#! /usr/bin/perl
use strict;

my ($thestring, $popsize, $str_len, @chr, $i, $str, @init_pop, @init_p
+op_fitness, $max_fit);
# GA To match the string:
$thestring = "the quick brown dog jumped over the lazy fox";

# step one, create an inital population of strings the same size
$popsize = 10;
$str_len = length($thestring);

# put list of valid characters into @chr
@chr = qw(a b c d e f g h i j k l m n o p q r s t u v w x y z);
@chr = (@chr, " ");

my $best_match = 0;

# fill each string with random character data
for ($i = 0; $i < $popsize; $i++) {
    $str = join '', map $chr[rand @chr], 1..44;
    
    # store the string and it's fitness
    $init_pop[$i][0] = $str;                        
    $init_pop[$i][1] = fitness($str);

    # keep track of the best match for later    
    if ($init_pop[$i][1] > $max_fit) {
        $best_match = $str;
        $max_fit = $init_pop[$i][1];
    }
}

# now, sort the population by fitness (desc) 
# THIS DOESN'T WORK
sort { $b->[1] <=> $a->[1] } @init_pop;

for ($i = 0; $i < $popsize; $i++) {
    print ($init_pop[$i][1], "\n");
}

# function to calculate the fitness of a given string, 
# that is how many characters match $thestring
sub fitness {
    my $test_str = $_[0];
    my ($j, $fit);
    
    $fit = 0;
        
    for ($j = 0; $j < $str_len; $j++) {
        if (substr($test_str, $j, 1) eq substr($thestring, $j, 1)) {  
+  
            $fit++;
        }
    
    }        
    return $fit;
}
[download]