libavutil/avstring.c

   1 /*
   2  * Copyright (c) 2000, 2001, 2002 Fabrice Bellard
   3  * Copyright (c) 2007 Mans Rullgard
   4  *
   5  * This file is part of FFmpeg.
   6  *
   7  * FFmpeg is free software; you can redistribute it and/or
   8  * modify it under the terms of the GNU Lesser General Public
   9  * License as published by the Free Software Foundation; either
  10  * version 2.1 of the License, or (at your option) any later version.
  11  *
  12  * FFmpeg 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 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 FFmpeg; if not, write to the Free Software
  19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  20  */
  21
  22 #include <limits.h>
  23 #include <stdarg.h>
  24 #include <stdint.h>
  25 #include <stdio.h>
  26 #include <string.h>
  27
  28 #include "config.h"
  29 #include "mem.h"
  30 #include "avassert.h"
  31 #include "avstring.h"
  32 #include "bprint.h"
  33 #include "error.h"
  34 #include "macros.h"
  35
  36 int av_strstart(const char *str, const char *pfx, const char **ptr)
  37 {
  38     while (*pfx && *pfx == *str) {
  39         pfx++;
  40         str++;
  41     }
  42     if (!*pfx && ptr)
  43         *ptr = str;
  44     return !*pfx;
  45 }
  46
  47 int av_stristart(const char *str, const char *pfx, const char **ptr)
  48 {
  49     while (*pfx && av_toupper((unsigned)*pfx) == av_toupper((unsigned)*str)) {
  50         pfx++;
  51         str++;
  52     }
  53     if (!*pfx && ptr)
  54         *ptr = str;
  55     return !*pfx;
  56 }
  57
  58 char *av_stristr(const char *s1, const char *s2)
  59 {
  60     if (!*s2)
  61         return (char*)(intptr_t)s1;
  62
  63     do
  64         if (av_stristart(s1, s2, NULL))
  65             return (char*)(intptr_t)s1;
  66     while (*s1++);
  67
  68     return NULL;
  69 }
  70
  71 char *av_strnstr(const char *haystack, const char *needle, size_t hay_length)
  72 {
  73     size_t needle_len = strlen(needle);
  74     if (!needle_len)
  75         return (char*)haystack;
  76     while (hay_length >= needle_len) {
  77         hay_length--;
  78         if (!memcmp(haystack, needle, needle_len))
  79             return (char*)haystack;
  80         haystack++;
  81     }
  82     return NULL;
  83 }
  84
  85 size_t av_strlcpy(char *dst, const char *src, size_t size)
  86 {
  87     size_t len = 0;
  88     while (++len < size && *src)
  89         *dst++ = *src++;
  90     if (len <= size)
  91         *dst = 0;
  92     return len + strlen(src) - 1;
  93 }
  94
  95 size_t av_strlcat(char *dst, const char *src, size_t size)
  96 {
  97     size_t len = strlen(dst);
  98     if (size <= len + 1)
  99         return len + strlen(src);
 100     return len + av_strlcpy(dst + len, src, size - len);
 101 }
 102
 103 size_t av_strlcatf(char *dst, size_t size, const char *fmt, ...)
 104 {
 105     size_t len = strlen(dst);
 106     va_list vl;
 107
 108     va_start(vl, fmt);
 109     len += vsnprintf(dst + len, size > len ? size - len : 0, fmt, vl);
 110     va_end(vl);
 111
 112     return len;
 113 }
 114
 115 char *av_asprintf(const char *fmt, ...)
 116 {
 117     char *p = NULL;
 118     va_list va;
 119     int len;
 120
 121     va_start(va, fmt);
 122     len = vsnprintf(NULL, 0, fmt, va);
 123     va_end(va);
 124     if (len < 0)
 125         goto end;
 126
 127     p = av_malloc(len + 1);
 128     if (!p)
 129         goto end;
 130
 131     va_start(va, fmt);
 132     len = vsnprintf(p, len + 1, fmt, va);
 133     va_end(va);
 134     if (len < 0)
 135         av_freep(&p);
 136
 137 end:
 138     return p;
 139 }
 140
 141 #define WHITESPACES " \n\t\r"
 142
 143 char *av_get_token(const char **buf, const char *term)
 144 {
 145     char *out     = av_realloc(NULL, strlen(*buf) + 1);
 146     char *ret     = out, *end = out;
 147     const char *p = *buf;
 148     if (!out)
 149         return NULL;
 150     p += strspn(p, WHITESPACES);
 151
 152     while (*p && !strspn(p, term)) {
 153         char c = *p++;
 154         if (c == '\\' && *p) {
 155             *out++ = *p++;
 156             end    = out;
 157         } else if (c == '\'') {
 158             while (*p && *p != '\'')
 159                 *out++ = *p++;
 160             if (*p) {
 161                 p++;
 162                 end = out;
 163             }
 164         } else {
 165             *out++ = c;
 166         }
 167     }
 168
 169     do
 170         *out-- = 0;
 171     while (out >= end && strspn(out, WHITESPACES));
 172
 173     *buf = p;
 174
 175     char *small_ret = av_realloc(ret, out - ret + 2);
 176     return small_ret ? small_ret : ret;
 177 }
 178
 179 char *av_strtok(char *s, const char *delim, char **saveptr)
 180 {
 181     char *tok;
 182
 183     if (!s && !(s = *saveptr))
 184         return NULL;
 185
 186     /* skip leading delimiters */
 187     s += strspn(s, delim);
 188
 189     /* s now points to the first non delimiter char, or to the end of the string */
 190     if (!*s) {
 191         *saveptr = NULL;
 192         return NULL;
 193     }
 194     tok = s++;
 195
 196     /* skip non delimiters */
 197     s += strcspn(s, delim);
 198     if (*s) {
 199         *s = 0;
 200         *saveptr = s+1;
 201     } else {
 202         *saveptr = NULL;
 203     }
 204
 205     return tok;
 206 }
 207
 208 int av_strcasecmp(const char *a, const char *b)
 209 {
 210     uint8_t c1, c2;
 211     do {
 212         c1 = av_tolower(*a++);
 213         c2 = av_tolower(*b++);
 214     } while (c1 && c1 == c2);
 215     return c1 - c2;
 216 }
 217
 218 int av_strncasecmp(const char *a, const char *b, size_t n)
 219 {
 220     uint8_t c1, c2;
 221     if (n <= 0)
 222         return 0;
 223     do {
 224         c1 = av_tolower(*a++);
 225         c2 = av_tolower(*b++);
 226     } while (--n && c1 && c1 == c2);
 227     return c1 - c2;
 228 }
 229
 230 char *av_strireplace(const char *str, const char *from, const char *to)
 231 {
 232     char *ret = NULL;
 233     const char *pstr2, *pstr = str;
 234     size_t tolen = strlen(to), fromlen = strlen(from);
 235     AVBPrint pbuf;
 236
 237     av_bprint_init(&pbuf, 1, AV_BPRINT_SIZE_UNLIMITED);
 238     while ((pstr2 = av_stristr(pstr, from))) {
 239         av_bprint_append_data(&pbuf, pstr, pstr2 - pstr);
 240         pstr = pstr2 + fromlen;
 241         av_bprint_append_data(&pbuf, to, tolen);
 242     }
 243     av_bprint_append_data(&pbuf, pstr, strlen(pstr));
 244     if (!av_bprint_is_complete(&pbuf)) {
 245         av_bprint_finalize(&pbuf, NULL);
 246     } else {
 247         av_bprint_finalize(&pbuf, &ret);
 248     }
 249
 250     return ret;
 251 }
 252
 253 const char *av_basename(const char *path)
 254 {
 255     char *p;
 256 #if HAVE_DOS_PATHS
 257     char *q, *d;
 258 #endif
 259
 260     if (!path || *path == '\0')
 261         return ".";
 262
 263     p = strrchr(path, '/');
 264 #if HAVE_DOS_PATHS
 265     q = strrchr(path, '\\');
 266     d = strchr(path, ':');
 267     p = FFMAX3(p, q, d);
 268 #endif
 269
 270     if (!p)
 271         return path;
 272
 273     return p + 1;
 274 }
 275
 276 const char *av_dirname(char *path)
 277 {
 278     char *p = path ? strrchr(path, '/') : NULL;
 279
 280 #if HAVE_DOS_PATHS
 281     char *q = path ? strrchr(path, '\\') : NULL;
 282     char *d = path ? strchr(path, ':')  : NULL;
 283
 284     d = d ? d + 1 : d;
 285
 286     p = FFMAX3(p, q, d);
 287 #endif
 288
 289     if (!p)
 290         return ".";
 291
 292     *p = '\0';
 293
 294     return path;
 295 }
 296
 297 char *av_append_path_component(const char *path, const char *component)
 298 {
 299     size_t p_len, c_len;
 300     char *fullpath;
 301
 302     if (!path)
 303         return av_strdup(component);
 304     if (!component)
 305         return av_strdup(path);
 306
 307     p_len = strlen(path);
 308     c_len = strlen(component);
 309     if (p_len > SIZE_MAX - c_len || p_len + c_len > SIZE_MAX - 2)
 310         return NULL;
 311     fullpath = av_malloc(p_len + c_len + 2);
 312     if (fullpath) {
 313         if (p_len) {
 314             av_strlcpy(fullpath, path, p_len + 1);
 315             if (c_len) {
 316                 if (fullpath[p_len - 1] != '/' && component[0] != '/')
 317                     fullpath[p_len++] = '/';
 318                 else if (fullpath[p_len - 1] == '/' && component[0] == '/')
 319                     p_len--;
 320             }
 321         }
 322         av_strlcpy(&fullpath[p_len], component, c_len + 1);
 323         fullpath[p_len + c_len] = 0;
 324     }
 325     return fullpath;
 326 }
 327
 328 int av_escape(char **dst, const char *src, const char *special_chars,
 329               enum AVEscapeMode mode, int flags)
 330 {
 331     AVBPrint dstbuf;
 332     int ret;
 333
 334     av_bprint_init(&dstbuf, 1, INT_MAX); /* (int)dstbuf.len must be >= 0 */
 335     av_bprint_escape(&dstbuf, src, special_chars, mode, flags);
 336
 337     if (!av_bprint_is_complete(&dstbuf)) {
 338         av_bprint_finalize(&dstbuf, NULL);
 339         return AVERROR(ENOMEM);
 340     }
 341     if ((ret = av_bprint_finalize(&dstbuf, dst)) < 0)
 342         return ret;
 343     return dstbuf.len;
 344 }
 345
 346 int av_match_name(const char *name, const char *names)
 347 {
 348     const char *p;
 349     size_t len, namelen;
 350
 351     if (!name || !names)
 352         return 0;
 353
 354     namelen = strlen(name);
 355     while (*names) {
 356         int negate = '-' == *names;
 357         p = strchr(names, ',');
 358         if (!p)
 359             p = names + strlen(names);
 360         names += negate;
 361         len = FFMAX(p - names, namelen);
 362         if (!av_strncasecmp(name, names, len) || !strncmp("ALL", names, FFMAX(3, p - names)))
 363             return !negate;
 364         names = p + (*p == ',');
 365     }
 366     return 0;
 367 }
 368
 369 int av_utf8_decode(int32_t *codep, const uint8_t **bufp, const uint8_t *buf_end,
 370                    unsigned int flags)
 371 {
 372     const uint8_t *p = *bufp;
 373     uint32_t top;
 374     uint64_t code;
 375     int ret = 0, tail_len;
 376     uint32_t overlong_encoding_mins[6] = {
 377         0x00000000, 0x00000080, 0x00000800, 0x00010000, 0x00200000, 0x04000000,
 378     };
 379
 380     if (p >= buf_end)
 381         return 0;
 382
 383     code = *p++;
 384
 385     /* first sequence byte starts with 10, or is 1111-1110 or 1111-1111,
 386        which is not admitted */
 387     if ((code & 0xc0) == 0x80 || code >= 0xFE) {
 388         ret = AVERROR(EILSEQ);
 389         goto end;
 390     }
 391     top = (code & 128) >> 1;
 392
 393     tail_len = 0;
 394     while (code & top) {
 395         int tmp;
 396         tail_len++;
 397         if (p >= buf_end) {
 398             (*bufp) ++;
 399             return AVERROR(EILSEQ); /* incomplete sequence */
 400         }
 401
 402         /* we assume the byte to be in the form 10xx-xxxx */
 403         tmp = *p++ - 128;   /* strip leading 1 */
 404         if (tmp>>6) {
 405             (*bufp) ++;
 406             return AVERROR(EILSEQ);
 407         }
 408         code = (code<<6) + tmp;
 409         top <<= 5;
 410     }
 411     code &= (top << 1) - 1;
 412
 413     /* check for overlong encodings */
 414     av_assert0(tail_len <= 5);
 415     if (code < overlong_encoding_mins[tail_len]) {
 416         ret = AVERROR(EILSEQ);
 417         goto end;
 418     }
 419
 420     if (code >= 1U<<31) {
 421         ret = AVERROR(EILSEQ);  /* out-of-range value */
 422         goto end;
 423     }
 424
 425     *codep = code;
 426
 427     if (code > 0x10FFFF &&
 428         !(flags & AV_UTF8_FLAG_ACCEPT_INVALID_BIG_CODES))
 429         ret = AVERROR(EILSEQ);
 430     if (code < 0x20 && code != 0x9 && code != 0xA && code != 0xD &&
 431         flags & AV_UTF8_FLAG_EXCLUDE_XML_INVALID_CONTROL_CODES)
 432         ret = AVERROR(EILSEQ);
 433     if (code >= 0xD800 && code <= 0xDFFF &&
 434         !(flags & AV_UTF8_FLAG_ACCEPT_SURROGATES))
 435         ret = AVERROR(EILSEQ);
 436     if ((code == 0xFFFE || code == 0xFFFF) &&
 437         !(flags & AV_UTF8_FLAG_ACCEPT_NON_CHARACTERS))
 438         ret = AVERROR(EILSEQ);
 439
 440 end:
 441     *bufp = p;
 442     return ret;
 443 }
 444
 445 int av_match_list(const char *name, const char *list, char separator)
 446 {
 447     const char *p, *q;
 448
 449     for (p = name; p && *p; ) {
 450         for (q = list; q && *q; ) {
 451             int k;
 452             for (k = 0; p[k] == q[k] || (p[k]*q[k] == 0 && p[k]+q[k] == separator); k++)
 453                 if (k && (!p[k] || p[k] == separator))
 454                     return 1;
 455             q = strchr(q, separator);
 456             if(q)
 457                 q++;
 458         }
 459         p = strchr(p, separator);
 460         if (p)
 461             p++;
 462     }
 463
 464     return 0;
 465 }