http://www.perlmonks.org?node_id=1114466


in reply to Re: The Boy Scout Rule
in thread The Boy Scout Rule

my $value = [ $x => $y ] -> [ $y <= $x ]; would not pass my code review.

I agree itís confusing the first time and the => should be a skinny comma :P instead. That said, the Schwartzian transform is even more confusing the first time you see it. No one in a post 5.6 Perl world would suggest rewriting it with a bunch of temp arrays and for blocks. So, I advocate simple little idioms like the above when they offer something more than clever/pretty code.

So, thinking to possibly defend the clever/pretty one, I tried a Benchmark, which I'm not necessarily doing right so someone please jump in if itís badly formed, and the one that might be the most semantically clear and I assumed would be the slowest is the fastest by a good measure. I forgot that List::Util is XS.

use strictures; use Benchmark "cmpthese"; use List::Util "min"; # This is XS. my @xy = ( [ 1, 0 ], [ 0, 1 ], [ 0, 0 ], [ 1, 1 ], [ 1_000_000, 999_999 ], [ 999_999, 1_000_000 ] ); my $m; # Avoid void in comparisons. cmpthese(10_000_000, { list_util => sub { $m = min(@$_) for @xy }, ternary => sub { $m = $_->[0] < $_->[1] ? $_->[0] : $_->[1] for +@xy }, clever => sub { $m = [ $_->[0], $_->[1] ]->[ $_->[0] <= $_->[1] + ] for @xy }, });
Rate clever ternary list_util clever 347584/s -- -56% -68% ternary 792393/s 128% -- -28% list_util 1096491/s 215% 38% --