TNRS should organize matches by taxa?

[jar398]

TNRS results should be organized by taxon, not name match. Currently you can get several results for a single OTT id, which is odd. See OpenTreeOfLife/opentree#860

I don't think we want to lose the match information, though. Maybe we're only talking about a factoring of the TNRS result information. All the matches for one taxon would be gathered together, so we get a list where each component is a taxon and a list of matches. Each match has fields is_synonym, score, nomenclature_code, and is_approximate_match.

TNRS documentation here: https://github.com/OpenTreeOfLife/germinator/wiki/TNRS-API-v3#match_names

Hacking the TNRS to do this is a pain, and this is a transformation that clients can do, so not obvious that the effort is best spent hacking taxomachine, or the client(s).