Add support for testing phyloreferences with different levels of strictness based on their status

[gaurav]

The latest update of the Curation Tool (phyloref/curation-tool#66) adds a system for tracking the state of phyloreferences (phyloref/curation-tool#25). JPhyloref should use this information to determine which phyloreferences should be tested. Based on the list in phyloref/curation-tool#25:

• pso:draft: incomplete drafts -- should not be tested
• pso:final-draft: final drafts -- try to resolve, but mark as TODO if they fail
• pso:under-review: attempts to test this phyloreference have failed -- try to resolve, but mark as TODO if they fail
• pso:submitted: attempts to test this phyloreference has succeeded -- should be tested
• pso:published: have been/are to be published to the Clade Ontology -- should be tested
• pso:retracted-from-publication: were previously published but are now deprecated -- try to resolve, but mark as TODO if they fail

[hlapp]

pso:retracted-from-publication: were previously published but are now deprecated -- try to resolve, but mark as TODO if they fail

Isn't this more or less the same as marking a term obsolete? Or are we not necessarily talking about classes in the Clade Ontology here, but PHYX files?

[gaurav]

Isn't this more or less the same as marking a term obsolete? Or are we not necessarily talking about classes in the Clade Ontology here, but PHYX files?

I was thinking of the case our JSON-to-OWL conversion code or (more likely) our specifier matching code changes in such a way that an existing phyloreference no longer works as expected. It is intended to be a temporary state, after which the definition is corrected by a curator and the publication is returned to pso:published. Now that I think about it, the publication should probably be returned to pso:under-review in that case, so you're probably right that pso:retracted-from-publication is unnecessary.

I've filed an issue to detect and use owl:deprecated (#15) and to incorporate that into the Curation Tool (phyloref/curation-tool#81). Now that a mechanism for marking statuses of phyloreferences has been merged in #9, I'm going to close this issue -- we can track improvements to that system separately in different issues if needed.