view release on metacpan or  search on metacpan

bin/bencher-tiny  view on Meta::CPAN

sub bench_items {
    require Benchmark;

    my $items = shift;

    # load modules first
    for my $item (@$items) {
        my $mod = $item->{module} or next;
        (my $mod_pm = "$mod.pm") =~ s!::!/!g;
        print "DEBUG: Loading module '$mod' ...\n" if $DEBUG;
        require $mod_pm;
    }

    my $func = $Opts{mode} eq 'timethese' ? 'timethese' : 'cmpthese';

    &{"Benchmark::$func"}(
        $Opts{count}, {
            map { ($_->{name} => $_->{code}) } @$items,
        },
    );
}

sub run {
    my $scenario = load_scenario();
    my $items = gen_items($scenario);
    bench_items($items);
}

parse_options();
run();

# ABSTRACT: Simple script to run benchmark scenario with Benchmark.pm
# PODNAME: bencher-tiny

__END__

=pod

=encoding UTF-8

=head1 NAME

bencher-tiny - Simple script to run benchmark scenario with Benchmark.pm

=head1 VERSION

This document describes version 0.003 of bencher-tiny (from Perl distribution Bencher-Tiny), released on 2016-10-20.

=head1 SYNOPSIS

After you install L<Bencher::Scenario::RefUtil>:

 % bencher-tiny -c 10000 RefUtil
                      Rate reftype(ARRAY) is_plain_arrayref is_arrayref ref(ARRAY)
 reftype(ARRAY)    11905/s             --              -36%        -42%       -43%
 is_plain_arrayref 18519/s            56%                --         -9%       -11%
 is_arrayref       20408/s            71%               10%          --        -2%
 ref(ARRAY)        20833/s            75%               12%          2%         --

 % bencher-tiny -c 10000 RefUtil --timethese
 Benchmark: timing 10000 iterations of is_arrayref, is_plain_arrayref, ref(ARRAY), reftype(ARRAY)...
 is_arrayref:  1 wallclock secs ( 0.48 usr +  0.00 sys =  0.48 CPU) @ 20833.33/s (n=10000)
 is_plain_arrayref:  0 wallclock secs ( 0.51 usr +  0.00 sys =  0.51 CPU) @ 19607.84/s (n=10000)
 ref(ARRAY):  1 wallclock secs ( 0.49 usr +  0.00 sys =  0.49 CPU) @ 20408.16/s (n=10000)
 reftype(ARRAY):  1 wallclock secs ( 0.84 usr +  0.00 sys =  0.84 CPU) @ 11904.76/s (n=10000)

=head1 DESCRIPTION

L<bencher> is a simple script to run benchmark scenario with L<Benchmark>
(Benchmark.pm). It lacks L<bencher> features like filtering
participants/datasets/modules, listing participants/datasets or other actions.
But it is small and does not have any non-core dependency.

=head1 EXIT CODES

0 on success.

99 on command-line options error.

=head1 OPTIONS

=head2 --count=i, -c

Specify count to pass to C<cmpthese()>.

=head2 --timethese

Use C<timethese()> instead of C<cmpthese()>.

=head2 --cmpthese

Use C<cmpthese()> (the default, so normally it is not necessary to use this
option).

=head1 ENVIRONMENT

=head2 DEBUG => bool

Will print debug messages when set to true.

=head1 HOMEPAGE

Please visit the project's homepage at L<https://metacpan.org/release/Bencher-Tiny>.

=head1 SOURCE

Source repository is at L<https://github.com/perlancar/perl-Bencher-Tiny>.

=head1 BUGS

Please report any bugs or feature requests on the bugtracker website L<https://rt.cpan.org/Public/Dist/Display.html?Name=Bencher-Tiny>

When submitting a bug or request, please include a test-file or a
patch to an existing test-file that illustrates the bug or desired
feature.

=head1 SEE ALSO

L<bencher>

=head1 AUTHOR