#!/usr/bin/perl

#- Simple cat archive with bzip2 for perl.
#- read file list and produce a perl script in stdout with hash data and $ARGV[0].cz2 archive file.
#- hash data is :
#-   { 'filename' => [ $off1, $siz1, $off2, $siz2 ], ... }
#- bzip2 -d between ..1 and ..2
use strict qw(subs vars refs);

sub main {
    my ($name, $maxsiz) = @_;
    my ($off1, $siz1, $off2, $siz2) = (0, 0, 0, 0);
    my @filelist = ();
    my %data = ();

    die "usage: $0 <name> <maxsiz>" unless $maxsiz > 100000;

    unlink "$name.cz2";

    foreach (<STDIN>) {
	chomp;

	my $file = $_; die "unable to find file $file" unless -e $file;
	$siz2 = -s $file;

	push @filelist, $file;
	$data{$file} = [ -1, -1, $off2, $siz2 ];

	if ($off2 + $siz2 > $maxsiz) { #- need compression.
	    system "cat @filelist | bzip2 -3 >tmp.bz2";
	    $siz1 = -s "tmp.bz2";

	    $data{$_} = [ $off1, $siz1, $data{$_}[2], $data{$_}[3] ] foreach @filelist;

	    system "cat tmp.bz2 >>$name.cz2";
	    $off1 += $siz1;
	    $off2 = 0; $siz2 = 0;
	    @filelist = ();
	}
	$off2 += $siz2;
    }
    if (scalar @filelist) {
	system "cat @filelist | bzip2 -3 >tmp.bz2";
	$siz1 = -s "tmp.bz2";

	$data{$_} = [ $off1, $siz1, $data{$_}[2], $data{$_}[3] ] foreach @filelist;

	system "cat tmp.bz2 >>$name.cz2";
	$off1 += $siz1;
	print STDERR "archive size of $name.cz2 is $off1\n";
    }

    local *OUTPUT;
    open OUTPUT, ">$name.cz2.pl";
    print OUTPUT "\%data = (\n";
    foreach (keys %data) {
	print OUTPUT "\'$_\' => [ ". join(', ', @{$data{$_}}) ." ],\n";
    }
    print OUTPUT ");\n";
}

main(@ARGV);
unlink "tmp.bz2";