FWIW, my reusable code would probably look like:
package Directory::Copy;
use strict;
use File::Copy;
use File::Find;
use File::Path;
sub copy_path {
my($origpath, $destpath) = @_;
find(sub {
my($file) = $File::Find::name;
(my $dest = $file) =~ s/^\Q$origpath/$destpath/;
if (-d $file) {
mkpath $dest;
} elsif (-f $file) {
copy $file, $dest;
}
}, $origpath);
}
__END__
As Bart noted, the /o is not really necessary, nor is a string eval. Even
if it were a performance hit, I'd prefer to ignore it in most code, as the
performance hit is probably less severe than the more work and less
readability of dealing with it. In a module on CPAN, such optimizations
would probably be worth it, esp. for older Perls, though.
--
Chris Nandor [EMAIL PROTECTED] http://pudge.net/
Open Source Development Network [EMAIL PROTECTED] http://osdn.com/
