bitfield.php

<?php

class Bitfield
{
	const FIELD_SIZE = 32;

	public static function getDistance(array $a, array $b)
	{
		$dist = 0;

		for ($i=0; $i<static::FIELD_SIZE; $i++) {
			if (!isset($a[$i]) && !isset($b[$i]))
				continue;

			if (!isset($a[$i]))
				$dist += static::countNonZero($b[$i]);
			elseif (!isset($b[$i]))
				$dist += static::countNonZero($a[$i]);
			else
				$dist += static::countNonZero($a[$i] ^ $b[$i]);
		}

		return $dist;
	}
	
	public static function getSimilarity(array $a, array $b)
	{
		$iden = 0;
		
		foreach($a as $i => $v)
			if (isset($b[$i]))
				$iden += static::countNonZero($a[$i] & $b[$i]);
		
		return $iden;
	}
	
	public static function countNonZero($num)
	{
		if ($num == 0)
			return 0;

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

		return $res;
	}

	public static function bitfieldToHash(array $bitfield)
	{
		for ($i=0; $i<static::FIELD_SIZE; $i++)
			empty($bitfield[$i]) && $bitfield[$i] = 0;
		
		ksort($bitfield); // thanks Obama!
		
		array_unshift($bitfield, 'N*');
		return base64_encode(call_user_func_array('pack', $bitfield));
	}
	
	public static function hashToBitfield($hash)
	{
		return @unpack('N32', base64_decode($hash)) ?: [];
	}
}

$h = Bitfield::bitfieldToHash([1 => 2, 10 => 33, 31 => 123]);
$b = Bitfield::hashToBitfield($h);

var_dump( $h, $b );