File::Find uses global variables (like $File::Find::name), so with these, you might encounter problems when starting two recursive instances of File::Find. It also changes the current directory of your process, which is another global state, but it restores the directory upon completion of the recursive descent. Other than that, I don't see any caveats. What happened when you tried it in your code (that you keep jealously hidden from our prying eyes)?

You likely won't find much of a speedup over your opendir, readdir and closedir solution, as File::Find uses the same functions. IO always is slow.

If you just want exactly all directories two levels down, have a look at glob:

my @directories_two_levels_down = grep { -d } glob './*/*';
[download]

Also, Yes, even you can use CPAN

Apologies for not posting the code - this has been constructed of code from different forums and a little tweaking. No comments as yet.

#!/usr/local/bin/perl -w

($#ARGV != 2) or die "Usage: $0 [directory] [rentention period in week
+s]\n";

use strict;
use warnings;
use File::Find;
use File::stat;
use Time::localtime;
use Time::Local;

my ($dir_path, $ret) = @ARGV;
my $rdate = time - ($ret * 24 * 60 * 60);
my $dir_depth = 1;
my %size;
my %date;
    
find( { preprocess => \&preprocess,
    wanted => \&wanted,}
    , $dir_path);

my @sorted = sort {$size{$b} <=> $size{$a}} keys %size;

print "\nList of directories older than: ".ctime($rdate)."\n\n";
printf "%-10s %-24s %s", "Size", "Modified Date", "Directory Path\n\n"
+;
    
foreach (@sorted)
{
    printf "%10d %s %s\n", $size{$_}, ctime($date{$_}), $_;
}

sub preprocess {
    my $depth = $File::Find::dir =~ tr[/][];
    return @_ if $depth < $dir_depth;
    return grep { not -d } @_ if $depth == $dir_depth;
    return;
}

sub wanted {
    my $depth = $File::Find::dir =~ tr[/][];
    my $mod_date =  (stat($File::Find::name)->mtime);
    if ( ($mod_date < $rdate) && ($depth == $dir_depth)) {
        $size{$File::Find::name} = ProcessDir($File::Find::name) if -d
+;
        $date{$File::Find::name} = $mod_date if -d;
    }
    return if $depth == $dir_depth;
}

sub ProcessDir
{
  my( $Path ) = @_;
  my @DirList;
  my $DirSize = 0;
  if( opendir( DIR, $Path ) )
  {
    while( my $Object = readdir( DIR ) )
    {
      next if( "." eq $Object || ".." eq $Object );
      my $ObjectPath = "$Path\\$Object";
      if( -d $ObjectPath )
      {
        push( @DirList, $ObjectPath )
      }
      else
      {
        $DirSize += -s $ObjectPath;
      }
    }
    closedir( DIR );
  }
  foreach my $ObjectPath ( @DirList )
  {
    $DirSize += ProcessDir( $ObjectPath );
  }
  return( $DirSize );
}
[download]

It is not complete but I was wondering if the sub ProcessDir could be more efficient.

I will investigate glob.

Cheers
P

As you seem to be on Windows, have you looked at Win32::DirSize?