in reply to (jeffa) Re: References and Arrays
in thread References and Arrays

Thanks, that second way you suggested works great! Not really sure if a simple array would work, I modified the code from the LWP::Parallel::UserAgent examples from cpan so I don't totally know what it is doing.

Perhaps you could take a look at my code and see what you think?!

http://www.perlmonks.org/index.pl?node_id=241752

I'd be happy to hear any comments you might have!

Cheers
Tom

Replies are listed 'Best First'.
Re: Re: (jeffa) Re: References and Arrays
by tachyon (Chancellor) on Mar 11, 2003 at 01:19 UTC

    Your code looks pretty good. use strit warnings and diagnostics or die is worth a read RE benefits of strict. foreach and for are synonyms. You use aliasing to $_ in one loop and a my $var in another. This is inconsistent but it makes for easier to read code to write for my $whatever (@widgets) {} as your code becomes more self documenting, especially if the contents of @widgets is not immediately apparent. This loop structure:

    $num_domains = scalar @domains; while ($num_domains > '0') { # Start Loopy Code [blah] my $new_domain = shift @domains; [blah] $num_domains = scalar @domains; # get the size of array }

    is pretty obtuse to me. while loops are fine but they will also go infinite if you don't satisfy the condition (easy to do). All you really need is a simple

    for ( 0 .. $#@domains ) { # do stuff } # or for my $root_domain ( @domains ) { # blah push @domains, @new_domains; }

    The second code snippet is nifty because if you say extract the links from a page and then add the domains to @domains the @domains list gets longer and....you have an instant spider. Without using recursion per se. You can use this iterating over a list/adding to the list in the loop technique to 'recurse' virtually any structure you like. There is an example at Link Checker FWIW that will spider a website. You can also walk a directory tree in a few lines with:

    my $root = 'd:/perl/'; my ( @dirs, @fails, @files); @dirs = ( $root ); for my $dir ( @dirs ) { opendir DIR, $dir or die $!;#do { push @fails, $dir; next }; while ( $_ = readdir DIR ) { next if -s "$dir/$_"; do{ push @dirs, "$dir/$_"; next } if -d "$dir/$_" and not m/^\.\.?$/; do{ push @files, "$dir/$_"; next } if -f "$dir/$_"; } } print "$_\n" for @dirs;

    Sure you can use File::Find but this is short, sweet and quite a bit faster.

    cheers

    tachyon

    s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print

      $ ln -s . "except_it_is_broken"

      I don't understand.. why does speed matter that much? Have you benchmarked?

      Makeshifts last the longest.