refactor to plug the back end v2 curation API

[marfox]

the gadget (AKA tool version 1) now supports the version 2 back-end services /suggest and /curate, see https://www.wikidata.org/wiki/Wikidata:Primary_sources_tool#Architecture

[Tpt]

Great \o/! We should wait for the deployment on a VPS in order to don't break Wikimedia terms of uses.

Do you plan to implement JSON output in the backend? It would allow to clean up a lot the frontend by avoiding to to there the TSV -> JSON conversion.

You should also replace all the usages of the old backend to the new one in order to don't have to maintain 2 backends at the same time.

[marfox]

@Tpt , I replied yesterday via e-mail, but it hasn't shown up here, so pasting below.

Great \o/! We should wait for the deployment on a VPS in order to don't break Wikimedia terms of uses.

Sounds good.

Do you plan to implement JSON output in the backend?

Sure thing! The Wikidata JSON output is already implemented.
However, it is not trivial to refactor the current front end, as it heavily relies on the TSV format, with a considerable number of functions using it.
We would love to get your help here, ideally refactoring functions from high-level ones like matchClaims, down to lower-level ones like parseFreebaseClaims.