FFmpeg
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
htmlsubtitles.c
Go to the documentation of this file.
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 
28 static int html_color_parse(void *log_ctx, const char *str)
29 {
30  uint8_t rgba[4];
31  int nb_sharps = 0;
32  while (str[nb_sharps] == '#')
33  nb_sharps++;
34  str += FFMAX(0, nb_sharps - 1);
35  if (av_parse_color(rgba, str, strcspn(str, "\" >"), log_ctx) < 0)
36  return -1;
37  return rgba[0] | rgba[1] << 8 | rgba[2] << 16;
38 }
39 
40 static void rstrip_spaces_buf(AVBPrint *buf)
41 {
42  if (av_bprint_is_complete(buf))
43  while (buf->len > 0 && buf->str[buf->len - 1] == ' ')
44  buf->str[--buf->len] = 0;
45 }
46 
47 /* skip all {\xxx} substrings except for {\an%d}
48  and all microdvd like styles such as {Y:xxx} */
49 static void handle_open_brace(AVBPrint *dst, const char **inp, int *an, int *closing_brace_missing)
50 {
51  int len = 0;
52  const char *in = *inp;
53 
54  *an += sscanf(in, "{\\an%*1u}%n", &len) >= 0 && len > 0;
55 
56  if (!*closing_brace_missing) {
57  if ( (*an != 1 && in[1] == '\\')
58  || (in[1] && strchr("CcFfoPSsYy", in[1]) && in[2] == ':')) {
59  char *bracep = strchr(in+2, '}');
60  if (bracep) {
61  *inp = bracep;
62  return;
63  } else
64  *closing_brace_missing = 1;
65  }
66  }
67 
68  av_bprint_chars(dst, *in, 1);
69 }
70 
71 struct font_tag {
72  char face[128];
73  int size;
74  uint32_t color;
75 };
76 
77 /*
78  * The general politic of the convert is to mask unsupported tags or formatting
79  * errors (but still alert the user/subtitles writer with an error/warning)
80  * without dropping any actual text content for the final user.
81  */
82 int ff_htmlmarkup_to_ass(void *log_ctx, AVBPrint *dst, const char *in)
83 {
84  char *param, buffer[128];
85  int len, tag_close, sptr = 0, line_start = 1, an = 0, end = 0;
86  int closing_brace_missing = 0;
87  int i, likely_a_tag;
88 
89  /*
90  * state stack is only present for fonts since they are the only tags where
91  * the state is not binary. Here is a typical use case:
92  *
93  * <font color="red" size=10>
94  * red 10
95  * <font size=50> RED AND BIG </font>
96  * red 10 again
97  * </font>
98  *
99  * On the other hand, using the state system for all the tags should be
100  * avoided because it breaks wrongly nested tags such as:
101  *
102  * <b> foo <i> bar </b> bla </i>
103  *
104  * We don't want to break here; instead, we will treat all these tags as
105  * binary state markers. Basically, "<b>" will activate bold, and "</b>"
106  * will deactivate it, whatever the current state.
107  *
108  * This will also prevents cases where we have a random closing tag
109  * remaining after the opening one was dropped. Yes, this happens and we
110  * still don't want to print a "</b>" at the end of the dialog event.
111  */
112  struct font_tag stack[16];
113 
114  memset(&stack[0], 0, sizeof(stack[0]));
115 
116  for (; !end && *in; in++) {
117  switch (*in) {
118  case '\r':
119  break;
120  case '\n':
121  if (line_start) {
122  end = 1;
123  break;
124  }
125  rstrip_spaces_buf(dst);
126  av_bprintf(dst, "\\N");
127  line_start = 1;
128  break;
129  case ' ':
130  if (!line_start)
131  av_bprint_chars(dst, *in, 1);
132  break;
133  case '{':
134  handle_open_brace(dst, &in, &an, &closing_brace_missing);
135  break;
136  case '<':
137  /*
138  * "<<" are likely latin guillemets in ASCII or some kind of random
139  * style effect; see sub/badsyntax.srt in the FATE samples
140  * directory for real test cases.
141  */
142 
143  likely_a_tag = 1;
144  for (i = 0; in[1] == '<'; i++) {
145  av_bprint_chars(dst, '<', 1);
146  likely_a_tag = 0;
147  in++;
148  }
149 
150  tag_close = in[1] == '/';
151  if (tag_close)
152  likely_a_tag = 1;
153 
154  av_assert0(in[0] == '<');
155 
156  len = 0;
157 
158  if (sscanf(in+tag_close+1, "%127[^<>]>%n", buffer, &len) >= 1 && len > 0) {
159  const int skip = len + tag_close;
160  const char *tagname = buffer;
161  while (*tagname == ' ') {
162  likely_a_tag = 0;
163  tagname++;
164  }
165  if ((param = strchr(tagname, ' ')))
166  *param++ = 0;
167 
168  /* Check if this is likely a tag */
169 #define LIKELY_A_TAG_CHAR(x) (((x) >= '0' && (x) <= '9') || \
170  ((x) >= 'a' && (x) <= 'z') || \
171  ((x) >= 'A' && (x) <= 'Z') || \
172  (x) == '_' || (x) == '/')
173  for (i = 0; tagname[i]; i++) {
174  if (!LIKELY_A_TAG_CHAR(tagname[i])) {
175  likely_a_tag = 0;
176  break;
177  }
178  }
179 
180  if (!av_strcasecmp(tagname, "font")) {
181  if (tag_close && sptr > 0) {
182  struct font_tag *cur_tag = &stack[sptr--];
183  struct font_tag *last_tag = &stack[sptr];
184 
185  if (cur_tag->size) {
186  if (!last_tag->size)
187  av_bprintf(dst, "{\\fs}");
188  else if (last_tag->size != cur_tag->size)
189  av_bprintf(dst, "{\\fs%d}", last_tag->size);
190  }
191 
192  if (cur_tag->color & 0xff000000) {
193  if (!(last_tag->color & 0xff000000))
194  av_bprintf(dst, "{\\c}");
195  else if (last_tag->color != cur_tag->color)
196  av_bprintf(dst, "{\\c&H%"PRIX32"&}", last_tag->color & 0xffffff);
197  }
198 
199  if (cur_tag->face[0]) {
200  if (!last_tag->face[0])
201  av_bprintf(dst, "{\\fn}");
202  else if (strcmp(last_tag->face, cur_tag->face))
203  av_bprintf(dst, "{\\fn%s}", last_tag->face);
204  }
205  } else if (!tag_close && sptr < FF_ARRAY_ELEMS(stack) - 1) {
206  struct font_tag *new_tag = &stack[sptr + 1];
207 
208  *new_tag = stack[sptr++];
209 
210  while (param) {
211  if (!av_strncasecmp(param, "size=", 5)) {
212  param += 5 + (param[5] == '"');
213  if (sscanf(param, "%u", &new_tag->size) == 1)
214  av_bprintf(dst, "{\\fs%u}", new_tag->size);
215  } else if (!av_strncasecmp(param, "color=", 6)) {
216  int color;
217  param += 6 + (param[6] == '"');
218  color = html_color_parse(log_ctx, param);
219  if (color >= 0) {
220  new_tag->color = 0xff000000 | color;
221  av_bprintf(dst, "{\\c&H%"PRIX32"&}", new_tag->color & 0xffffff);
222  }
223  } else if (!av_strncasecmp(param, "face=", 5)) {
224  param += 5 + (param[5] == '"');
225  len = strcspn(param,
226  param[-1] == '"' ? "\"" :" ");
227  av_strlcpy(new_tag->face, param,
228  FFMIN(sizeof(new_tag->face), len+1));
229  param += len;
230  av_bprintf(dst, "{\\fn%s}", new_tag->face);
231  }
232  if ((param = strchr(param, ' ')))
233  param++;
234  }
235  }
236  in += skip;
237  } else if (tagname[0] && !tagname[1] && strchr("bisu", av_tolower(tagname[0]))) {
238  av_bprintf(dst, "{\\%c%d}", (char)av_tolower(tagname[0]), !tag_close);
239  in += skip;
240  } else if (!av_strncasecmp(tagname, "br", 2) &&
241  (!tagname[2] || (tagname[2] == '/' && !tagname[3]))) {
242  av_bprintf(dst, "\\N");
243  in += skip;
244  } else if (likely_a_tag) {
245  if (!tag_close) // warn only once
246  av_log(log_ctx, AV_LOG_WARNING, "Unrecognized tag %s\n", tagname);
247  in += skip;
248  } else {
249  av_bprint_chars(dst, '<', 1);
250  }
251  } else {
252  av_bprint_chars(dst, *in, 1);
253  }
254  break;
255  default:
256  av_bprint_chars(dst, *in, 1);
257  break;
258  }
259  if (*in != ' ' && *in != '\r' && *in != '\n')
260  line_start = 0;
261  }
262 
263  if (!av_bprint_is_complete(dst))
264  return AVERROR(ENOMEM);
265 
266  while (dst->len >= 2 && !strncmp(&dst->str[dst->len - 2], "\\N", 2))
267  dst->len -= 2;
268  dst->str[dst->len] = 0;
269  rstrip_spaces_buf(dst);
270 
271  return 0;
272 }
void av_bprintf(AVBPrint *buf, const char *fmt,...)
Definition: bprint.c:94
#define AV_LOG_WARNING
Something somehow does not look correct.
Definition: log.h:182
int av_strncasecmp(const char *a, const char *b, size_t n)
Locale-independent case-insensitive compare.
Definition: avstring.c:223
static void rstrip_spaces_buf(AVBPrint *buf)
Definition: htmlsubtitles.c:40
#define av_assert0(cond)
assert() equivalent, that is always enabled.
Definition: avassert.h:37
uint8_t
static const uint32_t color[16+AV_CLASS_CATEGORY_NB]
Definition: log.c:92
static av_cold int end(AVCodecContext *avctx)
Definition: avrndec.c:90
#define av_log(a,...)
int av_parse_color(uint8_t *rgba_color, const char *color_string, int slen, void *log_ctx)
Put the RGBA values that correspond to color_string in rgba_color.
Definition: parseutils.c:354
char face[128]
Definition: htmlsubtitles.c:72
static av_const int av_tolower(int c)
Locale-independent conversion of ASCII characters to lowercase.
Definition: avstring.h:241
#define AVERROR(e)
Definition: error.h:43
static void handle_open_brace(AVBPrint *dst, const char **inp, int *an, int *closing_brace_missing)
Definition: htmlsubtitles.c:49
simple assert() macros that are a bit more flexible than ISO C assert().
#define FFMAX(a, b)
Definition: common.h:94
size_t av_strlcpy(char *dst, const char *src, size_t size)
Copy the string src to dst, but no more than size - 1 bytes, and null-terminate dst.
Definition: avstring.c:83
#define FFMIN(a, b)
Definition: common.h:96
int av_strcasecmp(const char *a, const char *b)
Locale-independent case-insensitive compare.
Definition: avstring.c:213
int ff_htmlmarkup_to_ass(void *log_ctx, AVBPrint *dst, const char *in)
Definition: htmlsubtitles.c:82
static int av_bprint_is_complete(const AVBPrint *buf)
Test if the print buffer is complete (not truncated).
Definition: bprint.h:185
#define FF_ARRAY_ELEMS(a)
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(constuint8_t *) pi-0x80)*(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(constint16_t *) pi >>8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t,*(constint16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(constint32_t *) pi >>24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t,*(constint32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(constfloat *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(constfloat *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(constfloat *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(constdouble *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(constdouble *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(constdouble *) pi *(1U<< 31))))#defineSET_CONV_FUNC_GROUP(ofmt, ifmt) staticvoidset_generic_function(AudioConvert *ac){}voidff_audio_convert_free(AudioConvert **ac){if(!*ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);}AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enumAVSampleFormatout_fmt, enumAVSampleFormatin_fmt, intchannels, intsample_rate, intapply_map){AudioConvert *ac;intin_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) returnNULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method!=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt)>2){ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc){av_free(ac);returnNULL;}returnac;}in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar){ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar?ac->channels:1;}elseif(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;elseac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);returnac;}intff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in){intuse_generic=1;intlen=in->nb_samples;intp;if(ac->dc){av_log(ac->avr, AV_LOG_TRACE,"%dsamples-audio_convert:%sto%s(dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));returnff_convert_dither(ac-> in
void * buf
Definition: avisynth_c.h:690
misc parsing utilities
static int html_color_parse(void *log_ctx, const char *str)
Definition: htmlsubtitles.c:28
common internal and external API header
uint32_t color
Definition: htmlsubtitles.c:74
int len
#define LIKELY_A_TAG_CHAR(x)
GLuint buffer
Definition: opengl_enc.c:102
void av_bprint_chars(AVBPrint *buf, char c, unsigned n)
Append char c n times to a print buffer.
Definition: bprint.c:140