Hello Sosi,

In addition to Text::Fuzzy, you should look at Algorithm::Diff. Here’s some proof-of-concept code:

#! perl
use strict;
use warnings;
use Algorithm::Diff qw( LCS );
use Data::Dump;
use constant MIN_MATCH => 0.5;

my @source =
(
    'John Ronald Reuel Tolkien',
    'John Ronald S Tolkien',
    'Trent Reznor',
    'Barack Hussein Obama II',
    'Barack Hussein II',
);

my @search =
(
    'John Ronald Reuel T',
    'Trent Reznor',
    'Barack Hussein II',
    'Barack Hussein Obama II',
    'No match here',
);

my %searches = map { $_ => [] } @search;

for my $s (@search)
{
    my @search_chars = split //, $s;
    my @matches;

    for my $source (@source)
    {
        my @source_chars = split //, $source;
        my @diff_chars   = LCS(\@search_chars, \@source_chars);
        my $diff         = join '', @diff_chars;

        if (!@{ $searches{$s} } ||
            length $diff > length $searches{$s}->[0])
        {
            $searches{$s}->[0] = $diff;
            $searches{$s}->[1] = $source;
        }
    }
}

for my $key (keys %searches)
{
    my $len_key   = length $key;
    my $len_match = length $searches{$key}->[0];

    delete $searches{$key} if ($len_match / $len_key) < MIN_MATCH;
}

dd \%searches;
[download]

Output:

23:19 >perl 1044_SoPW.pl
{
  "Barack Hussein II"       => ["Barack Hussein II", "Barack Hussein O
+bama II"],
  "Barack Hussein Obama II" => ["Barack Hussein Obama II", "Barack Hus
+sein Obama II"],
  "John Ronald Reuel T"     => ["John Ronald Reuel T", "John Ronald Re
+uel Tolkien"],
  "Trent Reznor"            => ["Trent Reznor", "Trent Reznor"],
}

23:19 >
[download]

Hope that helps,