Nice. Making the sort the last step like this is more efficient than my solution in the case of duplicates.

Elegant. I've learned so many simple but non-obvious design patterns just in a few weeks of coding exercises on here. Thanks for this one!

benchmarking is the only way to find that out.

I think a solution for maximum performance would look quite different and would be significantlyfaster (10x to 20x faster i would totally guess). Usually regular expressions are only faster if your programming language is not so performance (like early perl/php implementations I guess).

This was a million years ago, but I'm wondering about the performance of this solution. It's certainnly very readable and elegant once you know a little about regex. Are there multiple iterations involved? How does that work under the dashboard?