         guint32 c;
 359         char *out;
 360
 361         /* what if c is > 0xffff ? */
 362
 363         while ( (c = camel_utf8_getc(&ptr)) ) {
 364                 guint16 s = g_htons(c);
 365
 366                 g_byte_array_append(work, (unsigned char *) &s, 2);
 367         }
 368
 369         g_byte_array_append(work, (unsigned char *) "\000\000", 2);
 370         out = g_malloc(work->len);
 371         memcpy(out, work->data, work->len);
 372         g_byte_array_free(work, TRUE);
 373
 374         return out;
 375 }
 376
 377 /**
 378  * camel_ucs2_utf8:
 379  * @ptr:
 380  *
 381  * Convert a ucs2 string into a utf8 one.  The ucs2 string is treated
 382  * as network byte ordered, and terminated with a 16 bit NUL.
 383  *
 384  * Return value:
 385  **/
 386 char *camel_ucs2_utf8(const char *ptr)
 387 {
 388         guint16 *ucs = (guint16 *)ptr;
 389         guint32 c;
 390         GString *work = g_string_new("");
 391         char *out;
 392
 393         while ( (c = *ucs++) )
 394                 g_string_append_u(work, g_ntohs(c));
 395
 396         out = g_strdup(work->str);
 397         g_string_free(work, TRUE);
 398
 399         return out;
 400 }