This is an implementation of NFA-based regular-expression matching in pure python. At the moment, it implements a small subset of the normal regular expression syntax that only a mathematician would be satisfied with (although it should sufficient to recognize all regular languages, doing so wouldn't be any fun.)
It is not a finished product. Everything about it is subject to change. I would recommend not using it for anything.
Any names, especially, are subject to change.
It is substantially inspired by Russ Cox's articles on Implementing Regular Expressions, which you really should read if you find this code at all interesting.