word or character based, via longest common subsequences - View it on GitHub