No problem, nor any need for apology. I just wanted to mention that my approach would also work with the expanded test set. I think it's important to stress your point that any approach to this problem requires a testing framework that can easily exercise all the common, degenerate and corner cases.