/*
* Copyright (c) 2010 Aurelien Jacobs <aurel@gnuage.org>
* Copyright (c) 2017 Clément Bœsch <u@pkh.me>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "libavutil/avassert.h"
#include "libavutil/avstring.h"
#include "libavutil/common.h"
#include "libavutil/parseutils.h"
#include "htmlsubtitles.h"
#include <ctype.h>
static int html_color_parse(void *log_ctx, const char *str)
{
uint8_t rgba[4];
int nb_sharps = 0;
while (str[nb_sharps] == '#')
nb_sharps++;
str += FFMAX(0, nb_sharps - 1);
if (av_parse_color(rgba, str, strcspn(str, "\" >"), log_ctx) < 0)
return -1;
return rgba[0] | rgba[1] << 8 | rgba[2] << 16;
}
static void rstrip_spaces_buf(AVBPrint *buf)
{
if (av_bprint_is_complete(buf))
while (buf->len > 0 && buf->str[buf->len - 1] == ' ')
buf->str[--buf->len] = 0;
}
/*
* Fast code for scanning text enclosed in braces. Functionally
* equivalent to this sscanf call:
*
* sscanf(in, "{\\an%*1u}%n", &len) >= 0 && len > 0
*/
static int scanbraces(const char* in) {
if (strncmp(in, "{\\an", 4) != 0) {
return 0;
}
if (!av_isdigit(in[4])) {
return 0;
}
if (in[5] != '}') {
return 0;
}
return 1;
}
/* skip all {\xxx} substrings except for {\an%d}
and all microdvd like styles such as {Y:xxx} */
static void handle_open_brace(AVBPrint *dst, const char **inp, int *an, int *closing_brace_missing)
{
const char *in = *inp;
*an += scanbraces(in);
if (!*closing_brace_missing) {
if ( (*an != 1 && in[1] == '\\')
|| (in[1] && strchr("CcFfoPSsYy", in[1]) && in[2] == ':')) {
char *bracep = strchr(in+2, '}');
if (bracep) {
*inp = bracep;
return;
} else
*closing_brace_missing = 1;
}
}
av_bprint_chars(dst, *in, 1);
}
struct font_tag {
char face[128];
int size;
uint32_t color;
};
/*
* Fast code for scanning the rest of a tag. Functionally equivalent to
* this sscanf call:
*
* sscanf(in, "%127[^<>]>%n", buffer, lenp) == 2
*/
static int scantag(const char* in, char* buffer, int* lenp) {
int len;
for (len = 0; len < 128; len++) {
const char c = *in++;
switch (c) {
case '\0':
return 0;
case '<':
return 0;
case '>':
buffer[len] = '\0';
*lenp = len+1;
return 1;
default:
break;
}
buffer[len] = c;
}
return 0;
}
/*
* The general politic of the convert is to mask unsupported tags or formatting
* errors (but still alert the user/subtitles writer with an error/warning)
* without dropping any actual text content for the final user.
*/
int ff_htmlmarkup_to_ass(void *log_ctx, AVBPrint *dst, const char *in)
{
char *param, buffer[128];
int len, tag_close, sptr = 0, line_start = 1, an = 0, end = 0;
int closing_brace_missing = 0;
int i, likely_a_tag;
/*
* state stack is only present for fonts since they are the only tags where
* the state is not binary. Here is a typical use case:
*
* <font color="red" size=10>
* red 10
* <font size=50> RED AND BIG </font>
* red 10 again
* </font>
*
* On the other hand, using the state system for all the tags should be
* avoided because it breaks wrongly nested tags such as:
*
* <b> foo <i> bar </b> bla </i>
*
* We don't want to break here; instead, we will treat all these tags as
* binary state markers. Basically, "<b>" will activate bold, and "</b>"
* will deactivate it, whatever the current state.
*
* This will also prevents cases where we have a random closing tag
* remaining after the opening one was dropped. Yes, this happens and we
* still don't want to print a "</b>" at the end of the dialog event.
*/
struct font_tag stack[16];
memset(&stack[0], 0, sizeof(stack[0]));
for (; !end && *in; in++) {
switch (*in) {
case '\r':
break;
case '\n':
if (line_start) {
end = 1;
break;
}
rstrip_spaces_buf(dst);
av_bprintf(dst, "\\N");
line_start = 1;
break;
case ' ':
if (!line_start)
av_bprint_chars(dst, *in, 1);
break;
case '{':
handle_open_brace(dst, &in, &an, &closing_brace_missing);
break;
case '<':
/*
* "<<" are likely latin guillemets in ASCII or some kind of random
* style effect; see sub/badsyntax.srt in the FATE samples
* directory for real test cases.
*/
likely_a_tag = 1;
for (i = 0; in[1] == '<'; i++) {
av_bprint_chars(dst, '<', 1);
likely_a_tag = 0;
in++;
}
tag_close = in[1] == '/';
if (tag_close)
likely_a_tag = 1;
av_assert0(in[0] == '<');
len = 0;
if (scantag(in+tag_close+1, buffer, &len) && len > 0) {
const int skip = len + tag_close;
const char *tagname = buffer;
while (*tagname == ' ') {
likely_a_tag = 0;
tagname++;
}
if ((param = strchr(tagname, ' ')))
*param++ = 0;
/* Check if this is likely a tag */
#define LIKELY_A_TAG_CHAR(x) (((x) >= '0' && (x) <= '9') || \
((x) >= 'a' && (x) <= 'z') || \
((x) >= 'A' && (x) <= 'Z') || \
(x) == '_' || (x) == '/')
for (i = 0; tagname[i]; i++) {
if (!LIKELY_A_TAG_CHAR(tagname[i])) {
likely_a_tag = 0;
break;
}
}
if (!av_strcasecmp(tagname, "font")) {
if (tag_close && sptr > 0) {
struct font_tag *cur_tag = &stack[sptr--];
struct font_tag *last_tag = &stack[sptr];
if (cur_tag->size) {
if (!last_tag->size)
av_bprintf(dst, "{\\fs}");
else if (last_tag->size != cur_tag->size)
av_bprintf(dst, "{\\fs%d}", last_tag->size);
}
if (cur_tag->color & 0xff000000) {
if (!(last_tag->color & 0xff000000))
av_bprintf(dst, "{\\c}");
else if (last_tag->color != cur_tag->color)
av_bprintf(dst, "{\\c&H%"PRIX32"&}", last_tag->color & 0xffffff);
}
if (cur_tag->face[0]) {
if (!last_tag->face[0])
av_bprintf(dst, "{\\fn}");
else if (strcmp(last_tag->face, cur_tag->face))
av_bprintf(dst, "{\\fn%s}", last_tag->face);
}
} else if (!tag_close && sptr < FF_ARRAY_ELEMS(stack) - 1) {
struct font_tag *new_tag = &stack[sptr + 1];
*new_tag = stack[sptr++];
while (param) {
if (!av_strncasecmp(param, "size=", 5)) {
param += 5 + (param[5] == '"');
if (sscanf(param, "%u", &new_tag->size) == 1)
av_bprintf(dst, "{\\fs%u}", new_tag->size);
} else if (!av_strncasecmp(param, "color=", 6)) {
int color;
param += 6 + (param[6] == '"');
color = html_color_parse(log_ctx, param);
if (color >= 0) {
new_tag->color = 0xff000000 | color;
av_bprintf(dst, "{\\c&H%"PRIX32"&}", new_tag->color & 0xffffff);
}
} else if (!av_strncasecmp(param, "face=", 5)) {
param += 5 + (param[5] == '"');
len = strcspn(param,
param[-1] == '"' ? "\"" :" ");
av_strlcpy(new_tag->face, param,
FFMIN(sizeof(new_tag->face), len+1));
param += len;
av_bprintf(dst, "{\\fn%s}", new_tag->face);
}
if ((param = strchr(param, ' ')))
param++;
}
}
in += skip;
} else if (tagname[0] && !tagname[1] && strchr("bisu", av_tolower(tagname[0]))) {
av_bprintf(dst, "{\\%c%d}", (char)av_tolower(tagname[0]), !tag_close);
in += skip;
} else if (!av_strncasecmp(tagname, "br", 2) &&
(!tagname[2] || (tagname[2] == '/' && !tagname[3]))) {
av_bprintf(dst, "\\N");
in += skip;
} else if (likely_a_tag) {
if (!tag_close) // warn only once
av_log(log_ctx, AV_LOG_WARNING, "Unrecognized tag %s\n", tagname);
in += skip;
} else {
av_bprint_chars(dst, '<', 1);
}
} else {
av_bprint_chars(dst, *in, 1);
}
break;
default:
av_bprint_chars(dst, *in, 1);
break;
}
if (*in != ' ' && *in != '\r' && *in != '\n')
line_start = 0;
}
if (!av_bprint_is_complete(dst))
return AVERROR(ENOMEM);
while (dst->len >= 2 && !strncmp(&dst->str[dst->len - 2], "\\N", 2))
dst->len -= 2;
dst->str[dst->len] = 0;
rstrip_spaces_buf(dst);
return 0;
}