*laugh* - Thanks for the pointers and consider myself better learned. :)
Hmm ... now I better go & hunt down all those greps in my scripts...
Jason
If memory serves me right, on Tuesday 29 January 2002 13:26, Jeff 'japhy'
Pinyan wrote:
> On Jan 29, Jason Purdy said:
> >my ( @filea, @fileb, @inAButNotInB, @inBButNotInA );
>
> I don't see any hashes being used. This feels like it's going to be very
> inefficient.
>
> >open ( FILEA, shift ) || die "Cannot open 1st file: $!\n";
> >@filea = <FILEA>;
> >close ( FILEA );
> >
> >open ( FILEB, shift ) || die "Cannot open 2nd file: $!\n";
> >@fileb = <FILEB>;
> >close ( FILEB );
> >
> >foreach my $object ( @filea ) {
> > chomp $object;
> > push @inAButNotInB, $object if ( !grep(/$object/, @fileb) );
>
> And indeed it is. Do not use grep(/$foo/, LIST) to determine if $foo is
> in LIST, for a myriad of reasons. First, you're using a regex, so if $foo
> has any regex metacharacters in it, you'll get false negatives and false
> positives. Second, you're not anchoring the regex at all, so if LIST
> contains "rabbit" and $foo is "bit", you'll get a false positive. Third,
> grep() goes through the ENTIRE list. Imagine the consequences of
>
> if (grep $_ == 5, 1 .. 10_000) { ... }
>
> You've just scanned 99,995 more elements than necessary!
>
> The Perl FAQ suggests use of a hash. See my response to this same
> question.
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
