Re: Fastest data structure compare?

[eserte]

Using Storable is problematic, because it's not guaranteed that semantically equivalent data is serialized in the same way. Here's an example playing with the internal utf-8 flag --- while Test::More's is_deeply and Data::Compare get it right, checking the serialized Storable data fails:

#!/usr/bin/perl

use Data::Compare qw();
use Storable qw(nfreeze);
use Test::More qw(no_plan);

my $data1 = ["f\xfcbar"];
my $data2 = [substr("f\xfcbar\x{0100}", 0, -1)];

is_deeply($data1, $data2, "is_deeply test");
ok(Data::Compare::Compare($data1, $data2), "Data::Compare");
ok(nfreeze($data1) eq nfreeze($data2), "storable serialized");

I just benchmarked Test::More::is_deeply vs. Data::Compare and found that the latter is 3x faster for a data set which size is ~6MB as a storable-serialized file. This probably depends on the structure of the data set.

[Anonymous Monk]

Storable warns against this

if you happen to use your numbers as strings between two freezing operations on the same data structures, you will get different results.

There is no facility either to return all strings as utf8 sequences, or to attempt to convert utf8 data back to 8 bit and croak() if the conversion fails.

It is easily avoided if you use encoding/Encode

[eserte]

Encode may help in the utf8 case, but not in the case of different integer representation (as mentioned in the Storable manpage):

#!/usr/bin/perl

use Data::Compare qw();
use Storable qw(nfreeze);
use Test::More qw(no_plan);

my $data1 = { foobar => 1 };
my $data2 = { foobar => "1" };

$Storable::canonical = 1;

is_deeply($data1, $data2, "is_deeply test");
ok(Data::Compare::Compare($data1, $data2), "Data::Compare");
ok(nfreeze($data1) eq nfreeze($data2), "storable serialized");

[Anonymous Monk]

Great, you've read the bug section :)