--- amv.pl	2007/07/21 15:03:30	19
+++ amv.pl	2007/10/01 20:26:13	33
@@ -9,6 +9,8 @@
 # http://en.wikipedia.org/wiki/RIFF_(File_format)
 # http://www.obrador.com/essentialjpeg/HeaderInfo.htm
 # http://lists.helixcommunity.org/pipermail/datatype-dev/2005-January/001886.html
+# http://mpgedit.org/mpgedit/mpeg_format/mpeghdr.htm
+# http://wiki.multimedia.cx/index.php?title=IMA_ADPCM
 
 use strict;
 
@@ -17,25 +19,40 @@
 use File::Path;
 use Getopt::Long;
 
-my $dump = 0;
+my $dump_amv = 0;
+my $dump_video = 0;
+my $dump_jpeg = 0;
+my $dump_audio = 0;
 my $debug = 0;
+my $verbose = 0;
 my $dump_dir = '/tmp/dump/';
+my $dump_avi = "dump.avi";
 my $no_jpeg_header = 0;
 my $jpeg_q = 100;
+my $jpegtran;
 
 GetOptions(
-	"dump!"			=> \$dump,
+	"dump-amv!"		=> \$dump_amv,
+	"dump-video!"	=> \$dump_video,
+	"dump-jpeg!"	=> \$dump_jpeg,
+	"dump-audio!"	=> \$dump_audio,
 	"debug!"		=> \$debug,
 	"dump-dir=s"	=> \$dump_dir,
 	"no-jpeg-headers!" => \$no_jpeg_header,
+	"jpegtran=s"	=> \$jpegtran,
+	"verbose!"		=> \$verbose,
 );
 
 my $path = shift @ARGV || die "usage: $0 movie.amv\n";
 
+# by default, flip frames
+$jpegtran = '-flip vertical' unless defined($jpegtran);
 
 rmtree $dump_dir if -e $dump_dir;
 mkpath $dump_dir || die "can't create $dump_dir: $!";
 
+$| = 1;
+
 open(my $fh, '<', $path) || die "can't open $path: $!";
 
 # offset in file
@@ -45,8 +62,6 @@
 my $d;
 
 sub hex_dump {
-	return unless $dump;
-
 	my ( $bytes, $offset ) = @_;
 	return unless $bytes;
 
@@ -88,10 +103,13 @@
 	my $r_len = length($bytes);
 	confess "read $r_len bytes, expected $len" if $len != $r_len;
 
-	hex_dump( $bytes );
+	if ( $dump_amv ) {
+		print "## raw $len bytes\n";
+		hex_dump( $bytes );
+	}
 
 	if ( $bytes eq 'AMV_END_' ) {
-		print "> end of file marker AMV_END_\n" if $dump;
+		print "> end of file marker AMV_END_\n" if $dump_video;
 		$d->{eof}++;
 		return;
 	}
@@ -105,16 +123,20 @@
 	}
 }
 
+# my $len = next_part( 'boob' );
+# my ( $len, $part ) = next_part();
+
 sub next_part {
 	my ( $expected_part, $expected_len, $skip ) = @_;
 	my ( $part, $len ) = x(8,'A4V');
 	return unless $len;
-	confess "not $expected_part but $part" if $expected_part ne $part;
+	confess "not $expected_part but $part" if $expected_part && $expected_part ne $part;
 	if ( $expected_len ) {
 		confess "expected $expected_len bytes for $part got $len" if $len != $expected_len;
 	}
-	printf "<< %s - %d 0x%x bytes\n", $part, $len, $len;
+	printf "## next_part %s - %d 0x%x bytes\n", $part, $len, $len if $debug;
 	x($len) if $skip;
+	return ( $len, $part )  if wantarray;
 	return $len;
 }
 
@@ -131,7 +153,7 @@
 		$out .= chr($t);
 	}
 
-	if ( $dump ) {
+	if ( $dump_video ) {
 		print "## quantization table original\n";
 		hex_dump( $in );
 		print "## quantization table for $jpeg_q %\n";
@@ -156,14 +178,10 @@
 
 	if ( $#subframes < ( $join_subframes - 1 ) ) {
 		push @subframes, $data;
-		print "## saved $frame_nr/", $#subframes + 1, " subframe of ", length($data), " bytes\n";
+		print "## saved $frame_nr/", $#subframes + 1, " subframe of ", length($data), " bytes\n" if $debug;
 		return;
 	}
 
-	my $path = sprintf("$dump_dir/%04d.jpg", $frame_nr++ );
-
-	open(my $fh, '>', $path) || die "can't create $path: $!";
-
 	my $w = $d->{amvh}->{width} || die "no width?";
 	my $h = $d->{amvh}->{height} || confess "no height?";
 
@@ -286,7 +304,7 @@
 		0,			# Ah, Ai (not used)
 	);
 
-	if ( $dump ) {
+	if ( $dump_jpeg ) {
 		print "## created JPEG header...\n";
 		hex_dump( $header, 0 );
 	}
@@ -294,23 +312,144 @@
 	my $frame = join('', @subframes ) . $data;
 	@subframes = ();
 
+	my $path = sprintf("$dump_dir/%04d.jpg", $frame_nr );
+
+	my $fh;
+	if ( $jpegtran ) {
+		open($fh, '|-', "jpegtran $jpegtran > $path") || die "can't create $path: $!";
+	} else {
+		open($fh, '>', $path) || die "can't create $path: $!";
+	}
+
 	if ( ! $no_jpeg_header ) {
-		print $fh $header, $frame, "\xFF\xD9" || die "can't write jpeg $path: $!";
+		print $fh $header . $frame . "\xFF\xD9" || die "can't write jpeg $path: $!";
 	} else {
 		print $fh $frame || die "can't write raw jpeg $path: $!";
 	}
 	close $fh || die "can't close $path: $!";
-	print ">> created $frame_nr ", $no_jpeg_header ? 'raw' : '', " jpeg $path ", -s $path, " bytes\n";
+	print ">> created $frame_nr ", $no_jpeg_header ? 'raw' : '', " jpeg $path ", -s $path, " bytes\n" if $verbose;
 }
 
+#
+# IMA ADPCM decoder
+#
+
+my @index_adjust = ( -1, -1, -1, -1, 2, 4, 6, 8 );
+
+my @step_size = (
+	7, 8, 9, 10, 11, 12, 13, 14, 16, 17,
+	19, 21, 23, 25, 28, 31, 34, 37, 41, 45,
+	50, 55, 60, 66, 73, 80, 88, 97, 107, 118,
+	130, 143, 157, 173, 190, 209, 230, 253, 279, 307,
+	337, 371, 408, 449, 494, 544, 598, 658, 724, 796,
+	876, 963, 1060, 1166, 1282, 1411, 1552, 1707, 1878, 2066,
+	2272, 2499, 2749, 3024, 3327, 3660, 4026, 4428, 4871, 5358,
+	5894, 6484, 7132, 7845, 8630, 9493, 10442, 11487, 12635, 13899,
+	15289, 16818, 18500, 20350, 22385, 24623, 27086, 29794, 32767
+);
+
+my $pred_val = 0;
+my $step_idx = 0;
+
+# This code is "borrowed" from the ALSA library 
+# http://www.alsa-project.org
+
+sub adpcm_decode_sample {
+	my $code = shift;
+
+	my $pred_diff;	# Predicted difference to next sample
+	my $step;		# holds previous step_size value
+
+	# Separate sign and magnitude
+	my $sign = $code & 0x8;
+	$code &= 0x7;
+
+	# Computes pred_diff = (code + 0.5) * step / 4,
+	# but see comment in adpcm_coder.
+
+	$step = $step_size[$step_idx] || die "no step_size[$step_idx]";
+
+	# Compute difference and new predicted value 
+	$pred_diff = $step >> 3;
+	my $i = 0x4;
+	while( $i ) {
+		if ($code & $i) {
+			$pred_diff += $step;
+		}
+		$i >>= 1;
+		$step >>= 1;
+	}
+	$pred_val += $sign ? -$pred_diff : $pred_diff;
+
+	# Clamp output value 
+	if ($pred_val > 32767) {
+		$pred_val = 32767;
+	} elsif ($pred_val < -32768) {
+		$pred_val = -32768;
+	}
+
+	# Find new step_size index value
+	$step_idx += $index_adjust[$code];
+
+	if ($step_idx < 0) {
+		$step_idx = 0;
+	} elsif ($step_idx > 88) {
+		$step_idx = 88;
+	}
+	return $pred_val;
+}
+
+my $au_path = "$dump_dir/sound.au";
+open(my $au_fh, '>', $au_path) || die "can't open $au_path: $!";
+print $au_fh pack 'a4N5', (
+	# magic
+	'.snd',
+	# data offset
+	24,
+	# data size
+	-1,
+	# encoding - 16-bit linear PCM
+	3,
+	# sample rate
+	22050,
+	#channels
+	1,
+);
+
+sub audio_frame {
+	my $data = shift || die "no data?";
+
+	my ( $origin, $index, $bytes ) = unpack 'ssL', substr($data,0,8);
+
+	$pred_val = $origin;
+	$step_idx = $index;
+
+	my $size = 0;
+
+	foreach my $b ( map { ord($_) } split(//, substr($data,8)) ) {
+		print $au_fh pack 'n', adpcm_decode_sample( $b >> 4 );		
+		print $au_fh pack 'n', adpcm_decode_sample( $b & 15 );		
+		$size += 2;
+	}
+
+	warn "length isn't corrent $bytes != $size" if $bytes != $size;
+}
+
+#
+# read AMV file
+#
+
 my ( $riff, $amv ) = x(12, 'Z4x4Z4');
 die "$path not RIFF but $riff" if $riff ne 'RIFF';
 die "$path not AMV but $amv" if $amv ne 'AMV ';
 
+my $fps = 16;
+my $duration;
+
 while ( ! defined($d->{eof}) ) {
 	my ( $list, $name ) = x(12,'A4x4A4');
 	die "not LIST but $list" if $list ne 'LIST';
-	print "< $list * $name\n";
+	print "< $list * $name\n" if $verbose;
 
 	if ( $name eq 'hdrl' ) {
 
@@ -324,12 +463,15 @@
 			$h->{$n} = $v;
 		} x($len, 'Vx28VVVx8CCv');
 
-		printf "## %s %d*%d %s fps (%d ms/frame) %02d:%02d:%02d\n",
+		$duration = sprintf('%02d:%02d:%02d', $h->{hh}, $h->{mm}, $h->{ss} );
+
+		printf "## %s %d*%d %s fps (%d ms/frame) %s\n",
 			$path,
 			$h->{width}, $h->{height}, $h->{fps}, $h->{ms_per_frame},
-			$h->{hh}, $h->{mm}, $h->{ss};
+			$duration;
 
 		$d->{amvh} = $h;
+		$fps = $h->{fps};
 
 	} elsif ( $name eq 'strl' ) {
 
@@ -338,19 +480,63 @@
 
 	} elsif ( $name eq 'movi' ) {
 
-		while (1) {
+		my $have_parts = 1;
+
+		while ( $have_parts ) {
 			my $frame = $d->{movi}++;
-		
-			my $len = next_part( '00dc' );
-			last unless $len;
-			printf "<< %s 00dc - frame %d jpeg %d 0x%x bytes\n", $name, $frame, $len, $len;
-			mkjpg( x($len) );
 
-			$len = next_part( '01wb', 0, 1 );
-			printf "<< %s 01wb - frame %d audio %d 0x%x bytes\n", $name, $frame, $len, $len;
+			my $parts = 0;
+
+			while ( $parts < 2 ) {
+
+				my ( $len, $part ) = next_part();
+
+				if ( ! $len ) {
+					$have_parts = 0;
+					last;
+				}
+
+				if ( $part eq '00dc' ) {
+
+					printf "<< %s 00dc - part %d jpeg %d 0x%x bytes\n", $name, $frame, $len, $len if $verbose;
+					mkjpg( x($len) );
+					$parts++;
+
+				} elsif ( $part eq '01wb' ) {
+					printf "<< %s 01wb - part %d audio %d 0x%x bytes\n", $name, $frame, $len, $len if $verbose;
+
+					my $audio_frame = x( $len );
+
+					if ( $dump_audio ) {
+						printf "#### dumping audio frame %d 0x%x bytes\n", length($audio_frame), length($audio_frame);
+						hex_dump( $audio_frame );
+					}
+
+		#			print $audio_fh mp3_frame;
+					audio_frame( $audio_frame );
+
+					$parts++;
+				} else {
+					warn "unknown next part $part with $len bytes, skipping!";
+				}
+
+				warn "## #$frame_nr $name $part has $parts parts\n" if $debug;
+			}
+
+			$frame_nr++;
+
+			if ( $frame_nr % $fps == 0 ) {
+				print "\n" if ( ( $frame_nr / $fps ) % 60 == 0 );
+				print ".";
+			}
 		};
 
 	} else {
 		die "unknown $list $name";
 	}
 }
+
+my $cmd = "ffmpeg -r $fps -i $dump_dir/%04d.jpg -i $au_path -y $dump_avi";
+system($cmd) == 0 || die "can't convert frames to avi using $cmd: $!";
+
+print ">>>> created $frame_nr frames $dump_avi ", -s $dump_avi, "\n";