use strict;
use warnings;

use String::LCSS_XS qw( lcss );

my @array1 = ('This sentence is stored in array1') x 1_000;
my @array2 = (
    'Thank you for looking on this.',
    'This sentence is a test sentence.',
    'This sentence is stored in array2',
    'This is an other sentence.',
    'All variables are stored in array2',
) x 20_000;

preprocess( \@array1 );
preprocess( \@array2 );

process( \@array1, \@array2 );

exit 0;

my %words;
my %codes;
my $word_num;

sub process {
    my ( $array1, $array2 ) = @_;

    for my $elem1 (@$array1) {
        my $best_subseq       = "";
        my $best_subseq_words = 0;
        my $best_subseq_chars = 0;
        my $found             = 0;

        for my $elem2 (@$array2) {

            my $subseq = lcss( $elem1->[1], $elem2->[1] );

            if ( $subseq && length($subseq) > $best_subseq_words ) {
                my $real_subseq =
                  join ' ', map $codes{$_}, split //, $subseq;

                if ( length($real_subseq) > $best_subseq_chars ) {
                    $best_subseq       = $real_subseq;
                    $best_subseq_words = length($subseq);
                    $best_subseq_chars = length($real_subseq);
                    $found             = 1;
                }
            }
        }

        if ( $found == 1 ) {
            printf "<<%s>> is the lcss of\n\t<<%s>>\n",
              $best_subseq,
              $elem1->[0];
        }
    }
}

sub preprocess {
    my ($array) = @_;

    for my $i ( 0 .. $#$array ) {
        my $packed =
          join '', map {
            my $code = $words{$_};
            if ( not defined $code ) {
                $code = chr( $word_num += 1 );
                $words{$_}    = $code;
                $codes{$code} = $_;
            }
            $code;
          } $array->[$i] =~ /\w+/g;

        $array->[$i] = [ $array->[$i], $packed ];
    }
}