in reply to any-all

This is interesting, but beware... you are abusing sort's magic $a and $b variables. Even strict wouldn't complain, but you could have problems if the $code includes a sort.

Also don't use local there, use my. Its faster, and there is no reason to use local in this scenario.

Replies are listed 'Best First'.
RE: RE: any-all
by fundflow (Chaplain) on Sep 02, 2000 at 02:13 UTC
    Hmm.

    I'll add 'local' declaration to it.

    Thers little chance that someone will use sort there but who knows?
[reply]
      Its worse then that... your tests do the same amount of abuse. Have you considered doing this with a closure? 
      sub getAnyAll
{
    my $code = shift;
    my $Any = sub {
        my @A = @{$_[0]}; # I like to dereference early, it makes
        my @B = @{$_[1]}; # the rest of the code more readable.
        while( @A and @B )
        {
            return 1 if &$code( pop(@A), pop(@B) )
        }
        return 0 
    };

    my $All = sub {
        my @A = @{$_[0]};
        my @B = @{$_[1]};
        while( @A and @B )
        {
            return 0 unless &$code( pop(@A), pop(@B) )
        }
        return 1 
    };

    return ( $Any, $All );
}

      [download]
      that's untested, but gives you the idea. By the way... poping arrays passed by reference might not be such a good idea either...
[reply]
[d/l]
        1. What is a closure?

        2. Writing $a is nicer (=shorter and more readable) than $_[0]

        3. I'd like to iterate through both arrays, something like
        "for $a=(@A), $b=(@B)" and thus the pop. Hopefuly perl just pops a pointer array to the original.

        The goal is a short and compact sub that will do the above.
[reply]
[d/l]
[select]

In Section Cool Uses for Perl