libavcodec/htmlsubtitles.c

   1 /*
   2  * Copyright (c) 2010  Aurelien Jacobs <aurel@gnuage.org>
   3  * Copyright (c) 2017  Clément Bœsch <u@pkh.me>
   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 "libavutil/avassert.h"
  23 #include "libavutil/avstring.h"
  24 #include "libavutil/common.h"
  25 #include "libavutil/parseutils.h"
  26 #include "htmlsubtitles.h"
  27 #include <ctype.h>
  28
  29 static int html_color_parse(void *log_ctx, const char *str)
  30 {
  31     uint8_t rgba[4];
  32     int nb_sharps = 0;
  33     while (str[nb_sharps] == '#')
  34         nb_sharps++;
  35     str += FFMAX(0, nb_sharps - 1);
  36     if (av_parse_color(rgba, str, strcspn(str, "\" >"), log_ctx) < 0)
  37         return -1;
  38     return rgba[0] | rgba[1] << 8 | rgba[2] << 16;
  39 }
  40
  41 static void rstrip_spaces_buf(AVBPrint *buf)
  42 {
  43     if (av_bprint_is_complete(buf))
  44         while (buf->len > 0 && buf->str[buf->len - 1] == ' ')
  45             buf->str[--buf->len] = 0;
  46 }
  47
  48 /*
  49  * Fast code for scanning text enclosed in braces. Functionally
  50  * equivalent to this sscanf call:
  51  *
  52  * sscanf(in, "{\\an%*1u}%n", &len) >= 0 && len > 0
  53  */
  54 static int scanbraces(const char* in) {
  55     if (strncmp(in, "{\\an", 4) != 0) {
  56         return 0;
  57     }
  58     if (!av_isdigit(in[4])) {
  59         return 0;
  60     }
  61     if (in[5] != '}') {
  62         return 0;
  63     }
  64     return 1;
  65 }
  66
  67 /* skip all {\xxx} substrings except for {\an%d}
  68    and all microdvd like styles such as {Y:xxx} */
  69 static void handle_open_brace(AVBPrint *dst, const char **inp, int *an, int *closing_brace_missing)
  70 {
  71     const char *in = *inp;
  72
  73     *an += scanbraces(in);
  74
  75     if (!*closing_brace_missing) {
  76         if (   (*an != 1 && in[1] == '\\')
  77             || (in[1] && strchr("CcFfoPSsYy", in[1]) && in[2] == ':')) {
  78             char *bracep = strchr(in+2, '}');
  79             if (bracep) {
  80                 *inp = bracep;
  81                 return;
  82             } else
  83                 *closing_brace_missing = 1;
  84         }
  85     }
  86
  87     av_bprint_chars(dst, *in, 1);
  88 }
  89
  90 struct font_tag {
  91     char face[128];
  92     int size;
  93     uint32_t color;
  94 };
  95
  96 /*
  97  * Fast code for scanning the rest of a tag. Functionally equivalent to
  98  * this sscanf call:
  99  *
 100  * sscanf(in, "%127[^<>]>%n", buffer, lenp) == 2
 101  */
 102 static int scantag(const char* in, char* buffer, int* lenp) {
 103     int len;
 104
 105     for (len = 0; len < 128; len++) {
 106         const char c = *in++;
 107         switch (c) {
 108         case '\0':
 109             return 0;
 110         case '<':
 111             return 0;
 112         case '>':
 113             buffer[len] = '\0';
 114             *lenp = len+1;
 115             return 1;
 116         default:
 117             break;
 118         }
 119         buffer[len] = c;
 120     }
 121     return 0;
 122 }
 123
 124 /*
 125  * The general politic of the convert is to mask unsupported tags or formatting
 126  * errors (but still alert the user/subtitles writer with an error/warning)
 127  * without dropping any actual text content for the final user.
 128  */
 129 int ff_htmlmarkup_to_ass(void *log_ctx, AVBPrint *dst, const char *in)
 130 {
 131     char *param, buffer[128];
 132     int len, tag_close, sptr = 0, line_start = 1, an = 0, end = 0;
 133     int closing_brace_missing = 0;
 134     int i, likely_a_tag;
 135
 136     /*
 137      * state stack is only present for fonts since they are the only tags where
 138      * the state is not binary. Here is a typical use case:
 139      *
 140      *   <font color="red" size=10>
 141      *     red 10
 142      *     <font size=50> RED AND BIG </font>
 143      *     red 10 again
 144      *   </font>
 145      *
 146      * On the other hand, using the state system for all the tags should be
 147      * avoided because it breaks wrongly nested tags such as:
 148      *
 149      *   <b> foo <i> bar </b> bla </i>
 150      *
 151      * We don't want to break here; instead, we will treat all these tags as
 152      * binary state markers. Basically, "<b>" will activate bold, and "</b>"
 153      * will deactivate it, whatever the current state.
 154      *
 155      * This will also prevents cases where we have a random closing tag
 156      * remaining after the opening one was dropped. Yes, this happens and we
 157      * still don't want to print a "</b>" at the end of the dialog event.
 158      */
 159     struct font_tag stack[16];
 160
 161     memset(&stack[0], 0, sizeof(stack[0]));
 162
 163     for (; !end && *in; in++) {
 164         switch (*in) {
 165         case '\r':
 166             break;
 167         case '\n':
 168             if (line_start) {
 169                 end = 1;
 170                 break;
 171             }
 172             rstrip_spaces_buf(dst);
 173             av_bprintf(dst, "\\N");
 174             line_start = 1;
 175             break;
 176         case ' ':
 177             if (!line_start)
 178                 av_bprint_chars(dst, *in, 1);
 179             break;
 180         case '{':
 181             handle_open_brace(dst, &in, &an, &closing_brace_missing);
 182             break;
 183         case '<':
 184             /*
 185              * "<<" are likely latin guillemets in ASCII or some kind of random
 186              * style effect; see sub/badsyntax.srt in the FATE samples
 187              * directory for real test cases.
 188              */
 189
 190             likely_a_tag = 1;
 191             for (i = 0; in[1] == '<'; i++) {
 192                 av_bprint_chars(dst, '<', 1);
 193                 likely_a_tag = 0;
 194                 in++;
 195             }
 196
 197             tag_close = in[1] == '/';
 198             if (tag_close)
 199                 likely_a_tag = 1;
 200
 201             av_assert0(in[0] == '<');
 202
 203             len = 0;
 204
 205             if (scantag(in+tag_close+1, buffer, &len) && len > 0) {
 206                 const int skip = len + tag_close;
 207                 const char *tagname = buffer;
 208                 while (*tagname == ' ') {
 209                     likely_a_tag = 0;
 210                     tagname++;
 211                 }
 212                 if ((param = strchr(tagname, ' ')))
 213                     *param++ = 0;
 214
 215                 /* Check if this is likely a tag */
 216 #define LIKELY_A_TAG_CHAR(x) (((x) >= '0' && (x) <= '9') || \
 217                               ((x) >= 'a' && (x) <= 'z') || \
 218                               ((x) >= 'A' && (x) <= 'Z') || \
 219                                (x) == '_' || (x) == '/')
 220                 for (i = 0; tagname[i]; i++) {
 221                     if (!LIKELY_A_TAG_CHAR(tagname[i])) {
 222                         likely_a_tag = 0;
 223                         break;
 224                     }
 225                 }
 226
 227                 if (!av_strcasecmp(tagname, "font")) {
 228                     if (tag_close && sptr > 0) {
 229                         struct font_tag *cur_tag  = &stack[sptr--];
 230                         struct font_tag *last_tag = &stack[sptr];
 231
 232                         if (cur_tag->size) {
 233                             if (!last_tag->size)
 234                                 av_bprintf(dst, "{\\fs}");
 235                             else if (last_tag->size != cur_tag->size)
 236                                 av_bprintf(dst, "{\\fs%d}", last_tag->size);
 237                         }
 238
 239                         if (cur_tag->color & 0xff000000) {
 240                             if (!(last_tag->color & 0xff000000))
 241                                 av_bprintf(dst, "{\\c}");
 242                             else if (last_tag->color != cur_tag->color)
 243                                 av_bprintf(dst, "{\\c&H%"PRIX32"&}", last_tag->color & 0xffffff);
 244                         }
 245
 246                         if (cur_tag->face[0]) {
 247                             if (!last_tag->face[0])
 248                                 av_bprintf(dst, "{\\fn}");
 249                             else if (strcmp(last_tag->face, cur_tag->face))
 250                                 av_bprintf(dst, "{\\fn%s}", last_tag->face);
 251                         }
 252                     } else if (!tag_close && sptr < FF_ARRAY_ELEMS(stack) - 1) {
 253                         struct font_tag *new_tag = &stack[sptr + 1];
 254
 255                         *new_tag = stack[sptr++];
 256
 257                         while (param) {
 258                             if (!av_strncasecmp(param, "size=", 5)) {
 259                                 param += 5 + (param[5] == '"');
 260                                 if (sscanf(param, "%u", &new_tag->size) == 1)
 261                                     av_bprintf(dst, "{\\fs%u}", new_tag->size);
 262                             } else if (!av_strncasecmp(param, "color=", 6)) {
 263                                 int color;
 264                                 param += 6 + (param[6] == '"');
 265                                 color = html_color_parse(log_ctx, param);
 266                                 if (color >= 0) {
 267                                     new_tag->color = 0xff000000 | color;
 268                                     av_bprintf(dst, "{\\c&H%"PRIX32"&}", new_tag->color & 0xffffff);
 269                                 }
 270                             } else if (!av_strncasecmp(param, "face=", 5)) {
 271                                 param += 5 + (param[5] == '"');
 272                                 len = strcspn(param,
 273                                               param[-1] == '"' ? "\"" :" ");
 274                                 av_strlcpy(new_tag->face, param,
 275                                            FFMIN(sizeof(new_tag->face), len+1));
 276                                 param += len;
 277                                 av_bprintf(dst, "{\\fn%s}", new_tag->face);
 278                             }
 279                             if ((param = strchr(param, ' ')))
 280                                 param++;
 281                         }
 282                     }
 283                     in += skip;
 284                 } else if (tagname[0] && !tagname[1] && strchr("bisu", av_tolower(tagname[0]))) {
 285                     av_bprintf(dst, "{\\%c%d}", (char)av_tolower(tagname[0]), !tag_close);
 286                     in += skip;
 287                 } else if (!av_strncasecmp(tagname, "br", 2) &&
 288                            (!tagname[2] || (tagname[2] == '/' && !tagname[3]))) {
 289                     av_bprintf(dst, "\\N");
 290                     in += skip;
 291                 } else if (likely_a_tag) {
 292                     if (!tag_close) // warn only once
 293                         av_log(log_ctx, AV_LOG_WARNING, "Unrecognized tag %s\n", tagname);
 294                     in += skip;
 295                 } else {
 296                     av_bprint_chars(dst, '<', 1);
 297                 }
 298             } else {
 299                 av_bprint_chars(dst, *in, 1);
 300             }
 301             break;
 302         default:
 303             av_bprint_chars(dst, *in, 1);
 304             break;
 305         }
 306         if (*in != ' ' && *in != '\r' && *in != '\n')
 307             line_start = 0;
 308     }
 309
 310     if (!av_bprint_is_complete(dst))
 311         return AVERROR(ENOMEM);
 312
 313     while (dst->len >= 2 && !strncmp(&dst->str[dst->len - 2], "\\N", 2))
 314         dst->len -= 2;
 315     dst->str[dst->len] = 0;
 316     rstrip_spaces_buf(dst);
 317
 318     return 0;
 319 }