Hi borodache,

as a starting point, you might want to take a look at the following very basic subroutine which recursively traverses the subdirectory tree:

sub traverse_dir {
    my $dir = shift;
    my @entries = glob "$dir/*";
    for my $entry (@entries) {
        if (-f $entry) {
            print "$entry:", -s $entry, "\n";
        } elsif (-d $entry) {
            traverse_dir($entry);
        } else
            # perhaps do something for entries which are neither files
+ nor subdirectories, if that sort of thing exists
        }
    }
}
traverse_dir ".";
[download]

Many things would need to be added: management of symbolic links in accordance with your needs, adding accumulators for cumulative size of directories, etc.

Also note that I have used the -d, -f and -s file test operators for simplicity, but in a real program, I would probably use the stat (and possibly lstat) built-in function instead. Also, you might want to replace the glob function call by a opendir and readdir combination, together with appropriate calls to grep to get rid of entries such as . and .. (which are skipped by glob).

Update: see To glob or not to glob for a discussion of glob and some of its advantages and disadvantages.