Ok, I gave up on HTML: HTML is a load of rubbish, and
parsing all the rubbish that's out there would be enough to
drive the Dalai-Lama to eat a raw cow (ok I may be
exaggerating a little).
Using Jordi's help from months ago, I got an Xhtml module
that uses Xml.pm working in less than a couple of hours, and
it seems to be working quite well.
I'll submit another patch soon to update the tests.
Y.