So no big news, but here’s what the citeproc-py default "html+rdfa"
output (which is just the internal model serialized) looks like ATM:
Beyond the fact that we now have round-tripping of data
(notwithstanding ordering of the contributors), I’m thinking this
should help processing too. For example, I haven’t gotten to it, but
note that the name nodes carry the structured data (the content
attribute) with the initialized display form.
Bruce
Bruce