Keywords hamming distance

hamming_dist_keywords

<?php

$campaigns = [
	'foo' => ['keywords' => ['a','b','c']],
	'bar' => ['keywords' => ['b','c','e']],
	'baz' => ['keywords' => ['d','f','b']],
];

$allKeywords = array_unique(call_user_func_array('array_merge', array_map(function($c) { return $c['keywords']; }, $campaigns)));

asort($allKeywords);

$map = array_flip($allKeywords);

// keywords bitfield calculation
foreach ($campaigns as & $campaign) {
	$bits = 0;
	
	foreach($campaign['keywords'] as $k)
		$bits |= 1 << $map[$k];
	
	$campaign['bits'] = $bits;
}

$user = [
	'keywords' => ['c','d','e'],
	'bits' => 0,
];

foreach($user['keywords'] as $k) {
	if (!isset($map[$k]))
		continue;
	
	$user['bits'] |= 1 << $map[$k];
}

// calculate distance for each campaign
foreach($campaigns as $name => $campaign) {
	$distance = number_of_non_zero($campaign['bits'] ^ $user['bits']); // '^' is not 'xor' !!
	echo "Name: $name, distance: $distance\n"; 
}

function number_of_non_zero($num)
{
	if ($num == 0)
		return 0;
	
	$res = 0;
	do {
		$num &= $num -1;
		$res += 1;
	} while($num);
	
	return $res;
}