> I agree with the other comments about this, and I was thinking that this > construction is said to be better (wouldn't really affect anything here, but > for future reference): > > if p.localName is not None: > > Thoughts? Well, the test still fails with that change, doesn't it? Regards, Martin