Ah, yes, thanks for the report. The MultiTagMatcher needs to unpack QNames here:
https://github.com/lxml/lxml/blob/41cc5f378e2454ff1cabe5d227242cce211a3a2b/src/lxml/etree.pyx#L2720
I would just let it create the string and recurse. This is not performance critical at all. PR welcome, including tests, which should go near the".iter()" tests in "test_etree.py".
Ah, yes, thanks for the report. The MultiTagMatcher needs to unpack QNames here:
https:/ /github. com/lxml/ lxml/blob/ 41cc5f378e2454f f1cabe5d227242c ce211a3a2b/ src/lxml/ etree.pyx# L2720
I would just let it create the string and recurse. This is not performance critical at all.
PR welcome, including tests, which should go near the".iter()" tests in "test_etree.py".