#!/usr/bin/perl

=head1 NAME

dh_compress - compress files and fix symlinks in package build directories

=cut

use strict;
use warnings;
use Cwd qw(getcwd abs_path);
use File::Spec::Functions qw(abs2rel);
use Debian::Debhelper::Dh_Lib;

=head1 SYNOPSIS

B<dh_compress> [S<I<debhelper options>>] [B<-X>I<item>] [B<-A>] [S<I<file> ...>]

=head1 DESCRIPTION

B<dh_compress> is a debhelper program that is responsible for compressing
the files in package build directories, and makes sure that any symlinks
that pointed to the files before they were compressed are updated to point
to the new files.

By default, B<dh_compress> compresses files that Debian policy mandates should
be compressed, namely all files in F<usr/share/info>, F<usr/share/man>,
files in F<usr/share/doc> that are larger than 4k in size,
(except the F<copyright> file, F<.html> and other web files, image files, and files
that appear to be already compressed based on their extensions), and all
F<changelog> files. Plus PCF fonts underneath F<usr/share/fonts/X11/>

=head1 FILES

=over 4

=item debian/I<package>.compress

These files are deprecated.

If this file exists, the default files are not compressed. Instead, the
file is ran as a shell script, and all filenames that the shell script
outputs will be compressed. The shell script will be run from inside the
package build directory. Note though that using B<-X> is a much better idea in
general; you should only use a F<debian/package.compress> file if you really
need to.

=back

=head1 OPTIONS

=over 4

=item B<-X>I<item>, B<--exclude=>I<item>

Exclude files that contain F<item> anywhere in their filename from being
compressed. For example, B<-X.tiff> will exclude TIFF files from compression.
You may use this option multiple times to build up a list of things to
exclude.

=item B<-A>, B<--all>

Compress all files specified by command line parameters in ALL packages
acted on.

=item I<file> ...

Add these files to the list of files to compress.

=back

=head1 CONFORMS TO

Debian policy, version 3.0

=cut

init();

my $olddir;

foreach my $package (@{$dh{DOPACKAGES}}) {
	my $tmp=tmpdir($package);

	my $compress=pkgfile($package,"compress");

	# Run the file name gathering commands from within the directory
	# structure that will be effected.
	next unless -d $tmp;
	$olddir = getcwd() if not defined $olddir;
	verbose_print("cd $tmp");
	chdir($tmp) || error("Can't cd to $tmp: $!");

	# Figure out what files to compress.
	my @files;
	# First of all, deal with any files specified right on the command line.
	if (($package eq $dh{FIRSTPACKAGE} || $dh{PARAMS_ALL}) && @ARGV) {
		push @files, map { s{^/+}{}; $_ } @ARGV;
	}
	if ($compress) {
		# The compress file is a sh script that outputs the files to be compressed
		# (typically using find).
		warning("$compress is deprecated; use -X or avoid calling dh_compress instead");
		push @files, split(/\n/,`sh $olddir/$compress 2>/dev/null`);
	}
	else {
		# Note that all the excludes of odd things like _z 
		# are because gzip refuses to compress such files, assuming
		# they are zip files. I looked at the gzip source to get the
		# complete list of such extensions: ".gz", ".z", ".taz", 
		# ".tgz", "-gz", "-z", "_z"
		push @files, split(/\n/,`
			find usr/info usr/share/info usr/man usr/share/man usr/X11*/man -type f ! -iname "*.gz" \\
				! -iname "*.gif" ! -iname "*.png" ! -iname "*.jpg" \\
				! -iname "*.jpeg" \\
				2>/dev/null || true;
			find usr/share/doc \\
				\\( -type d -name _sources -prune -false \\) -o \\
				-type f \\( -size +4k -o -name "changelog*" -o -name "NEWS*" \\) \\
				\\( -name changelog.html -o ! -iname "*.htm*" \\) \\
				! -iname "*.xhtml" \\
				! -iname "*.gif" ! -iname "*.png" ! -iname "*.jpg" \\
				! -iname "*.jpeg" ! -iname "*.gz" ! -iname "*.taz" \\
				! -iname "*.tgz" ! -iname "*.z" ! -iname "*.bz2" \\
				! -iname "*-gz"  ! -iname "*-z" ! -iname "*_z" \\
				! -iname "*.epub" ! -iname "*.jar" ! -iname "*.zip" \\
				! -iname "*.odg" ! -iname "*.odp" ! -iname "*.odt" \\
				! -iname ".htaccess" ! -iname "*.css" \\
				! -iname "*.xz" ! -iname "*.lz" ! -iname "*.lzma" \\
				! -iname "*.svg" ! -iname "*.svgz" ! -iname "*.js" \\
				! -name "index.sgml" ! -name "objects.inv" ! -name "*.map" \\
				! -name "*.devhelp2" \\
				! -name "copyright" 2>/dev/null || true;
			find usr/share/fonts/X11 -type f -name "*.pcf" 2>/dev/null || true;
		`);
	}

	# Exclude files from compression.
	if (@files && defined($dh{EXCLUDE}) && $dh{EXCLUDE}) {
		my @new=();
		foreach (@files) {
			my $ok=1;
			foreach my $x (@{$dh{EXCLUDE}}) {
				if (/\Q$x\E/) {
					$ok='';
					last;
				}
			}
			push @new,$_ if $ok;
		}
		@files=@new;
	}
	
	# Look for files with hard links. If we are going to compress both,
	# we can preserve the hard link across the compression and save
	# space in the end.
	my @f=();
	my %hardlinks;
	my %seen;
	foreach (@files) {
		my ($dev, $inode, undef, $nlink)=stat($_);
		if (defined $nlink && $nlink > 1) {
			if (! $seen{"$inode.$dev"}) {
				$seen{"$inode.$dev"}=$_;
				push @f, $_;
			}
			else {
				# This is a hardlink.
				$hardlinks{$_}=$seen{"$inode.$dev"};
			}
		}
		else {
			push @f, $_;
		}
	}

	# normalize file names and remove duplicates
	my $norm_from_dir = $tmp;
	if ($norm_from_dir !~ m{^/}) {
		$norm_from_dir = "${olddir}/${tmp}";
	}
	my @normalized = normalize_paths($norm_from_dir, $tmp, @f);
	my %uniq_f; @uniq_f{@normalized} = ();
	@f = sort keys %uniq_f;

	# do it
	if (@f) {
		# Make executables not be anymore.
		xargs(\@f,"chmod","a-x");
		
		xargs(\@f,"gzip","-9nf");
	}
	
	# Now change over any files we can that used to be hard links so
	# they are again.
	foreach (keys %hardlinks) {
		# Remove old file.
		doit("rm","-f","$_");
		# Make new hardlink.
		doit("ln","$hardlinks{$_}.gz","$_.gz");
	}

	verbose_print("cd '$olddir'");
	chdir($olddir);

	# Fix up symlinks that were pointing to the uncompressed files.
	my %links = map { chomp; $_ => 1 } `find $tmp -type l`;
	my $changed;
	# Keep looping through looking for broken links until no more
	# changes are made. This is done in case there are links pointing
	# to links, pointing to compressed files.
	do {
		$changed = 0;
		foreach my $link (keys %links) {
			my ($directory) = $link =~ m:(.*)/:;
			my $linkval = readlink($link);
			if (! -e "$directory/$linkval" && -e "$directory/$linkval.gz") {
				doit("rm","-f",$link);
				doit("ln","-sf","$linkval.gz","$link.gz");
				delete $links{$link};
				$changed++;
			}
		}
	} while $changed;
}

sub normalize_paths {
	my ($cwd, $tmp, @paths) = @_;
	my @normalized;
	my $prefix = qr{\Q${tmp}/};

	for my $path (@paths) {
		my $abs = abs_path($path);
		if (not defined($abs)) {
			my $err = $!;
			my $alt = $path;
			if ($alt =~ s/^$prefix//) {
				$abs = abs_path($alt);
			}
			error(qq{Cannot resolve "$path": $err (relative to "${cwd}")})
				if (not defined($abs));
			warning(qq{Interpreted "$path" as "$alt"});
		}
		error("${abs} does not exist") if not -e $abs;
		push(@normalized, abs2rel($abs, $cwd));
	}
	return @normalized;
}

=head1 SEE ALSO

L<debhelper(7)>

This program is a part of debhelper.

=head1 AUTHOR

Joey Hess <joeyh@debian.org>

=cut

# Local Variables:
# indent-tabs-mode: t
# tab-width: 4
# cperl-indent-level: 4
# End: