Thank Fredrik for a brilliant job of re-implementing Perl's regex syntax into something that I assume is maintainable (haven't looked at the code myself) *and* Unicode compliant.
Yes, this seems to be a very good thing for Python.
Speedy regexp engines are notoriously hard to maintain cleanly, at least, so told me a few successive maintainers of GNU regexp. Difficult points are deterministic matching (avoiding backtracking) and POSIX compliance, and the longest match criterion in particular.
For one, I'm pretty happy with Python regexp implementation, even if it avoids the above points. It has other virtues that are well worth the trade, at least from the experience I have of it so far!
So, in a word, thanks too! :-)