use Benchmark;

my @array = qw(a b c d e f g h i j k l m n o p q r s t u v w z y z);
my %hash;

timethese(500000, {
               'btrott' => sub {
    @hash{@array} = (1) x @array;
},
               'httptech' => sub {
       for (@array) { $hash{$_} = 1 }
},
           });


Benchmark: timing 500000 iterations of btrott, httptech...
    btrott: 14 wallclock secs (14.58 usr +  0.00 sys = 14.58 CPU)
  httptech: 20 wallclock secs (20.68 usr +  0.01 sys = 20.69 CPU)
[download]

While we're on the topic of benchmarking, I'd just like to, for the hell of it, add the worst possible answer to this question:

map { $code .= "\$hash{$_} = 1;\n" } @array; eval "$code";
[download]

And now for a little program, borrowed from httptech:

use Benchmark;

my @array = qw(a b c d e f g h i j k l m n o p q r s t u v w z y z);
my %hash;

timethese(10000, {
        'btrott' => sub {
                @hash{@array} = (1) x @array;
        },
        'plaid' => sub {
                my $code
        map { $code .= "\$hash{$_} = 1;\n" } @array;
        eval "$code";
        }
});
[download]

Outputs

Benchmark: timing 10000 iterations of btrott, plaid...
    btrott:  1 wallclock secs ( 0.56 usr +  0.00 sys =  0.56 CPU)
     plaid: 23 wallclock secs (22.07 usr +  0.06 sys = 22.13 CPU)
[download]

Don't know why i threw this in. Guess I just didn't want to waste the effort:)

...and my old friend 'grep' Not quite as fast as btrott but still a decent contender. use Benchmark; my @array = qw(a b c d e f g h i j k l m n o p q r s t u v w z y z); my %hash; timethese( 500000, { 'btrott' => sub { @hash{@array} = (1) x @array; }, 'httptech' => sub { for (@array) { $hash{$_} = 1 } }, 'danimal' => sub { grep { $hash{$_}++ } @array; } }, ); Benchmark: timing 500000 iterations of btrott, danimal, httptech... btrott: 21 wallclock secs (21.38 usr + 0.00 sys = 21.38 CPU) danimal: 22 wallclock secs (22.57 usr + 0.00 sys = 22.57 CPU) httptech: 34 wallclock secs (34.10 usr + 0.00 sys = 34.10 CPU)

...and my old friend 'grep' Not quite as fast as btrott but still a decent contender.

use Benchmark;

my @array = qw(a b c d e f g h i j k l m n o p q r s t u v w z y z);
my %hash;

timethese(
           500000,
 {
           'btrott' => sub {
                                 @hash{@array} = (1) x @array;
},

           'httptech' => sub {
                                 for (@array) { $hash{$_} = 1 }
           },

           'danimal' => sub {
                             grep { $hash{$_}++ } @array;
           }
},
);

Benchmark: timing 500000 iterations of btrott, danimal, httptech...
    btrott: 21 wallclock secs (21.38 usr +  0.00 sys = 21.38 CPU)
   danimal: 22 wallclock secs (22.57 usr +  0.00 sys = 22.57 CPU)
  httptech: 34 wallclock secs (34.10 usr +  0.00 sys = 34.10 CPU)
[download]