Maybe this modification? I've assumed, that you'll need full path/file-names (see read_dir() in File::Slurp):

  ... 
  push @files, grep { -f  and  -s $_ >= $min_size  and  -s $_ <= $max_
+size }
               read_dir( $d, prefix => 1 );
  ...
[download]

Update: BTW: Instead of incrementing $c within the loop, you can have the same result by simply using:

...} # loop

my $c = @directories;
...
[download]