wwdc-download

#!/usr/bin/perl -s
#
# right-click on streaming video, select "Copy video address", and feed it to this script
#
# Example URL:
#	2016:	http://devstreaming.apple.com/videos/wwdc/2015/106z3yjwpfymnauri96m/106/hls_vod_mvp.m3u8
#	2017:	http://devstreaming.apple.com/videos/wwdc/2016/102w0bsn0ge83qfv7za/102/hls_vod_mvp.m3u8
#
# Not sure if this works outside WWDC.
#
# You need to be logged in to http://developer.apple.com
#
# hacked together by Reid / @clith - rae@tnir.org
#
my $base;
my $m3u;
my $ts;
my $ffmpegCmd = "ffmpeg -i 'concat:%s' -c copy -bsf:a aac_adtstoasc '%s'";
my $limit="400k";
my $limitOpt = "--limit-rate ${limit}";
my $baseCurlCmd = "curl -s";
# $baseCurlCmd .= " $limitOpt";
my $maxDownloadRetryCount = 4;
my $dateString;

local $SIG{HUP} = sub {
	chomp($dateString = `date`);
	print "\n\n[$dateString] Caught user ^C, deleting ${ts} and exiting..\n";
	unlink($ts);
	exit(0);
};

sub findBaseAudioVideo($) {
	my $url = shift;
	my $base = $url;
	$base =~ s#/[^/]*?$##;
	print "url = $url\n";
	print "base = $base\n";
	open CURL, "curl -s \"$url\"|" || die "$!: Can't open URL $url";
	my $audio = "";
	my $highestBandwidth=0;
	my $video = "";
	while(<CURL>) {
		chomp;
		if(m#GROUP-ID="program_audio".*URI="(.+)"#) {
			$audio = $1;
			next;
		}
		if(m#EXT-X-STREAM-INF.*\bBANDWIDTH=(\d+)#) {
			if($1 > $highestBandwidth) {
				$highestBandwidth = $1;
				$video = <CURL>;
				next;
			}
		}
	}
	close(CURL);
	print("base = $base\n");
	print("audio = $audio\n");
	print("video = $video\n");
	return ($base, $audio, $video)
} 

#print "## Limiting downloads to ${limit}B/s\n";
foreach (@ARGV) {
	unless(m#^(.*)/(\d+)/([\w.]+)$#) {
		unless(m#^(.*)/(m3u8)/([\w.]+)$#) {
			unless(m#^(.*)/([\w.]+)$#) {
				warn "Can't read url $_\n";
				next;
			}
			else {
				print "$1 -- $2\n";
				# $base = "$1/$2";
				$session = "m3u8";
				# $m3u = $3;
			}
		} else {
			print "$1 -- $2 -- $3\n";
			# $base = "$1/$2";
			$session = "$2";
			# $m3u = $3;
		}
	} else {
		# $base = "$1/$2";
		$session = "$2";
		# $m3u = $3;
	}
	# $curlCmd = "${baseCurlCmd} --referer $_";
	print("1 session=${session}\n");
	$session="101" if $session eq "m3u8";
	# print("2 session=${session} [$base / $m3u]\n");
	print("2 session=${session}\n");
	
	# my @videoList = split ' ', `$curlCmd "$base/$m3u"`;
	# my @videoList = split ' ', `$curlCmd "$_"`;
	# my $vidbase = "${base}/${vidSize}";
	# my $vidm3u = "${vidbase}/${vidSize}.m3u8";
	# my $stream_m3u = `$curlCmd "$_"`;
	# my $vidm3u = `$curlCmd "$_"|grep -v '^#'| sort -r -s | head -1`;

	mkdir $session unless -d $session;
	chdir $session || warn "was not able to chdir: $!\n";
	chomp(my $date = `date`);
	print "\n# Starting $session [$tsfiles[0] - ${lastTs}] on ${date}\n";
	print "$_\n";

	my ($base, $audio, $video) = findBaseAudioVideo($_);
	foreach my $vidm3u ("$video", "$audio") {

		chomp($vidm3u);
		print("vidm3u = $vidm3u\n");
		$vidbase = ${vidm3u};
		$vidbase =~ s#/[^/]*$##;
		print("tslist-cmd: $curlCmd \"$base/$vidm3u\" | grep -v '^#'");
		my $tslist = `$curlCmd "$base/$vidm3u"|grep -v '^#'`;
		my @tsfiles = split ' ', $tslist;
		my $lastTs = $tsfiles[$#tsfiles -1];

		# skip already-compiled session videos
		my @allFiles=<*>;
		print("3 session=${session}\n");
		my @sessionVideoList = grep(/${session}./, @allFiles);
		if(scalar @sessionVideoList > 0) {
			print "Skipping \"@sessionVideoList\"\n---\n";
			my $cmd = "echo grep \"${session}\"";
			system($cmd);
			next;
		} else {
			print "Downloading \"$session\"\n";
		}

		my $stopped = 0;
		my $complete = 0;
		my $column = 1;
		my $lastTs = "";
		foreach $ts (@tsfiles) {
			my $tsurl = "$base/$ts";
			print("tsurl = $tsurl\n");
			my $shouldSkip=0;
			if(-r $ts) {
				print " [$ts] ";
				$shouldSkip=1;
				$complete = 1;
			} else {
				print "  $ts  ";
				$complete = 0;
			}
			$column++;
			if($column > 10) {
				$column=1;
				print "\n";
			}
			next if $shouldSkip;
			for (my $downloadAttemptCount = 1; $downloadAttemptCount < $maxDownloadRetryCount; $downloadAttemptCount++) {
				my $systemValue = system("$curlCmd -O \"$tsurl\"");
				# print "$curlCmd -O \"$tsurl\"\n";
				if($systemValue) {
					my $sig = $? & 127;
					$lastTs = $tsurl;
					chomp($dateString = `date`);
					printf("\n\n[$dateString] Error: failed (signal %d) to download $ts (try ${downloadAttemptCount} of $maxDownloadRetryCount) [$systemValue]\n", $sig);
					print "  Exited with value ", $? >> 8, "\n";
					print "  $ts  \n";
					$columns=2;
					if($sig == 2) {
						print "^C\n";
						print "\n[$dateString] Cleaning up and exiting; deleting $ts..\n";
						unlink $ts;
						exit(0);
					}
				} else {
					$complete = 1;
					last;
				}
			}
		}
		if($complete) {
			my $episode=$session;
			$episode =~ s/^(\d)/${1}x/;
			my $cmd = sprintf($ffmpegCmd, join("|", @tsfiles), "../${episode}..${hdExt}.mp4");
			my $printCmd = sprintf($ffmpegCmd, "<$#tsfiles files>", "../${episode}..${hdExt}.mp4");
			chomp($dateString = `date`);
			print "\n\n[$dateString] Creating mp4\n[$dateString] $printCmd\n\n";
			if(system($cmd)) {
				warn "Creation of ${episode} mp4 failed: $!\n";
			} else {
				chomp($dateString = `date`);
				print "\n[$dateString] Creation of ${episode} mp4 complete\n\n";
			}
		} else {
			chomp($dateString = `date`);
			print "\n\n[$dateString] Error: failed to download $ts (try #${downloadAttemptCount} of $maxDownloadRetryCount)\n";
			print "$lastTs\n";
			print "[$dateString] Halting all downloads.\n";
			unlink $ts if $ts ne "";
			last;
		}
		chdir "..";
	}
}