Spooky, that sample works for me too it is just a redacted/cut down representation of the larger original. Maybe there were some newlines or tabs that were purged via cut/paste… Cant see anything invalid in the original, that still shows the “mixed content is not supported” error…
Had a closer look, in the original machine-generated xml the elements were indented with some kind of space characters and the parser failed. Changing these to tab characters and the parser worked fine… Just a bit confusing as the failing XML appears to be valid UTF-8 XML.
Thanks anyway,
Mike