The parsing and namespace cleaning isn't terribly expensive, but it does add up. It was adding an extra 5sec when processing 100,000 basic message stanzas.
Matchers are now PEP8 compliant and have documentation.