WP_Text_Diff_Renderer_Table::compute_string_distance( string $string1, string $string2 ): int

Source

public function compute_string_distance( $string1, $string2 ) {
	// Use an md5 hash of the strings for a count cache, as it's fast to generate, and collisions aren't a concern.
	$count_key1 = md5( $string1 );
	$count_key2 = md5( $string2 );

	// Cache vectors containing character frequency for all chars in each string.
	if ( ! isset( $this->count_cache[ $count_key1 ] ) ) {
		$this->count_cache[ $count_key1 ] = count_chars( $string1 );
	}
	if ( ! isset( $this->count_cache[ $count_key2 ] ) ) {
		$this->count_cache[ $count_key2 ] = count_chars( $string2 );
	}

	$chars1 = $this->count_cache[ $count_key1 ];
	$chars2 = $this->count_cache[ $count_key2 ];

	$difference_key = md5( implode( ',', $chars1 ) . ':' . implode( ',', $chars2 ) );
	if ( ! isset( $this->difference_cache[ $difference_key ] ) ) {
		// L1-norm of difference vector.
		$this->difference_cache[ $difference_key ] = array_sum( array_map( array( $this, 'difference' ), $chars1, $chars2 ) );
	}

	$difference = $this->difference_cache[ $difference_key ];

	// $string1 has zero length? Odd. Give huge penalty by not dividing.
	if ( ! $string1 ) {
		return $difference;
	}

	// Return distance per character (of string1).
	return $difference / strlen( $string1 );
}

View all references View on Trac View on GitHub