#!perl
# see http://www.perlmonks.org/?node=rsyncsize
use strict; use warnings; use List::Util "min"; 
my $g = $ARGV[0] or die q
(rsyncsize.pl -- utility to find which directories and files use the 
most disk space on a remote file system accessed by rsync.
Usage: perl rsyncsize.pl host:path 
or host::path or any other form of pathname rsync accepts
);
open my $R, "-|", qw"rsync -zr", $g or die "error executing rsync"; 
my %h; my $c = 0; 
while(<$R>){ 
	if (!/^\S{10}/) { warn $_; next; } $c++; 
	chomp; my $n = substr $_,43; 
	substr($_,11) =~ /^\s*(\d+)/ or die; my $s = $1; 
	while ($n =~ m"/|$"g) 
		{ my $p = $`; $h{$p} += $s; }
} 
close $R or die "error from rsync"; 
warn "got $c files listed"; 
my @m = sort { $h{$b} <=> $h{$a} } keys%h; 
for my $n (reverse @m[0 .. min(@m-1,9999)]) { 
	printf "%10.0f %s\n", $h{$n}, $n; 
}
__END__