A (WIP) "correct" implementation of regular expressions in Python using FSAs

This is a cool project I've been working on recently. Essentially, this is an implementation of regular expressions in Python using finite state automata instead of backtracking. This yields a runtime complexity of O(mn) (or O(n) if you compile the regex) compared to the O(2^n) which backtracking has for a worst case.

Another "benefit" (for Python programmers and PyPi) is that the implementation is not* in C, it's in Python.