Python notebook scratchpad towards mapping a schema.org dataset description into DCAT

Following today's discussion, and guided by Simon's term level mappings,
I've set up a quick Python notebook on Google Colab using rdflib and
sparql, which explores use of SPARQL CONSTRUCT to map the scheme flavour
into something like DCAT.

It works with a very basic core, but I haven't figured out what can be done
around options (e.g. mapping schema:license to dct:license and what happens
if that term isn't used...). Should it be one complex query or several of
them, ... or look to other mechanisms like shex mappings instead?

https://colab.research.google.com/drive/1FKoN3HR69vqVN3fVH36E3MRkB7wSXGEt
has all I've managed so far, feedback/edits welcomed...

Dan