Also you can't accidentally offset the index within the iterative code e.g

for(my $i = 0; $i < @ARGVl $i++) {
    if($ARGV[$i] =~ /meets some condition/) {
        # ack! the index has been offset
        $i += 2;
    }
}
[download]

for my $i (0 .. $#ARGV) {
    if($ARGV[$i] =~ /meets some condition/) {
        # doesn't matter as $i will be set again in the next iteration
        $i += 2;
    }
}
[download]

Here's a quick benchmark for those in need of speed

use Benchmark qw(cmpthese); 
my @list = 1..10000;
my $r;

cmpthese(-10, {
    cstyle => sub {
        for(my $i = 0; $i < @list; $i++) {
            $r = $list[$i];
        }
    },
    perlstyle => sub {
        for my $i (@list) {
            $r = $i;
        }
    }
});

__output__

Benchmark: running cstyle, perlstyle, each for at least 10 CPU seconds
+...
    cstyle: 15 wallclock secs (10.57 usr +  0.05 sys = 10.62 CPU) @ 36
+.06/s (n=383)
 perlstyle: 15 wallclock secs (10.10 usr +  0.03 sys = 10.13 CPU) @ 86
+.77/s (n=879)
            Rate    cstyle perlstyle
cstyle    36.1/s        --      -58%
perlstyle 86.8/s      141%        --
[download]

_________
broquaint

update: added benchmark because it's just so darned easy

use strict;
use Benchmark;

timethese(10, {
   'c-style'    => sub {
      for (my $i = 0; $i < 1_000_000; $i++) { }
   },
   'list style' => sub {
      for my $i (0..1_000_000) { }
   },
});
[download]

jeffa

L-LL-L--L-LL-L--L-LL-L--
-R--R-RR-R--R-RR-R--R-RR
B--B--B--B--B--B--B--B--
H---H---H---H---H---H---
(the triplet paradiddle with high-hat)

Am I correct in understanding that the three argument for is actually converted into a while continue? With the added overhead of two scoped blocks?

Yves / DeMerphq
---
Writing a good benchmark isnt as easy as it might look.

B::Deparse is your friend here.

$ perl -MO=Deparse -e 'for ($x = 0; $x < 10; $x++) {}'
$x = 0;
while ($x < 10) {
    ();
}
continue {
    ++$x
}
[download]

So the answer would be yes. As to whether there's extra scope overhead, well, we'll leave that as an exercise for the reader.

for (@ARGV) { .. }
[download]

-- 
Joost       downtime n. The period during which a system
            is error-free and immune from user input.
[download]

Perhaps because its just not the same thing? Your's sets $_ to the elements of the array, while Popcorn Dave's sets $_ to one more than the indices of the array (which probably isn't the desired result). As broquaint suggests, I tend to use the for(0..$#ARGV) version if I happen to need the indices.

--
I'd like to be able to assign to an luser

In the original code, the first argument was an option, the rest were files, so that wouldn't have worked in this particular instance.

Some people fall from grace. I prefer a running start...