#! perl -slw
use strict;
use threads;
use threads::shared;
use Thread::Queue;
use Time::HiRes qw[ sleep time ];
use Text::LevenshteinXS qw(distance);

our $M //= 4;
our $S //= 25;
our $L //= 100;
our $C //= 50;

my $minimum_score = $C;
my $max_childs    = $M;

sub worker {
    my $Q = shift;
    while( $_ = $Q->dequeue ) {
        my( $srcId, $srcSeq, $dstId, $dstSeq ) = map split( ':' ), split;

        my $dist = distance( $srcSeq, $dstSeq );
        my $score = 100 - ( $dist * 100 / length( $dstSeq ) || 1 );

        if( $score >= $minimum_score ) {
            print "$srcId looks like $dstId";
        }
    }
}

srand 1;

my $Q = new Thread::Queue;
my @workers = map async( \&worker, $Q ), 1 .. $max_childs;

my %contacts;
get_db_data( \%contacts );
my @ids = keys %contacts;

print STDERR "Starting with max: $max_childs concurrent threads";
my $start = time;

for my $idn1 ( 0 .. $#ids ) {
    my $srcId = $ids[ $idn1 ];
    for my $idn2 ( $idn1 + 1 .. $#ids ) {
        my $dstid = $ids[ $idn2 ];
        sleep 0.1 while $Q->pending > 10;
        $Q->enqueue( "$srcId:$contacts{ $srcId } $dstId:$contacts{ $dstId }" );
    }
}

$Q->enqueue( (undef) x $max_childs );
$_->join for @workers;

printf STDERR "Took %6f seconds\n", time() - $start;
exit;

sub get_db_data { ## mock_up
    my $ref = shift;
    $ref->{ $_ } = join '', map{ ( 'A', 'C', 'G', 'T' )[ rand 4 ] } 1 .. $L
        for 1 .. $S;
    return;
}

__END__
c:\test>866682-2
Starting with max: 4 concurrent threads
21 looks like 23
17 looks like 12
2 looks like 12
23 looks like 21
23 looks like 14
23 looks like 4
16 looks like 12
16 looks like 15
6 looks like 12
12 looks like 17
12 looks like 2
12 looks like 16
12 looks like 6
14 looks like 23
15 looks like 16
4 looks like 23
4 looks like 5
5 looks like 4
Took 1.167890 seconds