Hmm, maybe I like this shift business and I'll use it. It makes sense.
but I never thought about it because I don't like manipulating arrays
without a good reason. It's micro optimisation and readability is better
(sometimes) although that added line does disturb me a bit.
So I benchmarked it. of course it's blazing fast in any case. I did a minimal
work
in the subs. I'm not benchmarking king but the results surprised me. I
ran with a count of 10^6, 10^7 and 5 . 10^7
The surprise was that there was an up to 6% speed difference and the
biggest surprise is the inversion of which method is fastest.
I'm curious, someone care to comment?
Nadim.
Rate shift copy
shift 735294/s -- -1%
copy 740741/s 1% --
Rate copy shift
copy 720461/s -- -3%
shift 739645/s 3% --
Rate shift copy
shift 696379/s -- -6%
copy 742611/s 7% --
with this code:
cmpthese
(
50_000_000,
{
'shift' => sub {_shift(@args)},
'copy' => sub {_copy(@args)},
}
);
sub _shift {my $self = shift; my ($x, $y, $z)=...@_ ; my $r = $x + $y + $z ;
return $r ;}
sub _copy {my ($self, $x, $y,$z) =...@_; my $r = $x + $y + $z ;return $r ;}
+++ Johan Vromans [Fri, Sep 10, 2010 at 01:06:23PM +0200]:
> Eric Wilhelm <[email protected]> writes:
>
> > So, say the first half of the innards of foo() are being refactored to
> > bar(), which will also take the original parameters in the same form:
> >
> > sub foo {
> > my $self = shift;
> > my %bar = $self->bar(@_);
> >
>
> Don't forget SUPER:
>
> sub foo {
> my $self = shift;
> my %bar = $self->SUPER::foo(@_);
>
> -- Johan
