Acme-CPANModules-OrderedHash
view release on metacpan or search on metacpan
lib/Acme/CPANModules/OrderedHash.pm view on Meta::CPAN
},
},
{
module => 'Tree::RB::XS',
description => <<'MARKDOWN',
Multi-purpose tree data structure which can record insertion order and act as an
ordered hash. Use `track_recent => 1, keys_in_recent_order => 1` options. Can
be used as a tied hash, or as an object (faster).
MARKDOWN
bench_code => sub {
my ($op, $numkeys, $numrep) = @_;
my $tree= Tree::RB::XS->new(compare_fn => 'str', track_recent => 1, keys_in_recent_order => 1);
for (1..$numkeys) { $tree->insert("key$_") }
if ($op eq 'delete') {
for (1..$numkeys) { $tree->delete("key$_") }
} elsif ($op eq 'keys') {
for (1..$numrep) { my @keys= $tree->keys }
} elsif ($op eq 'iterate') {
for (1..$numrep) { my $iter = $tree->iter; while (my $v = $iter->next) {} }
}
},
},
],
bench_datasets => [
{name=>'insert 1000 pairs', argv => ['insert', 1000]},
{name=>'insert 1000 pairs + delete', argv => ['delete', 1000]},
{name=>'insert 1000 pairs + return keys 100 times', argv => ['keys', 1000, 100]},
{name=>'insert 1000 pairs + iterate 10 times', argv => ['iterate', 1000, 10], exclude_participant_tags => ['no_iterate']},
],
};
1;
# ABSTRACT: List of modules that provide ordered hash data type
__END__
=pod
=encoding UTF-8
=head1 NAME
Acme::CPANModules::OrderedHash - List of modules that provide ordered hash data type
=head1 VERSION
This document describes version 0.004 of Acme::CPANModules::OrderedHash (from Perl distribution Acme-CPANModules-OrderedHash), released on 2025-04-15.
=head1 SYNOPSIS
To run benchmark with default option:
% bencher --cpanmodules-module OrderedHash
To run module startup overhead benchmark:
% bencher --module-startup --cpanmodules-module OrderedHash
For more options (dump scenario, list/include/exclude/add participants, list/include/exclude/add datasets, etc), see L<bencher> or run C<bencher --help>.
=head1 DESCRIPTION
When you ask a Perl's hash for the list of keys, the answer comes back
unordered. In fact, Perl explicitly randomizes the order of keys it returns
everytime. The random ordering is a (security) feature, not a bug. However,
sometimes you want to know the order of insertion. These modules provide you
with an ordered hash; most of them implement it by recording the order of
insertion of keys in an additional array.
Other related modules:
L<Tie::SortHash> - will automatically sort keys when you call C<keys()>,
C<values()>, C<each()>. But this module does not maintain insertion order.
=head1 ACME::CPANMODULES ENTRIES
=over
=item L<Tie::IxHash>
=item L<Hash::Ordered>
=item L<Tie::Hash::Indexed>
Provides two interfaces: tied hash and OO.
=item L<Tie::LLHash>
=item L<Tie::StoredOrderHash>
=item L<Array::OrdHash>
Provide something closest to PHP's associative array, where you can refer
elements by key or by numeric index, and insertion order is remembered.
=item L<List::Unique::DeterministicOrder>
Provide a list, not hash.
=item L<Tree::RB::XS>
Multi-purpose tree data structure which can record insertion order and act as an
ordered hash. Use C<< track_recent =E<gt> 1, keys_in_recent_order =E<gt> 1 >> options. Can
be used as a tied hash, or as an object (faster).
=back
=head1 BENCHMARKED MODULES
Version numbers shown below are the versions used when running the sample benchmark.
L<Tie::IxHash> 1.23
lib/Acme/CPANModules/OrderedHash.pm view on Meta::CPAN
T:S: participant=Tie::StoredOrderHash
TH:I: participant=Tie::Hash::Indexed
TR:X: participant=Tree::RB::XS
The above result presented as chart:
=begin html
<img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAtAAAAH4CAMAAABUnipoAAAAIGNIUk0AAHomAACAhAAA+gAAAIDoAAB1MAAA6mAAADqYAAAXcJy6UTwAAADJUExURf///wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA...
=end html
Result formatted as table (split, part 4 of 4):
#table4#
{dataset=>"insert 1000 pairs + return keys 100 times"}
+----------------------+-----------+-----------+-----------------------+-----------------------+-----------+---------+
| participant | rate (/s) | time (ms) | pct_faster_vs_slowest | pct_slower_vs_fastest | errors | samples |
+----------------------+-----------+-----------+-----------------------+-----------------------+-----------+---------+
| Tie::StoredOrderHash | 17 | 58 | 0.00% | 1439.14% | 6.1e-05 | 20 |
| Tie::LLHash | 20 | 50 | 16.39% | 1222.37% | 7.3e-05 | 20 |
| Array::OrdHash | 25 | 40 | 44.54% | 964.81% | 0.00011 | 21 |
| Tie::IxHash | 26.8 | 37.3 | 54.99% | 893.08% | 3.3e-05 | 20 |
| Tie::Hash::Indexed | 44 | 23 | 154.54% | 504.67% | 2.7e-05 | 20 |
| Hash::Ordered | 135 | 7.43 | 678.48% | 97.71% | 7.1e-06 | 20 |
| Tree::RB::XS | 270 | 3.8 | 1439.14% | 0.00% | 4.3e-06 | 20 |
+----------------------+-----------+-----------+-----------------------+-----------------------+-----------+---------+
The above result formatted in L<Benchmark.pm|Benchmark> style:
Rate T:S T:L A:O T:I TH:I H:O TR:X
T:S 17/s -- -13% -31% -35% -60% -87% -93%
T:L 20/s 15% -- -19% -25% -54% -85% -92%
A:O 25/s 44% 25% -- -6% -42% -81% -90%
T:I 26.8/s 55% 34% 7% -- -38% -80% -89%
TH:I 44/s 152% 117% 73% 62% -- -67% -83%
H:O 135/s 680% 572% 438% 402% 209% -- -48%
TR:X 270/s 1426% 1215% 952% 881% 505% 95% --
Legends:
A:O: participant=Array::OrdHash
H:O: participant=Hash::Ordered
T:I: participant=Tie::IxHash
T:L: participant=Tie::LLHash
T:S: participant=Tie::StoredOrderHash
TH:I: participant=Tie::Hash::Indexed
TR:X: participant=Tree::RB::XS
The above result presented as chart:
=begin html
<img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAtAAAAH4CAMAAABUnipoAAAAIGNIUk0AAHomAACAhAAA+gAAAIDoAAB1MAAA6mAAADqYAAAXcJy6UTwAAADMUExURf///wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA...
=end html
=head2 Sample benchmark #2
Benchmark command (benchmarking module startup overhead):
% bencher --cpanmodules-module OrderedHash --module-startup
Result formatted as table:
#table5#
+----------------------+-----------+-------------------+-----------------------+-----------------------+-----------+---------+
| participant | time (ms) | mod_overhead_time | pct_faster_vs_slowest | pct_slower_vs_fastest | errors | samples |
+----------------------+-----------+-------------------+-----------------------+-----------------------+-----------+---------+
| Hash::Ordered | 14 | 6 | 0.00% | 80.85% | 0.00011 | 20 |
| Tie::Hash::Indexed | 13 | 5 | 3.99% | 73.91% | 9.5e-05 | 21 |
| Array::OrdHash | 13 | 5 | 9.26% | 65.51% | 9.4e-05 | 20 |
| Tree::RB::XS | 12 | 4 | 9.34% | 65.39% | 9.6e-05 | 20 |
| Tie::LLHash | 12 | 4 | 12.48% | 60.77% | 8.7e-05 | 20 |
| Tie::IxHash | 12 | 4 | 13.93% | 58.73% | 3.8e-05 | 20 |
| Tie::StoredOrderHash | 10 | 2 | 39.06% | 30.05% | 0.0001 | 20 |
| perl -e1 (baseline) | 8 | 0 | 80.85% | 0.00% | 7.8e-05 | 21 |
+----------------------+-----------+-------------------+-----------------------+-----------------------+-----------+---------+
The above result formatted in L<Benchmark.pm|Benchmark> style:
Rate H:O TH:I A:O TR:X T:L T:I T:S perl -e1 (baseline)
H:O 71.4/s -- -7% -7% -14% -14% -14% -28% -42%
TH:I 76.9/s 7% -- 0% -7% -7% -7% -23% -38%
A:O 76.9/s 7% 0% -- -7% -7% -7% -23% -38%
TR:X 83.3/s 16% 8% 8% -- 0% 0% -16% -33%
T:L 83.3/s 16% 8% 8% 0% -- 0% -16% -33%
T:I 83.3/s 16% 8% 8% 0% 0% -- -16% -33%
T:S 100.0/s 39% 30% 30% 19% 19% 19% -- -19%
perl -e1 (baseline) 125.0/s 75% 62% 62% 50% 50% 50% 25% --
Legends:
A:O: mod_overhead_time=5 participant=Array::OrdHash
H:O: mod_overhead_time=6 participant=Hash::Ordered
T:I: mod_overhead_time=4 participant=Tie::IxHash
T:L: mod_overhead_time=4 participant=Tie::LLHash
T:S: mod_overhead_time=2 participant=Tie::StoredOrderHash
TH:I: mod_overhead_time=5 participant=Tie::Hash::Indexed
TR:X: mod_overhead_time=4 participant=Tree::RB::XS
perl -e1 (baseline): mod_overhead_time=0 participant=perl -e1 (baseline)
The above result presented as chart:
=begin html
<img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAtAAAAH4CAMAAABUnipoAAAAIGNIUk0AAHomAACAhAAA+gAAAIDoAAB1MAAA6mAAADqYAAAXcJy6UTwAAACKUExURf///wAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAABYAHwYACAAAAAAAAAAAAAAAACYANx...
=end html
To display as an interactive HTML table on a browser, you can add option C<--format html+datatables>.
=head1 FAQ
=head2 What is an Acme::CPANModules::* module?
An Acme::CPANModules::* module, like this module, contains just a list of module
names that share a common characteristics. It is a way to categorize modules and
document CPAN. See L<Acme::CPANModules> for more details.
=head2 What are ways to use this Acme::CPANModules module?
( run in 0.806 second using v1.01-cache-2.11-cpan-cdf2f3d4e48 )