less efficient in perl 5.10.0 onwards (attaching of PERL_MAGIC_arylen magic to the AV).

Is that going to revert?

Is that going to revert?

No. It's a trade off that made $#a slightly slower and bigger while making arrays smaller and faster (as long as you don't access them via the $# mechanism).

Dave.

Thank you for your response.

arrays smaller and faster (as long as you don't access them via the $#

That's unfortunate.

Is it really impossible to make read-only references to $#.. equivalent (in performance, cost and side-effects), to ( Scalar @.. - 1 )?

Ie. Could not the application of magic to an array, and all the penalties that go with it, be deferred until it is used in an lvalue context?

---

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

RIP PCW It is as I've been saying!(Audio until 20090817)

@a in scalar context just pushes its size onto the stack. $#a in scalar context could be an lvalue or an rvalue, so a 'magic' var is pushed onto the stack that updates the array's size if assigned to. This could probably be optimised when pp_av2arylen is called in rvalue context.

Dave.

Ah, of course. I suppose there are cases where it's not immediately apparent to the parser if a given occurrence is an lvalue or rvalue?

-QM
--
Quantum Mechanics: The dreams stuff is made of

This could probably be optimised when pp_av2arylen is called in rvalue context.

Done by commits 02d85cc37a4acecafdc2f0b45640b03cd1f4ac71 and 28c5b5bcd7f52e6b2219508a1066cd0ccc8dd19a. You'll see these changes in 5.11.2 (next month's dev release) and 5.12 (early to mid 2010)

Interesting! That's the kind of answer I was hoping someone would post. For anyone who's keeping score, here's a benchmark and my results:

#!/usr/bin/perl

use strict;
use warnings;
use Benchmark qw(cmpthese);

my $i=0;
cmpthese(-10,
    {
        deref => sub {
            my $aryptr = [qw(Once upon a time in a galaxy far far away
+)];
            if ( @{ $aryptr } )
            {
                $i=1;
            }
            $aryptr = [];
            if ( @{ $aryptr } )
            {
                $i=-1;
            }
        },
        arylen => sub {
            my $aryptr = [qw(Once upon a time in a galaxy far far away
+)];
            if ( $#{ $aryptr } >= 0 )
            {
                $i=1;
            }
            $aryptr = [];
            if ( $#{ $aryptr } >= 0 )
            {
                $i=-1;
            }
        },
    }
);

__END__
           Rate arylen  deref
arylen  95903/s     --   -28%
deref  132418/s    38%     --
[download]

The point is that $#a should be faster than @a-1. That's not what you are testing. You have have an extra operation in your $#a test instead of having one less.

#!/usr/bin/perl

use strict;
use warnings;

use Benchmark qw(cmpthese);

our @a = qw( Once upon a time in a galaxy far far away );

cmpthese(-3, {
    cnt_minus_1 => 'my $x = @a-1;',
    last_idx    => 'my $x = $#a;',
});
[download]

                 Rate    last_idx cnt_minus_1
last_idx    3225597/s          --        -25%
cnt_minus_1 4308243/s         34%          --
[download]