src/pulse/utf8.c

   1 /***
   2   This file is part of PulseAudio.
   3
   4   Copyright 2006 Lennart Poettering
   5   Copyright 2006 Pierre Ossman <ossman@cendio.se> for Cendio AB
   6
   7   PulseAudio is free software; you can redistribute it and/or modify
   8   it under the terms of the GNU Lesser General Public License as
   9   published by the Free Software Foundation; either version 2.1 of the
  10   License, or (at your option) any later version.
  11
  12   PulseAudio is distributed in the hope that it will be useful, but
  13   WITHOUT ANY WARRANTY; without even the implied warranty of
  14   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
  15   Lesser General Public License for more details.
  16
  17   You should have received a copy of the GNU Lesser General Public
  18   License along with PulseAudio; if not, write to the Free Software
  19   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307
  20   USA.
  21 ***/
  22
  23 /* This file is based on the GLIB utf8 validation functions. The
  24  * original license text follows. */
  25
  26 /* gutf8.c - Operations on UTF-8 strings.
  27  *
  28  * Copyright (C) 1999 Tom Tromey
  29  * Copyright (C) 2000 Red Hat, Inc.
  30  *
  31  * This library is free software; you can redistribute it and/or
  32  * modify it under the terms of the GNU Lesser General Public
  33  * License as published by the Free Software Foundation; either
  34  * version 2 of the License, or (at your option) any later version.
  35  *
  36  * This library is distributed in the hope that it will be useful,
  37  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  38  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.         See the GNU
  39  * Lesser General Public License for more details.
  40  *
  41  * You should have received a copy of the GNU Lesser General Public
  42  * License along with this library; if not, write to the
  43  * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  44  * Boston, MA 02111-1307, USA.
  45  */
  46
  47 #ifdef HAVE_CONFIG_H
  48 #include <config.h>
  49 #endif
  50
  51 #include <errno.h>
  52 #include <stdlib.h>
  53 #include <inttypes.h>
  54 #include <string.h>
  55
  56 #ifdef HAVE_ICONV
  57 #include <iconv.h>
  58 #endif
  59
  60 #include <pulse/xmalloc.h>
  61 #include <pulsecore/macro.h>
  62
  63 #include "utf8.h"
  64
  65 #define FILTER_CHAR '_'
  66
  67 static inline pa_bool_t is_unicode_valid(uint32_t ch) {
  68
  69     if (ch >= 0x110000) /* End of unicode space */
  70         return FALSE;
  71     if ((ch & 0xFFFFF800) == 0xD800) /* Reserved area for UTF-16 */
  72         return FALSE;
  73     if ((ch >= 0xFDD0) && (ch <= 0xFDEF)) /* Reserved */
  74         return FALSE;
  75     if ((ch & 0xFFFE) == 0xFFFE) /* BOM (Byte Order Mark) */
  76         return FALSE;
  77
  78     return TRUE;
  79 }
  80
  81 static inline pa_bool_t is_continuation_char(uint8_t ch) {
  82     if ((ch & 0xc0) != 0x80) /* 10xxxxxx */
  83         return FALSE;
  84     return TRUE;
  85 }
  86
  87 static inline void merge_continuation_char(uint32_t *u_ch, uint8_t ch) {
  88     *u_ch <<= 6;
  89     *u_ch |= ch & 0x3f;
  90 }
  91
  92 static char* utf8_validate(const char *str, char *output) {
  93     uint32_t val = 0;
  94     uint32_t min = 0;
  95     const uint8_t *p, *last;
  96     int size;
  97     uint8_t *o;
  98
  99     pa_assert(str);
 100
 101     o = (uint8_t*) output;
 102     for (p = (const uint8_t*) str; *p; p++) {
 103         if (*p < 128) {
 104             if (o)
 105                 *o = *p;
 106         } else {
 107             last = p;
 108
 109             if ((*p & 0xe0) == 0xc0) { /* 110xxxxx two-char seq. */
 110                 size = 2;
 111                 min = 128;
 112                 val = (uint32_t) (*p & 0x1e);
 113                 goto ONE_REMAINING;
 114             } else if ((*p & 0xf0) == 0xe0) { /* 1110xxxx three-char seq.*/
 115                 size = 3;
 116                 min = (1 << 11);
 117                 val = (uint32_t) (*p & 0x0f);
 118                 goto TWO_REMAINING;
 119             } else if ((*p & 0xf8) == 0xf0) { /* 11110xxx four-char seq */
 120                 size = 4;
 121                 min = (1 << 16);
 122                 val = (uint32_t) (*p & 0x07);
 123             } else {
 124                 size = 1;
 125                 goto error;
 126             }
 127
 128             p++;
 129             if (!is_continuation_char(*p))
 130                 goto error;
 131             merge_continuation_char(&val, *p);
 132
 133 TWO_REMAINING:
 134             p++;
 135             if (!is_continuation_char(*p))
 136                 goto error;
 137             merge_continuation_char(&val, *p);
 138
 139 ONE_REMAINING:
 140             p++;
 141             if (!is_continuation_char(*p))
 142                 goto error;
 143             merge_continuation_char(&val, *p);
 144
 145             if (val < min)
 146                 goto error;
 147
 148             if (!is_unicode_valid(val))
 149                 goto error;
 150
 151             if (o) {
 152                 memcpy(o, last, (size_t) size);
 153                 o += size - 1;
 154             }
 155
 156             if (o)
 157                 o++;
 158
 159             continue;
 160
 161 error:
 162             if (o) {
 163                 *o = FILTER_CHAR;
 164                 p = last; /* We retry at the next character */
 165             } else
 166                 goto failure;
 167         }
 168
 169         if (o)
 170             o++;
 171     }
 172
 173     if (o) {
 174         *o = '\0';
 175         return output;
 176     }
 177
 178     return (char*) str;
 179
 180 failure:
 181     return NULL;
 182 }
 183
 184 char* pa_utf8_valid (const char *str) {
 185     return utf8_validate(str, NULL);
 186 }
 187
 188 char* pa_utf8_filter (const char *str) {
 189     char *new_str;
 190
 191     pa_assert(str);
 192     new_str = pa_xmalloc(strlen(str) + 1);
 193     return utf8_validate(str, new_str);
 194 }
 195
 196 #ifdef HAVE_ICONV
 197
 198 static char* iconv_simple(const char *str, const char *to, const char *from) {
 199     char *new_str;
 200     size_t len, inlen;
 201     iconv_t cd;
 202     ICONV_CONST char *inbuf;
 203     char *outbuf;
 204     size_t res, inbytes, outbytes;
 205
 206     pa_assert(str);
 207     pa_assert(to);
 208     pa_assert(from);
 209
 210     cd = iconv_open(to, from);
 211     if (cd == (iconv_t)-1)
 212         return NULL;
 213
 214     inlen = len = strlen(str) + 1;
 215     new_str = pa_xmalloc(len);
 216
 217     for (;;) {
 218         inbuf = (ICONV_CONST char*) str; /* Brain dead prototype for iconv() */
 219         inbytes = inlen;
 220         outbuf = new_str;
 221         outbytes = len;
 222
 223         res = iconv(cd, &inbuf, &inbytes, &outbuf, &outbytes);
 224
 225         if (res != (size_t)-1)
 226             break;
 227
 228         if (errno != E2BIG) {
 229             pa_xfree(new_str);
 230             new_str = NULL;
 231             break;
 232         }
 233
 234         pa_assert(inbytes != 0);
 235
 236         len += inbytes;
 237         new_str = pa_xrealloc(new_str, len);
 238     }
 239
 240     iconv_close(cd);
 241
 242     return new_str;
 243 }
 244
 245 char* pa_utf8_to_locale (const char *str) {
 246     return iconv_simple(str, "", "UTF-8");
 247 }
 248
 249 char* pa_locale_to_utf8 (const char *str) {
 250     return iconv_simple(str, "UTF-8", "");
 251 }
 252
 253 #else
 254
 255 char* pa_utf8_to_locale (const char *str) {
 256     pa_assert(str);
 257     return NULL;
 258 }
 259
 260 char* pa_locale_to_utf8 (const char *str) {
 261     pa_assert(str);
 262     return NULL;
 263 }
 264
 265 #endif
 266
 267 char *pa_ascii_valid(const char *str) {
 268     const char *p;
 269     pa_assert(str);
 270
 271     for (p = str; *p; p++)
 272         if ((unsigned char) *p >= 128)
 273             return NULL;
 274
 275     return (char*) str;
 276 }
 277
 278 char *pa_ascii_filter(const char *str) {
 279     char *r, *s, *d;
 280     pa_assert(str);
 281
 282     r = pa_xstrdup(str);
 283
 284     for (s = r, d = r; *s; s++)
 285         if ((unsigned char) *s < 128)
 286             *(d++) = *s;
 287
 288     *d = 0;
 289
 290     return r;
 291 }