libsubtitles-perl 1.04-1 / usr / share / perl5 / Subtitles.pm

# $Id: Subtitles.pm,v 1.22 2012/02/14 13:21:48 dk Exp $
package Subtitles;
use strict;
require Exporter;
use vars qw(@ISA @EXPORT @EXPORT_OK @codecs $VERSION);
@ISA = qw(Exporter);
@EXPORT = qw(codecs time2str);
@EXPORT_OK = qw(codecs time2hms time2shms hms2time time2str);
$VERSION = '1.04';
		
use Encode;


push @codecs, map { "Subtitles::Codec::$_" } qw( srt mdvd sub2 smi idx);

#
# package-oriented API
#

sub time2hms
{
	shift if $#_ == 1; # package and object
	my $time = $_[0];
	$time = 0 if $time < 0;
	$time += .0005;
	return int($time/3600),int(($time%3600)/60),int($time%60),int(($time-int($time))*1000),
}

sub time2shms
{
	shift if $#_ == 1; # package and object
	my $time = $_[0];
	my $sign;
	if ( $time < 0) {
		$sign = -1;
		$time = -$time;
	} else {
		$sign = 1;
	}
	$time += .0005;
	return $sign,int($time/3600),int(($time%3600)/60),int($time%60),int(($time-int($time))*1000),
}

sub hms2time
{
	shift if $#_ == 4; # package and object
	my ( $h, $m, $s, $ms) = @_;
	return $h * 3600 + $m * 60 + $s + $ms / 1000;
}


sub time2str
{
	shift if $#_ == 1; # package and object
	my $time = $_[0];
	my $is_minus = '';
	$time = -$time, $is_minus = "-" if $time < 0;
	return sprintf ( "$is_minus%02d:%02d:%02d.%03d", time2hms($time));
}

sub codecs { @codecs }

#
# object-oriented API
#

sub new
{
	my $class = shift;
	return bless {
		codec  => undef,
		@_,
		text   => [],
		from   => [],
		to     => [],
		class  => $class,
	}, $class;
}

sub load
{
	my ( $self, $fh, $codec) = @_;
	$self-> clear;
	local $/;
	my $content = <$fh>;
	if ( $content =~ s/^(\xff\xfe|\xfe\xff)//) {
		# found a 16-bit bom
		my $le = ( $1 eq "\xff\xfe" ) ? 'v*' : 'n*';
		$content = join('', map { chr } unpack($le, $content));
	} elsif ( $content =~ s/^\xef\xbb\xbf//) {
		# found a utf-8 bom
		Encode::_utf8_on($content);
	}
	my @content;
	for ( split "\n", $content) {
		s/[\s\n\r]+$//;
		push @content, $_;
	}
	unless ( defined $codec) {
		for ( @content) {
			my $line = $_;
			for ( @codecs) {
				next unless $_-> match( $line);
				$codec = $_;
			}
		}
	}
	unless ( defined $codec) {
		$@ = "No suitable codec is found";
		return undef;
	}
	my $ret;
	eval {
		$ret = $codec-> read( $self, \@content);
	};
	return undef if $@ or !defined $ret;
	# validate
	if ( @{$self->{from}} == 0) {
		$@ = "Empty subtitle";
		return undef;
	}
	if ( @{$self->{from}} != @{$self->{to}}) {
		if ( @{$self->{from}} == @{$self->{to}} + 1) {
			push @{$self->{to}}, $self->{from}->[-1] + 2; # fix a dangling tail 
		} else {
			my $a = @{$self->{from}}; 
			my $b = @{$self->{to}}; 
			$@ = "Number of 'from' ($a) and 'to' ($b) timeframe positions is different";
			return undef;
		}
	}
	if ( @{$self->{from}} != @{$self->{text}}) {
		if ( @{$self->{from}} == @{$self->{text}} + 1) {
			push @{$self->{text}}, ''; # fix a dangling tail
		} else {
			my $a = @{$self->{from}}; 
			my $b = @{$self->{text}}; 
			$@ = "Number of timeframes ($a) is different from the number of text lines ($b)";
			return undef;
		}
	}
	$self->{codec} = $codec;
	return 1;
}

sub codec
{
	return $_[0]-> {codec} unless $#_;
	my ( $self, $codec) = @_;
	my %c = map { $_ => 1 } @codecs;
	return unless exists $c{$codec};
	return if defined $self->{codec} && $self->{codec} eq $codec;
	$self->{codec}-> downgrade($self, $codec) if defined $self->{codec};
	$self->{codec} = $codec;
}

sub rate
{
	return $_[0]-> {rate} unless $#_;
	return if defined $_[1] && $_[1] <= 0;
	$_[0]->{rate} = $_[1];
}

# parses 
#   SS
#   MM:SS
#   HH:MM:SS
#   HH:MM:SS,msec
#   MM:SS,msec
# into time
sub parse_time
{
	my ( $self, $time) = @_;
	my $sign = 1;
	$sign = -1 if $time =~ s/^-//;
	if ( $time =~ m/^(?:(\d{1,2}):)?(?:(\d{1,2}):)?(\d{1,2})(?:[\,\.\:](\d{1,3}))?$/) {
		my ( $h, $m, $s, $ms) = ( $1, $2, $3, $4);
		( $h, $m) = ( $m, $h) if defined $h && ! defined $m;
		$h  = 0 unless defined $h;
		$m  = 0 unless defined $m;
		$ms = '0' unless defined $ms;
		$ms .= '0' while length($ms) < 3;
		return $sign * ( $h * 3600 + $m * 60 + $s + $ms / 1000);
	} elsif ( $self && $self-> {codec}) {
		my $t = $self->{codec}->time($time);
		return $sign * $t if defined $t;
	}
	undef;
}

sub shift { $_[0]-> transform( 1, $_[1]) }
sub scale { $_[0]-> transform( $_[1], 0) }

sub lines { scalar @{$_[0]->{text}} }

# applies linear (y = ax+b) transformation within a scope
sub transform
{
	my ( $self, $a, $b, $qfrom, $qto) = @_;
	return if $a == 1 && $b == 0;
	$qfrom = 0 unless defined $qfrom;
	$qto   = $self->{to}->[-1] unless defined $qto;
	my $i;
	my $n = $self-> lines;
	my $from = $self->{from};
	my $to   = $self->{to};
	for ( $i = 0; $i < $n; $i++) {
		next if $$from[$i] > $qto || $$to[$i] < $qfrom;
		$$from[$i] = $a * $$from[$i] + $b;
		$$to[$i]   = $a * $$to[$i] + $b;
	}
}

sub dup
{
	my ( $self, $clear) = @_;
	if ( $clear) {
		return bless { 
		%$self,
		text => [],
		from => [],
		to   => [],
		}, $self-> {class};
	} else {
		return bless { 
		%$self,
		text => [ @{$self->{text}}],
		from => [ @{$self->{from}}],
		to   => [ @{$self->{to}}],
		}, $self-> {class};
	}
}

sub clear
{
	my $self = $_[0];
	$self-> {text} = [];
	$self-> {from} = [];
	$self-> {to}   = [];
}

sub join
{
	my ( $self, $guest, $time_between) = @_;
	$time_between = 2 unless defined $time_between;
	my $delta = $time_between + $self-> length;
	push @{$self->{text}}, @{$guest->{text}};
	push @{$self->{from}}, map { $_ + $delta } @{$guest->{from}};
	push @{$self->{to}},   map { $_ + $delta } @{$guest->{to}};
}

sub split
{
	my ( $self, $where) = @_;

	my ( $s1, $s2) = ( $self-> dup(1), $self-> dup(1));

	my $i;
	my $n = $self->lines;
	my $t = $self->{to};
	my ( $end, $begin);

	$end = $n - 1;
	for ( $i = 0; $i < $n; $i++) {
		next if $$t[$i] <= $where;
		$begin = $i;
		$end = $i - 1;
		last;
	}

	if ( defined $end && $end >= 0) {
		@{$s1->{text}} = @{$self->{text}}[0..$end];
		@{$s1->{from}} = @{$self->{from}}[0..$end];
		@{$s1->{to}}   = @{$self->{to}}[0..$end];
	}
	if ( defined $begin && $begin < $n) {
		@{$s2->{text}} = @{$self->{text}}[$begin..$n-1];
		@{$s2->{from}} = @{$self->{from}}[$begin..$n-1];
		@{$s2->{to}}   = @{$self->{to}}[$begin..$n-1];
		$s2-> shift( -$where);
	}
	($s1,$s2);
}

sub length
{
	my $self = $_[0];
	return @{$self->{to}} ? $self->{to}->[-1] : 0;
}

sub save
{
	my ( $self, $fh) = @_;
	my $content;
	eval {
		$content = $self-> {codec}-> write( $self);
		die "no content" unless defined $content and @$content;

		$content = CORE::join("\n", @$content);
		if ( Encode::is_utf8($content)) {
			# bomify
			print $fh "\xef\xbb\xbf" or die "write error:$!";
			binmode $fh, ':utf8';
		}

		print $fh $content, "\n" or die "write error:$!"; 
	};

	return $@ ? 0 : 1;
}

package Subtitles::Codec;
use vars qw(@ISA);

sub match 
{
	my ( $self, $line) = @_;
	undef;
}

sub read 
{
	my ( $self, $sub, $content) = @_;
	die "abstract method call";
}

sub write 
{
	my ( $self, $sub) = @_;
	die "abstract method call";
}

sub time { undef }

sub downgrade {}

package Subtitles::Codec::srt;
use vars qw(@ISA);
@ISA=qw(Subtitles::Codec);

sub match
{
	$_[1] =~ m/^(\d\d):(\d\d):(\d\d)[.,](\d\d\d)\s*-->\s*(\d\d):(\d\d):(\d\d)[.,](\d\d\d)/;
}

sub read
{
	my ( $self, $sub, $content) = @_;

	my $stage = 0;
	my $num = 1;
	my $line = 0;
# 0:   
# 1: 1
# 2: 00:00:04,073 --> 00:00:05,781
# 3: Subtitle

	for ( @$content) {
		$line++;
		if ( $stage == 0) {
			next unless length;
			die "Invalid line numbering at line $line\n" unless m/^\d+$/;
			$num++;
			$stage++;
		} elsif ( $stage == 1) {
			die "Invalid timing at line $line\n" unless
				m/^(\d\d):(\d\d):(\d\d)[.,](\d\d\d)\s*-->\s*(\d\d):(\d\d):(\d\d)[.,](\d\d\d)/;
			push @{$sub->{from}}, Subtitles::hms2time( $1, $2, $3, $4);
			push @{$sub->{to}},   Subtitles::hms2time( $5, $6, $7, $8); 
			$stage++;
		} elsif ( $stage == 2) {
			if ( length) {
				push @{$sub->{text}}, $_;
				$stage++;
			} else {
				push @{$sub->{text}}, '';
				$stage = 0;
			}
		} else {
			if ( length) {
				$sub->{text}->[-1] .= "\n$_";
			} else {
				$stage = 0;
			}
		}
	}
	1;
}

sub write
{
	my ( $self, $sub) = @_;

	my $n = @{$sub->{text}};
	my $i;
	my @ret;
	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	for ( $i = 0; $i < $n; $i++) {
		push @ret, 
			$i + 1,
			sprintf ( "%02d:%02d:%02d,%03d --> %02d:%02d:%02d,%03d",
				Subtitles::time2hms($from->[$i]),
				Subtitles::time2hms($to->[$i]),
			),
			split ("\n", $text->[$i]),
			''
		;
	}
	\@ret;
}

package Subtitles::Codec::mdvd;
use vars qw(@ISA);
@ISA=qw(Subtitles::Codec);

sub match
{
	$_[1] =~ m/^[{\[]\d+[}\]][{\[]\d*[}\]]/; 
}

sub read
{
	my ( $self, $sub, $content) = @_;

	my $line = 0;
# {3724}{3774}Text

	my $fps = $sub->{rate} ? $sub->{rate} : 23.976;
	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};

	for ( @$content) {
		$line++;
		unless ( m/^[{\[](\d+)[}\]][{\[](\d*)[}\]](.*)$/) {
			warn "Invalid input at line $line\n";
			next;
		}
		push @$from, $1/$fps;
		push @$to,   length($2) ? ($2/$fps) : ($1+1)/$fps;
		my $t = $3;
		$t=~ s/\|\s*/\n/g;
		push @$text, $t;
	}
	1;
}

sub write
{
	my ( $self, $sub) = @_;
	
	my $fps = $sub->{rate} ? $sub->{rate} : 23.976;

	my $n = @{$sub->{text}};
	my $i;
	my @ret;
	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	for ( $i = 0; $i < $n; $i++) {
		my $t = $text->[$i];
		$t =~ s/\n/\|/g;
		push @ret, 
			sprintf ( "{%d}{%d}%s",
				int( $from->[$i] * $fps + .5),
				int( $to->[$i]   * $fps + .5),
				$t
			);
	}
	\@ret;
}

package Subtitles::Codec::sub2;
use vars qw(@ISA);
@ISA=qw(Subtitles::Codec);

sub match
{
	$_[1] =~ m/^\[(SUBTITLE|COLF)\]/i or
	$_[1] =~ m/^(\d\d):(\d\d):(\d\d)\.(\d\d),(\d\d):(\d\d):(\d\d)\.(\d\d)/;
}

sub read
{
	my ( $self, $sub, $content) = @_;

	my $line = 0;
# [INFORMATION]
# [AUTHOR]
# [SOURCE]
# [PRG]
# [FILEPATH]
# [DELAY]
# [CD TRACK]
# [COMMENT]
# [END INFORMATION]
# 
# [SUBTITLE]
# [COLF]&HFFFFFF,[STYLE]no,[SIZE]18,[FONT]Arial
# 00:04:10.26,00:04:13.57
# Welcome to Gattaca.

	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	my @header;

	my $read_header = 1;
	my $state = 0;

	for ( @$content) {
		$line++;
		if ( $read_header) {
			if ( m/^(\d\d):(\d\d):(\d\d)\.(\d\d)\,(\d\d):(\d\d):(\d\d)\.(\d\d)/) {
				$read_header = 0;
				goto BODY;
			}
			push @header, $_;
		} else {
		BODY:
			if ( $state == 0) {
				next unless length;
				die "Invalid timing at line $line\n" unless
					m/^(\d\d):(\d\d):(\d\d)\.(\d\d)\,(\d\d):(\d\d):(\d\d)\.(\d\d)/;
				push @$from, Subtitles::hms2time( $1, $2, $3, $4 * 10);
				push @$to,   Subtitles::hms2time( $5, $6, $7, $8 * 10); 
				$state = 1;
			} else {
				s/\[br\]\s*/\n/g;
				push @$text, $_;
				$state = 0;
			}
		}
	}

	$sub->{sub2}->{header} = \@header;
	1;
}

sub write
{
	my ( $self, $sub) = @_;

	my $n = @{$sub->{text}};
	my $i;
	my @ret;
	if ( $sub->{sub2}->{header}) {
		@ret = @{$sub->{sub2}->{header}};
	} else {
	@ret = split "\n", <<HEADER;
[INFORMATION]
[AUTHOR]
[SOURCE]
[PRG]
[FILEPATH]
[DELAY]
[CD TRACK]
[COMMENT]
[END INFORMATION]

[SUBTITLE]
[STYLE]no,[SIZE]18
HEADER
	}
	
	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	for ( $i = 0; $i < $n; $i++) {
		my ($fh,$fm,$fs,$fms) = Subtitles::time2hms($from->[$i]);
		my ($th,$tm,$ts,$tms) = Subtitles::time2hms($to->[$i]);
		$fms = int ( $fms / 10);
		$tms = int ( $tms / 10);
		my $t = $text->[$i];
		$t =~ s/\n/[br]/g;
		push @ret, 
			sprintf ( "%02d:%02d:%02d.%02d,%02d:%02d:%02d.%02d",
			$fh,$fm,$fs,$fms,
			$th,$tm,$ts,$tms
			),
			$t,
			''
		;
	}
	\@ret;
}

package Subtitles::Codec::smi;
use vars qw(@ISA);
@ISA=qw(Subtitles::Codec);

sub match
{
	$_[1] =~ m/^<SAMI>/i;
}

sub read
{
	my ( $self, $sub, $content) = @_;

# <SAMI>
# <HEAD>
#    <STYLE TYPE="Text/css">
#    <!--
#       P {margin-left: 29pt; margin-right: 29pt; font-size: 24pt; text-align: center; font-family: Tahoma; font-weight: bold; color: #FFFFFF; background-color: #FFFFFF;}
#       .SUBTTL {Name: 'Subtitles'; Lang: en-US; SAMIType: CC;}
#    -->
#    </STYLE>
# </HEAD>
# <BODY>
#    <SYNC START=2002>
#       <P CLASS=SUBTTL>Juon - A curse born of a strong grudge held by someone<br>who died.
#   <SYNC START=7707>
#      <P CLASS=SUBTTL>&nbsp;
#   <SYNC START=7741>
# </BODY>
# </SAMI>

	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	my (@header,@footer);

	my $read_header = 1;
	my $read_footer = 0;

	my $body = '';

	# extract body to inspect closer
	for ( @$content) {
		if ( $read_header) {
			if ( m/<BODY>/i) {
				$read_header = 0;
			}
			push @header, $_;
		} elsif ( $read_footer) {
			push @footer, $_;
		} else {
			if ( m/<\/BODY>/) {
				push @footer, $_;
				$read_footer = 1; 
				next;
			}
			$body .= $_;
		}
	}
	
	# parse body
	my $sync = 0;
	my $line = '';
	while ( $body =~ m/\G(?:(?:(\s*)<\s*([^\>]*)\s*>)|([^<>]*))/gcs) {
		if ( defined $2 and length $2) {
			my $t = $1;
			$_ = $2;
			if ( m/^sync\s+start\s*=\s*(\d+)/i) {
				$sub->{smi}->{s1gap} = length $t
					unless defined $sub->{smi}->{s1gap};
				my $s = $1;
				die "Inconsistency near '$_' ( is less than previous sync $sync )\n"
					if $s < $sync;
				if ( $line !~ /^[\n\s]*$/s) {
					$line =~ s/[\n\s]+$//s;
					push @$from, $sync / 1000;
					push @$to,   $s / 1000;
					push @$text, $line;
				}
				$sync = $s;
				$line = '';
			} elsif ( m/^p\s+class\s*\=\s*(\S+)/i) {
				$sub->{smi}->{s2gap} = length $t
					unless defined $sub->{smi}->{s2gap};
				$sub-> {smi}-> {class} = $1
					unless defined $sub->{smi}->{class};
			} elsif ( m/^\s*br\s*/i) {
				$line .= "\n";
			}
		} elsif ( defined $3 and length $3) {
			$_ = $3;
			s/&nsbp;/ /g;
			$line .= $_;
		}
	}

	$sub->{smi}->{header} = \@header;
	$sub->{smi}->{footer} = \@footer;
	return 1;
}

sub write
{
	my ( $self, $sub) = @_;

	my $n = @{$sub->{text}};
	my $i;
	my @ret;
	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};

	my $smi_class = defined ($sub->{smi}->{class}) ? $sub->{smi}->{class} : 'SUBTTL';
	if ( $sub->{smi}->{header}) {
		@ret = @{$sub->{smi}->{header}};
	} else {
	@ret = split "\n", <<HEADER;
<SAMI>
<HEAD>
   <STYLE TYPE="Text/css">
   <!--
      P {
       margin-left: 29pt; 
       margin-right: 29pt; 
       font-size: 24pt; 
       text-align: center; 
       font-family: Tahoma; 
       font-weight: bold; 
       color: #FFFFFF; 
       background-color: #FFFFFF;
     }
     .SUBTTL {Name: 'Subtitles'; Lang: en-US; SAMIType: CC;}
   -->
   </STYLE>
</HEAD>
<BODY>
HEADER
	}

	my $s1 = ' ' x ( $sub->{smi}->{s1gap} || 0);
	my $s2 = ' ' x ( $sub->{smi}->{s2gap} || 0);
	for ( $i = 0; $i < $n; $i++) {
		my $f = int($$from[$i] * 1000 + .5);
		my $t = int($$to[$i] * 1000 + .5);
		my $x = $$text[$i];
		$x =~ s/\n/<br>/g;
		push @ret, 
			"$s1<SYNC Start=$f>",
			"$s2<P Class=$smi_class>$x";
		push @ret,
			"$s1<SYNC Start=$t>",
			"$s2<P Class=$smi_class>&nbsp;"
			if $i == $n - 1 || int($$from[$i+1] * 1000 + .5) != $t;
		;
	}
	if ( $sub->{smi}->{footer}) {
		push @ret, @{$sub->{smi}->{footer}};
	} else {
		push @ret, split "\n", <<FOOTER;
</BODY>
</SAMI>
FOOTER
	}
	\@ret;
}

sub downgrade
{
	for ( @{$_[1]->{text}}) {
		s/<[^\>]*>//g;
		s/{[^\}]*}//g;
	}
}

package Subtitles::Codec::idx;
use vars qw(@ISA);
@ISA=qw(Subtitles::Codec);

sub match
{
	$_[1] =~ m/^\s*\#\s*VobSub index file/
}

sub read
{
	my ( $self, $sub, $content) = @_;

	my $line = 0;
# # VobSub index file, v7 (do not modify this line!)
# # 
# # To repair desyncronization, you can insert gaps this way:
# # (it usually happens after vob id changes)
# # 
# #	 delay: [sign]hh:mm:ss:ms
# # 
# # Where:
# #	 [sign]: +, - (optional)
# #	 hh: hours (0 <= hh)
# #	 mm/ss: minutes/seconds (0 <= mm/ss <= 59)
# #	 ms: milliseconds (0 <= ms <= 999)
# # 
# #	 Note: You can't position a sub before the previous with a negative value.
# # 
# # You can also modify timestamps or delete a few subs you don't like.
# # Just make sure they stay in increasing order.
# 
# 
# # Settings
# 
# # Original frame size
# size: 720x576
# 
# # Origin, relative to the upper-left corner, can be overloaded by aligment
# org: 0, 0
# 
# # Image scaling (hor,ver), origin is at the upper-left corner or at the alignment coord (x, y)
# scale: 100%, 100%
# 
# # Alpha blending
# alpha: 100%
# 
# # Smoothing for very blocky images (use OLD for no filtering)
# smooth: OFF
# 
# # In millisecs
# fadein/out: 50, 50
# 
# # Force subtitle placement relative to (org.x, org.y)
# align: OFF at LEFT TOP
# 
# # For correcting non-progressive desync. (in millisecs or hh:mm:ss:ms)
# # Note: Not effective in DirectVobSub, use "delay: ... " instead.
# time offset: 0
# 
# # ON: displays only forced subtitles, OFF: shows everything
# forced subs: OFF
# 
# # The original palette of the DVD
# palette: 0000e1, e83f07, 000000, fdfdfd, 033a03, ea12eb, faff1a, 095d76, 7c7c7c, e0e0e0, 701f03, 077307, 00006c, cc0ae9, d2ab0f, 730972
# 
# # Custom colors (transp idxs and the four colors)
# custom colors: OFF, tridx: 1000, colors: fdfdfd, 000000, e0e0e0, faff1a
# 
# # Language index in use
# langidx: 0
# 
# # Dansk
# id: da, index: 0
# # Decomment next line to activate alternative name in DirectVobSub / Windows Media Player 6.x
# # alt: Dansk
# # Vob/Cell ID: 3, 1 (PTS: 0)
# timestamp: 00:00:44:280, filepos: 000000000
# timestamp: 00:00:50:520, filepos: 000003000

	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	my @header;

	my $read_header = 1;
	my $state = 0;

	my @comments;

	for ( @$content) {
		if ( m/^\s*timestamp\:\s*(\d\d)\:(\d\d)\:(\d\d)\:(\d+).*?filepos\:\s*(.*)$/) {
			push @$from, Subtitles::hms2time( $1, $2, $3, $4);
			push @$text, $5;
		} else {
			push @comments, [ scalar @$from, $_ ];
		}
		$line++;
	}

	for ( $line = 0; $line < @$from - 1; $line++) {
			$$to[$line] = $$from[$line + 1] - 0.002;
	}
	push @$to, $$from[-1] + 2.0 if @$from;

	$sub->{idx}->{comments} = \@comments;

	1;
}

sub write
{
	my ( $self, $sub) = @_;

	die "The idx format subtitles cannot be created from the other formats\n"
			unless $sub->{idx}->{comments};

	my $from = $sub->{from};
	my $to   = $sub->{to};
	my $text = $sub->{text};
	my $c    = $sub->{idx}->{comments};
	my ( $i, $j);
	my $n = @$text;
	my @ret;
	for ( $i = $j = 0; $i < $n; $i++) {
		push @ret, $$c[$j++][1] while $j < @$c and $$c[$j][0] <= $i;
		push @ret, sprintf( "timestamp: %02d:%02d:%02d:%03d, filepos: %s",
			Subtitles::time2hms($from->[$i]), $text->[$i]);
	}
	\@ret;
}

1;

=pod

=head1 NAME

Subtitles - handle video subtitles in various text formats

=head1 DESCRIPTION

Video files (avi mpeg etc) are sometimes accompanied with subtitles, which are
currently very popular as text files. C<Subtitles> provides means for simple
loading, re-timing, and storing these subtitle files.  A command-line tool
F<subs> for the same purpose and using C<Subtitles> interface is included in
the distribution.

The module supports C<srt>, C<sub>, C<smi>, and C<mdvd> subtitle formats.

Time values are floats, in seconds with millisecond precision.

=head1 SYNOPSIS

	use Subtitles;
	
	my $sub = Subtitles->new();

	open F, 'Ichi The Killer.sub' or die "Cannot read:$!";
	die "Cannot load:$@\n" unless $sub-> load(\*F);
	close F;

	# back two minutes
	$sub-> shift( $sub-> parse_time('-02:00')); 

	# re-frame from 25 fps
	$sub-> scale( 23.976 / 25 );

	# or both
	$sub-> transform( -120, 0.96);
	$sub-> transform( -120, 0.96, 0, $sub-> length - 60);

	# split in 2
	my ( $part1, $part2) = $sub-> split( $self-> length / 2);

	# join back with 5-second gap 
	$part1-> join( $part2, 5);

	# save
	open F, "> out.sub" or die "Cannot write:$!\n";
	$part1-> save( \*F);
	close F;

	# report
	print "sub is ", time2str( $sub-> length);

=head1 API

=head2 Package methods

=over

=item codecs

Returns array of installed codecs.

=item hms2time HOURS, MINUTES, SECONDS, MILLISECONDS

Combines four parameters into float time in seconds.

=item time2hms TIME

Splits time into four integers, - hours, minutes, seconds, and milliseconds.
If time is less than zero, zero times are returned.

=item time2shms

Splits time into five integers, - time sign, hours, minutes, seconds, and milliseconds.

=item time2str TIME

Converts time to a human-readable string.

=back

=head2 Object methods

=over

=item clear

Removes all content

=item codec [ STRING ]

If STRING is not defined, returns currently associated codec.
Otherwise, sets the new codec in association. The STRING is
the codec's package name, such as C<Subtitles::Codec::srt>.

=item dup [ CLEAR ]

Duplicates object instance in deep-copy fashion. If CLEAR
flag is set, timeframes are not copied.

=item join GUEST, GAP

Adds content of object GUEST at the end of the list of subtitles with GAP in seconds.

=item length

Returns length of subtitle span.

=item load FH [ CODEC ]

Reads subtitle content into object. If successful, returns 1;
otherwise undef is returned and C<$@> contains the error.

By default, tries to deduce which codec to use; to point the
selection explicitly CODEC string is to be used.

=item lines

Returns number of subtitle cues.

=item new

Creates a new instance. To force a particular
codec, supply C<codec> string here.

=item parse_time STRING

Parses STRING which is either a C<[[HH:]MM:]SS[,MSEC]> string
or string in a format specific to a codec, for example, number
of a frame. 

=item rate FPS

Forces a particluar frame-per-second rate, if a codec
can make use of it.

=item save FH

Writes content of instance into FH file handle,
using the associated codec.

=item scale A

Changes time-scale. If A is 2, the subtitles 
go off 2 times slower, if 0.5 - two times faster, etc.

=item shift B

Shifts timings by B seconds. B can be negative.

=item split TIME

Splits the content of the instance between
two newly created instances of the same class,
by TIME, and returns these. The both resulting 
subtitles begin at time 0.

=item transform A, B [FROM, TO]

Applies linear transformation to the time-scale,
such as C<u = At + B> where C<t> is the original 
time and C<u> is the result. If FROM and TO 
brackets are set, the changes are applied only
to the lines in the timeframe between these.

=back

=head1 BUGS

This is alpha code, more a proof-of-concept rather
that anything else, so most surely bugs are lurking.

Anyway: not all subtitle types are recognized.
The modules doesn't handle multi-language subtitles.

=head1 SEE ALSO

L<subs> - command-line wrapper for this module

=head1 THANKS

L<http://dvd.box.sk/>, L<http://subs.2ya.com>.

=head1 AUTHOR

Dmitry Karasik, E<lt>dmitry@karasik.eu.orgE<gt>.

=cut