videolan
/
vlc
mirror of https://code.videolan.org/videolan/vlc
1
Fork 0
Code Releases Activity
vlc/modules/codec/subsdec.c

1087 lines
38 KiB
C
Raw Permalink Blame History

/*****************************************************************************
* subsdec.c : text subtitle decoder
*****************************************************************************
* Copyright (C) 2000-2006 VLC authors and VideoLAN
*
* Authors: Gildas Bazin <gbazin@videolan.org>
* Samuel Hocevar <sam@zoy.org>
* Derk-Jan Hartman <hartman at videolan dot org>
* Bernie Purcell <bitmap@videolan.org>
*
* This program 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.
*
* This program 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 this program; if not, write to the Free Software Foundation,
* Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301, USA.
*****************************************************************************/
/*****************************************************************************
* Preamble
*****************************************************************************/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
#include <limits.h>
#include <errno.h>
#include <ctype.h>
#include <vlc_common.h>
#include <vlc_plugin.h>
#include <vlc_codec.h>
#include <vlc_charset.h>
#include <vlc_xml.h>
#include "substext.h"
/*****************************************************************************
* Module descriptor.
*****************************************************************************/
static const char *const ppsz_encodings[] = {
"",
"system",
"UTF-8",
"UTF-16",
"UTF-16BE",
"UTF-16LE",
"GB18030",
"ISO-8859-15",
"Windows-1252",
"IBM850",
"ISO-8859-2",
"Windows-1250",
"ISO-8859-3",
"ISO-8859-10",
"Windows-1251",
"KOI8-R",
"KOI8-U",
"ISO-8859-6",
"Windows-1256",
"ISO-8859-7",
"Windows-1253",
"ISO-8859-8",
"Windows-1255",
"ISO-8859-9",
"Windows-1254",
"ISO-8859-11",
"Windows-874",
"ISO-8859-13",
"Windows-1257",
"ISO-8859-14",
"ISO-8859-16",
"ISO-2022-CN-EXT",
"EUC-CN",
"ISO-2022-JP-2",
"EUC-JP",
"Shift_JIS",
"CP949",
"ISO-2022-KR",
"Big5",
"ISO-2022-TW",
"Big5-HKSCS",
"VISCII",
"Windows-1258",
};
static const char *const ppsz_encoding_names[] = {
/* xgettext:
The character encoding name in parenthesis corresponds to that used for
the GetACP translation. "Windows-1252" applies to Western European
languages using the Latin alphabet. */
N_("Default (Windows-1252)"),
N_("System codeset"),
N_("Universal (UTF-8)"),
N_("Universal (UTF-16)"),
N_("Universal (big endian UTF-16)"),
N_("Universal (little endian UTF-16)"),
N_("Universal, Chinese (GB18030)"),
/* ISO 8859 and the likes */
/* 1 */
N_("Western European (Latin-9)"), /* mostly superset of Latin-1 */
N_("Western European (Windows-1252)"),
N_("Western European (IBM 00850)"),
/* 2 */
N_("Eastern European (Latin-2)"),
N_("Eastern European (Windows-1250)"),
/* 3 */
N_("Esperanto (Latin-3)"),
/* 4 */
N_("Nordic (Latin-6)"), /* Latin 6 supersedes Latin 4 */
/* 5 */
N_("Cyrillic (Windows-1251)"), /* ISO 8859-5 is not practically used */
N_("Russian (KOI8-R)"),
N_("Ukrainian (KOI8-U)"),
/* 6 */
N_("Arabic (ISO 8859-6)"),
N_("Arabic (Windows-1256)"),
/* 7 */
N_("Greek (ISO 8859-7)"),
N_("Greek (Windows-1253)"),
/* 8 */
N_("Hebrew (ISO 8859-8)"),
N_("Hebrew (Windows-1255)"),
/* 9 */
N_("Turkish (ISO 8859-9)"),
N_("Turkish (Windows-1254)"),
/* 10 -> 4 */
/* 11 */
N_("Thai (TIS 620-2533/ISO 8859-11)"),
N_("Thai (Windows-874)"),
/* 13 */
N_("Baltic (Latin-7)"),
N_("Baltic (Windows-1257)"),
/* 12 -> /dev/null */
/* 14 */
N_("Celtic (Latin-8)"),
/* 15 -> 1 */
/* 16 */
N_("South-Eastern European (Latin-10)"),
/* CJK families */
N_("Simplified Chinese (ISO-2022-CN-EXT)"),
N_("Simplified Chinese Unix (EUC-CN)"),
N_("Japanese (7-bits JIS/ISO-2022-JP-2)"),
N_("Japanese Unix (EUC-JP)"),
N_("Japanese (Shift JIS)"),
N_("Korean (EUC-KR/CP949)"),
N_("Korean (ISO-2022-KR)"),
N_("Traditional Chinese (Big5)"),
N_("Traditional Chinese Unix (EUC-TW)"),
N_("Hong-Kong Supplementary (HKSCS)"),
/* Other */
N_("Vietnamese (VISCII)"),
N_("Vietnamese (Windows-1258)"),
};
static const int pi_justification[] = { -1, 0, SUBPICTURE_ALIGN_LEFT, SUBPICTURE_ALIGN_RIGHT };
static const char *const ppsz_justification_text[] = {
N_("Auto"),N_("Center"),N_("Left"),N_("Right")
};
#define ENCODING_TEXT N_("Subtitle text encoding")
#define ENCODING_LONGTEXT N_("Set the encoding used in text subtitles")
#define ALIGN_TEXT N_("Subtitle justification")
#define ALIGN_LONGTEXT N_("Set the justification of subtitles")
#define AUTODETECT_UTF8_TEXT N_("UTF-8 subtitle autodetection")
#define AUTODETECT_UTF8_LONGTEXT N_("This enables automatic detection of " \
"UTF-8 encoding within subtitle files.")
static int OpenDecoder ( vlc_object_t * );
static void CloseDecoder ( vlc_object_t * );
vlc_module_begin ()
set_shortname( N_("Subtitles"))
set_description( N_("Text subtitle decoder") )
set_capability( "spu decoder", 50 )
set_callbacks( OpenDecoder, CloseDecoder )
set_subcategory( SUBCAT_INPUT_SCODEC )
add_integer( "subsdec-align", -1, ALIGN_TEXT, ALIGN_LONGTEXT )
change_integer_list( pi_justification, ppsz_justification_text )
add_string( "subsdec-encoding", "",
ENCODING_TEXT, ENCODING_LONGTEXT )
change_string_list( ppsz_encodings, ppsz_encoding_names )
add_bool( "subsdec-autodetect-utf8", true,
AUTODETECT_UTF8_TEXT, AUTODETECT_UTF8_LONGTEXT )
vlc_module_end ()
/*****************************************************************************
* Local prototypes
*****************************************************************************/
#define NO_BREAKING_SPACE "&#160;"
typedef struct
{
int i_align; /* Subtitles alignment on the vout */
vlc_iconv_t iconv_handle; /* handle to iconv instance */
bool b_autodetect_utf8;
} decoder_sys_t;
static int DecodeBlock ( decoder_t *, block_t * );
static subpicture_t *ParseText ( decoder_t *, block_t * );
static text_segment_t *ParseSubtitles(int *pi_align, const char * );
/*****************************************************************************
* OpenDecoder: probe the decoder and return score
*****************************************************************************
* Tries to launch a decoder and return score so that the interface is able
* to chose.
*****************************************************************************/
static int OpenDecoder( vlc_object_t *p_this )
{
decoder_t *p_dec = (decoder_t*)p_this;
decoder_sys_t *p_sys;
switch( p_dec->fmt_in->i_codec )
{
case VLC_CODEC_SUBT:
case VLC_CODEC_ITU_T140:
break;
default:
return VLC_EGENERIC;
}
/* Allocate the memory needed to store the decoder's structure */
p_dec->p_sys = p_sys = calloc( 1, sizeof( *p_sys ) );
if( p_sys == NULL )
return VLC_ENOMEM;
p_dec->pf_decode = DecodeBlock;
p_dec->fmt_out.i_codec = 0;
/* init of p_sys */
p_sys->i_align = -1;
p_sys->iconv_handle = (vlc_iconv_t)-1;
p_sys->b_autodetect_utf8 = false;
const char *encoding;
char *var = NULL;
/* First try demux-specified encoding */
if( p_dec->fmt_in->i_codec == VLC_CODEC_ITU_T140 )
encoding = "UTF-8"; /* IUT T.140 is always using UTF-8 */
else
if( p_dec->fmt_in->subs.psz_encoding && *p_dec->fmt_in->subs.psz_encoding )
{
encoding = p_dec->fmt_in->subs.psz_encoding;
msg_Dbg (p_dec, "trying demuxer-specified character encoding: %s",
encoding);
}
else
{
/* Second, try configured encoding */
if ((var = var_InheritString (p_dec, "subsdec-encoding")) != NULL)
{
msg_Dbg (p_dec, "trying configured character encoding: %s", var);
if (!strcmp (var, "system"))
{
free (var);
var = NULL;
encoding = "";
/* ^ iconv() treats "" as nl_langinfo(CODESET) */
}
else
encoding = var;
}
else
/* Third, try "local" encoding */
{
/* xgettext:
The Windows ANSI code page most commonly used for this language.
VLC uses this as a guess of the subtitle files character set
(if UTF-8 and UTF-16 autodetection fails).
Western European languages normally use "CP1252", which is a
Microsoft-variant of ISO 8859-1. That suits the Latin alphabet.
Other scripts use other code pages.
This MUST be a valid iconv character set. If unsure, please refer
the VideoLAN translators mailing list. */
encoding = vlc_pgettext("GetACP", "CP1252");
msg_Dbg (p_dec, "trying default character encoding: %s", encoding);
}
/* Check UTF-8 autodetection */
if (var_InheritBool (p_dec, "subsdec-autodetect-utf8"))
{
msg_Dbg (p_dec, "using automatic UTF-8 detection");
p_sys->b_autodetect_utf8 = true;
}
}
if (strcasecmp (encoding, "UTF-8") && strcasecmp (encoding, "utf8"))
{
p_sys->iconv_handle = vlc_iconv_open ("UTF-8", encoding);
if (p_sys->iconv_handle == (vlc_iconv_t)(-1))
msg_Err (p_dec, "cannot convert from %s: %s", encoding,
vlc_strerror_c(errno));
}
free (var);
p_sys->i_align = var_InheritInteger( p_dec, "subsdec-align" );
return VLC_SUCCESS;
}
/****************************************************************************
* DecodeBlock: the whole thing
****************************************************************************
* This function must be fed with complete subtitles units.
****************************************************************************/
static int DecodeBlock( decoder_t *p_dec, block_t *p_block )
{
subpicture_t *p_spu;
if( p_block == NULL ) /* No Drain */
return VLCDEC_SUCCESS;
if( p_block->i_flags & BLOCK_FLAG_CORRUPTED )
{
block_Release( p_block );
return VLCDEC_SUCCESS;
}
p_spu = ParseText( p_dec, p_block );
block_Release( p_block );
if( p_spu != NULL )
decoder_QueueSub( p_dec, p_spu );
return VLCDEC_SUCCESS;
}
/*****************************************************************************
* CloseDecoder: clean up the decoder
*****************************************************************************/
static void CloseDecoder( vlc_object_t *p_this )
{
decoder_t *p_dec = (decoder_t *)p_this;
decoder_sys_t *p_sys = p_dec->p_sys;
if( p_sys->iconv_handle != (vlc_iconv_t)-1 )
vlc_iconv_close( p_sys->iconv_handle );
free( p_sys );
}
/*****************************************************************************
* ParseText: parse an text subtitle packet and send it to the video output
*****************************************************************************/
static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
{
decoder_sys_t *p_sys = p_dec->p_sys;
subpicture_t *p_spu = NULL;
if( p_block->i_flags & BLOCK_FLAG_CORRUPTED )
return NULL;
/* We cannot display a subpicture with no date */
if( p_block->i_pts == VLC_TICK_INVALID )
{
msg_Warn( p_dec, "subtitle without a date" );
return NULL;
}
/* Check validity of packet data */
/* An "empty" line containing only \0 can be used to force
and ephemer picture from the screen */
if( p_block->i_buffer < 1 )
{
msg_Warn( p_dec, "no subtitle data" );
return NULL;
}
char *psz_subtitle = NULL;
/* Should be resiliant against bad subtitles */
if( p_sys->iconv_handle == (vlc_iconv_t)-1 ||
p_sys->b_autodetect_utf8 )
{
psz_subtitle = malloc( p_block->i_buffer + 1 );
if( psz_subtitle == NULL )
return NULL;
memcpy( psz_subtitle, p_block->p_buffer, p_block->i_buffer );
psz_subtitle[p_block->i_buffer] = '\0';
}
if( p_sys->iconv_handle == (vlc_iconv_t)-1 )
{
if (EnsureUTF8( psz_subtitle ) == NULL)
{
msg_Err( p_dec, "failed to convert subtitle encoding.\n"
"Try manually setting a character-encoding "
"before you open the file." );
}
}
else
{
if( p_sys->b_autodetect_utf8 )
{
if( IsUTF8( psz_subtitle ) == NULL )
{
msg_Dbg( p_dec, "invalid UTF-8 sequence: "
"disabling UTF-8 subtitles autodetection" );
p_sys->b_autodetect_utf8 = false;
}
}
if( !p_sys->b_autodetect_utf8 )
{
size_t inbytes_left = p_block->i_buffer;
size_t outbytes_left = 6 * inbytes_left;
char *psz_new_subtitle = xmalloc( outbytes_left + 1 );
char *psz_convert_buffer_out = psz_new_subtitle;
const char *psz_convert_buffer_in =
psz_subtitle ? psz_subtitle : (char *)p_block->p_buffer;
size_t ret = vlc_iconv( p_sys->iconv_handle,
&psz_convert_buffer_in, &inbytes_left,
&psz_convert_buffer_out, &outbytes_left );
*psz_convert_buffer_out++ = '\0';
free( psz_subtitle );
if( ( ret == (size_t)(-1) ) || inbytes_left )
{
free( psz_new_subtitle );
msg_Err( p_dec, "failed to convert subtitle encoding.\n"
"Try manually setting a character-encoding "
"before you open the file." );
return NULL;
}
psz_subtitle = realloc( psz_new_subtitle,
psz_convert_buffer_out - psz_new_subtitle );
if( !psz_subtitle )
psz_subtitle = psz_new_subtitle;
}
}
/* Create the subpicture unit */
p_spu = decoder_NewSubpictureText( p_dec );
if( !p_spu )
{
free( psz_subtitle );
return NULL;
}
p_spu->i_start = p_block->i_pts;
p_spu->i_stop = p_block->i_pts + p_block->i_length;
p_spu->b_ephemer = (p_block->i_length == VLC_TICK_INVALID);
subtext_updater_sys_t *p_spu_sys = p_spu->updater.sys;
int i_inline_align = -1;
p_spu_sys->region.p_segments = ParseSubtitles( &i_inline_align, psz_subtitle );
free( psz_subtitle );
p_spu_sys->region.b_absolute = false;
if( p_sys->i_align >= 0 ) /* bottom ; left, right or centered */
{
p_spu_sys->region.align = SUBPICTURE_ALIGN_BOTTOM | p_sys->i_align;
p_spu_sys->region.inner_align = p_sys->i_align;
}
else if( i_inline_align >= 0 )
{
p_spu_sys->region.align = i_inline_align;
p_spu_sys->region.inner_align = i_inline_align;
}
else /* default, bottom ; centered */
{
p_spu_sys->region.align = SUBPICTURE_ALIGN_BOTTOM;
p_spu_sys->region.inner_align = 0;
}
return p_spu;
}
static bool AppendCharacter( text_segment_t* p_segment, char c )
{
char* tmp;
if ( asprintf( &tmp, "%s%c", p_segment->psz_text ? p_segment->psz_text : "", c ) < 0 )
return false;
free( p_segment->psz_text );
p_segment->psz_text = tmp;
return true;
}
static bool AppendString( text_segment_t* p_segment, const char* psz_str )
{
char* tmp;
if ( asprintf( &tmp, "%s%s", p_segment->psz_text ? p_segment->psz_text : "", psz_str ) < 0 )
return false;
free( p_segment->psz_text );
p_segment->psz_text = tmp;
return true;
}
static char* ConsumeAttribute( const char** ppsz_subtitle, char** ppsz_attribute_value )
{
const char* psz_subtitle = *ppsz_subtitle;
char* psz_attribute_name;
*ppsz_attribute_value = NULL;
while (*psz_subtitle == ' ')
psz_subtitle++;
size_t attr_len = 0;
char delimiter;
while ( *psz_subtitle && isalpha( *psz_subtitle ) )
{
psz_subtitle++;
attr_len++;
}
if ( !*psz_subtitle || attr_len == 0 )
return NULL;
psz_attribute_name = malloc( attr_len + 1 );
if ( unlikely( !psz_attribute_name ) )
return NULL;
strncpy( psz_attribute_name, psz_subtitle - attr_len, attr_len );
psz_attribute_name[attr_len] = 0;
// Skip over to the attribute value
while ( *psz_subtitle && *psz_subtitle != '=' )
psz_subtitle++;
if ( !*psz_subtitle )
{
*ppsz_subtitle = psz_subtitle;
return psz_attribute_name;
}
// Skip the '=' sign
psz_subtitle++;
// Aknoledge the delimiter if any
while ( *psz_subtitle && isspace( *psz_subtitle) )
psz_subtitle++;
if ( *psz_subtitle == '\'' || *psz_subtitle == '"' )
{
// Save the delimiter and skip it
delimiter = *psz_subtitle;
psz_subtitle++;
}
else
delimiter = 0;
// Skip spaces, just in case
while ( *psz_subtitle && isspace( *psz_subtitle ) )
psz_subtitle++;
attr_len = 0;
while ( *psz_subtitle && ( ( delimiter != 0 && *psz_subtitle != delimiter ) ||
( delimiter == 0 && ( !isspace(*psz_subtitle) && *psz_subtitle != '>' ) ) ) )
{
psz_subtitle++;
attr_len++;
}
if ( attr_len == 0 )
{
*ppsz_subtitle = psz_subtitle;
return psz_attribute_name;
}
if ( unlikely( !( *ppsz_attribute_value = malloc( attr_len + 1 ) ) ) )
{
free( psz_attribute_name );
return NULL;
}
strncpy( *ppsz_attribute_value, psz_subtitle - attr_len, attr_len );
(*ppsz_attribute_value)[attr_len] = 0;
// Finally, skip over the final delimiter
if (delimiter != 0 && *psz_subtitle)
psz_subtitle++;
*ppsz_subtitle = psz_subtitle;
return psz_attribute_name;
}
// Returns the next tag and consume the string up to after the tag name, or
// returns NULL and doesn't advance if the angle bracket was not a tag opening
// For instance, if psz_subtitle == "<some_tag attribute=value>"
// GetTag will return "some_tag", and will advance up to the first 'a' in "attribute"
// The returned value must be freed.
static char* GetTag( const char** ppsz_subtitle, bool b_closing )
{
const char* psz_subtitle = *ppsz_subtitle;
if ( *psz_subtitle != '<' )
return NULL;
// Skip the '<'
psz_subtitle++;
if ( b_closing && *psz_subtitle == '/' )
psz_subtitle++;
// Skip potential spaces
while ( *psz_subtitle == ' ' )
psz_subtitle++;
// Now we need to verify if what comes next is a valid tag:
if ( !isalpha( *psz_subtitle ) )
return NULL;
size_t tag_size = 1;
while ( isalnum( psz_subtitle[tag_size] ) || psz_subtitle[tag_size] == '_' )
tag_size++;
char* psz_tagname = vlc_alloc( tag_size + 1, sizeof( *psz_tagname ) );
if ( unlikely( !psz_tagname ) )
return NULL;
strncpy( psz_tagname, psz_subtitle, tag_size );
psz_tagname[tag_size] = 0;
psz_subtitle += tag_size;
*ppsz_subtitle = psz_subtitle;
return psz_tagname;
}
static bool IsClosed( const char* psz_subtitle, const char* psz_tagname )
{
const char* psz_tagpos = strcasestr( psz_subtitle, psz_tagname );
if ( !psz_tagpos )
return false;
// Search for '</' and '>' immediately before & after (minding the potential spaces)
const char* psz_endtag = psz_tagpos + strlen( psz_tagname );
while ( *psz_endtag == ' ' )
psz_endtag++;
if ( *psz_endtag != '>' )
return false;
// Skip back before the tag itself
psz_tagpos--;
while ( *psz_tagpos == ' ' && psz_tagpos > psz_subtitle )
psz_tagpos--;
if ( *psz_tagpos-- != '/' )
return false;
if ( *psz_tagpos != '<' )
return false;
return true;
}
typedef struct tag_stack tag_stack_t;
struct tag_stack
{
char* psz_tagname;
tag_stack_t *p_next;
};
static void AppendTag( tag_stack_t **pp_stack, char* psz_tagname )
{
tag_stack_t* p_elem = malloc( sizeof( *p_elem ) );
if ( unlikely( !p_elem ) )
return;
p_elem->p_next = *pp_stack;
p_elem->psz_tagname = psz_tagname;
*pp_stack = p_elem;
}
static bool HasTag( tag_stack_t **pp_stack, const char* psz_tagname )
{
tag_stack_t *p_prev = NULL;
for ( tag_stack_t* p_current = *pp_stack; p_current; p_current = p_current->p_next )
{
if ( !strcasecmp( psz_tagname, p_current->psz_tagname ) )
{
if ( p_current == *pp_stack )
{
*pp_stack = p_current->p_next;
}
else
{
p_prev->p_next = p_current->p_next;
}
free( p_current->psz_tagname );
free( p_current );
return true;
}
p_prev = p_current;
}
return false;
}
/*
* mini style stack implementation
*/
typedef struct style_stack style_stack_t;
struct style_stack
{
text_style_t* p_style;
style_stack_t* p_next;
};
static text_style_t* DuplicateAndPushStyle(style_stack_t** pp_stack)
{
text_style_t* p_dup = ( *pp_stack ) ? text_style_Duplicate( (*pp_stack)->p_style ) : text_style_Create( STYLE_NO_DEFAULTS );
if ( unlikely( !p_dup ) )
return NULL;
style_stack_t* p_entry = malloc( sizeof( *p_entry ) );
if ( unlikely( !p_entry ) )
{
text_style_Delete( p_dup );
return NULL;
}
// Give the style ownership to the segment.
p_entry->p_style = p_dup;
p_entry->p_next = *pp_stack;
*pp_stack = p_entry;
return p_dup;
}
static void PopStyle(style_stack_t** pp_stack)
{
style_stack_t* p_old = *pp_stack;
if ( !p_old )
return;
*pp_stack = p_old->p_next;
// Don't free the style, it is now owned by the text_segment_t
free( p_old );
}
static text_segment_t* NewTextSegmentPushStyle( text_segment_t* p_segment, style_stack_t** pp_stack )
{
text_segment_t* p_new = text_segment_New( NULL );
if ( unlikely( p_new == NULL ) )
return NULL;
text_style_t* p_style = DuplicateAndPushStyle( pp_stack );
p_new->style = p_style;
p_segment->p_next = p_new;
return p_new;
}
static text_segment_t* NewTextSegmentPopStyle( text_segment_t* p_segment, style_stack_t** pp_stack )
{
text_segment_t* p_new = text_segment_New( NULL );
if ( unlikely( p_new == NULL ) )
return NULL;
// We shouldn't have an empty stack since this happens when closing a tag,
// but better be safe than sorry if (/when) we encounter a broken subtitle file.
PopStyle( pp_stack );
text_style_t* p_dup = ( *pp_stack ) ? text_style_Duplicate( (*pp_stack)->p_style ) : text_style_Create( STYLE_NO_DEFAULTS );
p_new->style = p_dup;
p_segment->p_next = p_new;
return p_new;
}
static text_segment_t* ParseSubtitles( int *pi_align, const char *psz_subtitle )
{
text_segment_t* p_segment;
text_segment_t* p_first_segment;
style_stack_t* p_stack = NULL;
tag_stack_t* p_tag_stack = NULL;
//FIXME: Remove initial allocation? Might make the below code more complicated
p_first_segment = p_segment = text_segment_New( "" );
*pi_align = -1;
/* */
while( *psz_subtitle )
{
/* HTML extensions */
if( *psz_subtitle == '<' )
{
char *psz_tagname = GetTag( &psz_subtitle, false );
if ( psz_tagname != NULL )
{
if( !strcasecmp( psz_tagname, "br" ) )
{
if ( !AppendCharacter( p_segment, '\n' ) )
{
free( psz_tagname );
goto fail;
}
}
else if( !strcasecmp( psz_tagname, "b" ) )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_BOLD;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
}
else if( !strcasecmp( psz_tagname, "i" ) )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_ITALIC;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
}
else if( !strcasecmp( psz_tagname, "u" ) )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_UNDERLINE;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
}
else if( !strcasecmp( psz_tagname, "s" ) )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_STRIKEOUT;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
}
else if( !strcasecmp( psz_tagname, "font" ) )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
char* psz_attribute_name;
char* psz_attribute_value;
while( ( psz_attribute_name = ConsumeAttribute( &psz_subtitle, &psz_attribute_value ) ) )
{
if ( !psz_attribute_value )
{
free( psz_attribute_name );
continue;
}
if ( !strcasecmp( psz_attribute_name, "face" ) )
{
free(p_segment->style->psz_fontname);
p_segment->style->psz_fontname = psz_attribute_value;
// We don't want to free the attribute value since it has become our fontname
psz_attribute_value = NULL;
}
else if ( !strcasecmp( psz_attribute_name, "family" ) )
{
free(p_segment->style->psz_monofontname);
p_segment->style->psz_monofontname = psz_attribute_value;
psz_attribute_value = NULL;
}
else if ( !strcasecmp( psz_attribute_name, "size" ) )
{
int size = atoi( psz_attribute_value );
if( size )
{
p_segment->style->i_font_size = size;
p_segment->style->f_font_relsize = STYLE_DEFAULT_REL_FONT_SIZE *
STYLE_DEFAULT_FONT_SIZE / p_segment->style->i_font_size;
}
}
else if ( !strcasecmp( psz_attribute_name, "color" ) )
{
p_segment->style->i_font_color = vlc_html_color( psz_attribute_value, NULL );
p_segment->style->i_features |= STYLE_HAS_FONT_COLOR;
}
else if ( !strcasecmp( psz_attribute_name, "outline-color" ) )
{
p_segment->style->i_outline_color = vlc_html_color( psz_attribute_value, NULL );
p_segment->style->i_features |= STYLE_HAS_OUTLINE_COLOR;
}
else if ( !strcasecmp( psz_attribute_name, "shadow-color" ) )
{
p_segment->style->i_shadow_color = vlc_html_color( psz_attribute_value, NULL );
p_segment->style->i_features |= STYLE_HAS_SHADOW_COLOR;
}
else if ( !strcasecmp( psz_attribute_name, "outline-level" ) )
{
p_segment->style->i_outline_width = atoi( psz_attribute_value );
}
else if ( !strcasecmp( psz_attribute_name, "shadow-level" ) )
{
p_segment->style->i_shadow_width = atoi( psz_attribute_value );
}
else if ( !strcasecmp( psz_attribute_name, "back-color" ) )
{
p_segment->style->i_background_color = vlc_html_color( psz_attribute_value, NULL );
p_segment->style->i_features |= STYLE_HAS_BACKGROUND_COLOR;
}
else if ( !strcasecmp( psz_attribute_name, "alpha" ) )
{
p_segment->style->i_font_alpha = atoi( psz_attribute_value );
p_segment->style->i_features |= STYLE_HAS_FONT_ALPHA;
}
free( psz_attribute_name );
free( psz_attribute_value );
}
}
else
{
// This is an unknown tag. We need to hide it if it's properly closed, and display it otherwise
if ( !IsClosed( psz_subtitle, psz_tagname ) )
{
AppendCharacter( p_segment, '<' );
AppendString( p_segment, psz_tagname );
AppendCharacter( p_segment, '>' );
}
else
{
AppendTag( &p_tag_stack, psz_tagname );
// We don't want to free the tagname now, it will be freed when the tag
// gets poped from the stack.
psz_tagname = NULL;
}
// In any case, fall through and skip to the closing tag.
}
// Skip potential spaces & end tag
while ( *psz_subtitle && *psz_subtitle != '>' )
psz_subtitle++;
if ( *psz_subtitle == '>' )
psz_subtitle++;
free( psz_tagname );
}
else if( !strncmp( psz_subtitle, "</", 2 ))
{
char* psz_closetagname = GetTag( &psz_subtitle, true );
if ( psz_closetagname != NULL )
{
if ( !strcasecmp( psz_closetagname, "b" ) ||
!strcasecmp( psz_closetagname, "i" ) ||
!strcasecmp( psz_closetagname, "u" ) ||
!strcasecmp( psz_closetagname, "s" ) ||
!strcasecmp( psz_closetagname, "font" ) )
{
// A closing tag for one of the tags we handle, meaning
// we pushed a style onto the stack earlier
p_segment = NewTextSegmentPopStyle( p_segment, &p_stack );
}
else
{
// Unknown closing tag. If it is closing an unknown tag, ignore it. Otherwise, display it
if ( !HasTag( &p_tag_stack, psz_closetagname ) )
{
AppendString( p_segment, "</" );
AppendString( p_segment, psz_closetagname );
AppendCharacter( p_segment, '>' );
}
}
while ( *psz_subtitle == ' ' )
psz_subtitle++;
if ( *psz_subtitle == '>' )
psz_subtitle++;
free( psz_closetagname );
}
else
{
/**
* This doesn't appear to be a valid tag closing syntax.
* Simply append the text
*/
AppendString( p_segment, "</" );
psz_subtitle += 2;
}
}
else
{
/* We have an unknown tag, just append it, and move on.
* The rest of the string won't be recognized as a tag, and
* we will ignore unknown closing tag
*/
AppendCharacter( p_segment, '<' );
psz_subtitle++;
}
}
/* SSA extensions */
else if( psz_subtitle[0] == '{' && psz_subtitle[1] == '\\' &&
strchr( psz_subtitle, '}' ) )
{
/* Check for forced alignment */
if( *pi_align < 0 &&
!strncmp( psz_subtitle, "{\\an", 4 ) && psz_subtitle[4] >= '1' && psz_subtitle[4] <= '9' && psz_subtitle[5] == '}' )
{
static const int pi_vertical[3] = { SUBPICTURE_ALIGN_BOTTOM, 0, SUBPICTURE_ALIGN_TOP };
static const int pi_horizontal[3] = { SUBPICTURE_ALIGN_LEFT, 0, SUBPICTURE_ALIGN_RIGHT };
const int i_id = psz_subtitle[4] - '1';
*pi_align = pi_vertical[i_id/3] | pi_horizontal[i_id%3];
}
/* TODO fr -> rotation */
/* Hide {\stupidity} */
psz_subtitle = strchr( psz_subtitle, '}' ) + 1;
}
/* MicroDVD extensions */
/* FIXME:
* - Currently, we don't do difference between X and x, and we should:
* Capital Letters applies to the whole text and not one line
* - We don't support Position and Coordinates
* - We don't support the DEFAULT flag (HEADER)
*/
else if( psz_subtitle[0] == '{' && psz_subtitle[1] != 0 &&
psz_subtitle[2] == ':' && strchr( &psz_subtitle[2], '}' ) )
{
const char *psz_tag_end = strchr( &psz_subtitle[2], '}' );
size_t i_len = psz_tag_end - &psz_subtitle[3];
if( psz_subtitle[1] == 'Y' || psz_subtitle[1] == 'y' )
{
if( psz_subtitle[3] == 'i' )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_ITALIC;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
psz_subtitle++;
}
if( psz_subtitle[3] == 'b' )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_BOLD;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
psz_subtitle++;
}
if( psz_subtitle[3] == 'u' )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_style_flags |= STYLE_UNDERLINE;
p_segment->style->i_features |= STYLE_HAS_FLAGS;
psz_subtitle++;
}
}
else if( (psz_subtitle[1] == 'C' || psz_subtitle[1] == 'c' )
&& psz_subtitle[3] == '$' && i_len >= 7 )
{
/* Yes, they use BBGGRR, instead of RRGGBB */
char psz_color[7];
psz_color[0] = psz_subtitle[8]; psz_color[1] = psz_subtitle[9];
psz_color[2] = psz_subtitle[6]; psz_color[3] = psz_subtitle[7];
psz_color[4] = psz_subtitle[4]; psz_color[5] = psz_subtitle[5];
psz_color[6] = '\0';
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_font_color = vlc_html_color( psz_color, NULL );
p_segment->style->i_features |= STYLE_HAS_FONT_COLOR;
}
else if( psz_subtitle[1] == 'F' || psz_subtitle[1] == 'f' )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
free(p_segment->style->psz_fontname);
p_segment->style->psz_fontname = strndup( &psz_subtitle[3], i_len );
}
else if( psz_subtitle[1] == 'S' || psz_subtitle[1] == 's' )
{
int size = atoi( &psz_subtitle[3] );
if( size )
{
p_segment = NewTextSegmentPushStyle( p_segment, &p_stack );
p_segment->style->i_font_size = size;
p_segment->style->f_font_relsize = STYLE_DEFAULT_REL_FONT_SIZE *
STYLE_DEFAULT_FONT_SIZE / p_segment->style->i_font_size;
}
}
/* Currently unsupported since we don't have access to the i_align flag here
else if( psz_subtitle[1] == 'P' )
{
if( psz_subtitle[3] == "1" )
i_align = SUBPICTURE_ALIGN_TOP;
else if( psz_subtitle[3] == "0" )
i_align = SUBPICTURE_ALIGN_BOTTOM;
} */
// Hide other {x:y} atrocities, notably {o:x}
psz_subtitle = psz_tag_end + 1;
}
else
{
if( *psz_subtitle == '\n' || !strncasecmp( psz_subtitle, "\\n", 2 ) )
{
if ( !AppendCharacter( p_segment, '\n' ) )
goto fail;
if ( *psz_subtitle == '\n' )
psz_subtitle++;
else
psz_subtitle += 2;
}
else if( !strncasecmp( psz_subtitle, "\\h", 2 ) )
{
if ( !AppendString( p_segment, "\xC2\xA0" ) )
goto fail;
psz_subtitle += 2;
}
else
{
//FIXME: Highly inneficient
AppendCharacter( p_segment, *psz_subtitle );
psz_subtitle++;
}
}
}
while ( p_stack )
PopStyle( &p_stack );
while ( p_tag_stack )
{
tag_stack_t *p_tag = p_tag_stack;
p_tag_stack = p_tag_stack->p_next;
free( p_tag->psz_tagname );
free( p_tag );
}
return p_first_segment;
fail:
text_segment_ChainDelete( p_first_segment );
return NULL;
}
View Git Blame Copy Permalink