Support contributors field

[balhoff]

See INCATools/dead_simple_owl_design_patterns#46

[matentzn]

Can I already add the contributors field without breaking anything?

[dosumis]

Only once the pull request has been merged.

Based on discussion on pull request, this is still TBD:

• Specify ORCIDs (Seems reasonable to me)
• If we do specify ORCIDs do we specify a full IRI or Curie, or both?
• Recommend that if translated to OWL, the result should be a list of dc:contributor annotations on the term. (I think this is a no brainer.)

[matentzn]

Okay, I will raise the issue tomorrow at the meeting

[dosumis]

Currently we just suggest ORCIDs:

description: A list of authors of a pattern. We recommend that each author is specified using an ORCID (full URL). We do not recommend that this list is instantiated in terms generated using a pattern.

Making this mandatory & specifying how will allow us to add checks. OTOH - this is meant to be a very general spec that anyone could use for OWL patterns, so may be better to allow for some flexibility.

[matentzn]

I think validation should be just URI. This is good linked data standard and people can add orcid or whatever else they want to identify themselves.

[balhoff]

A CURIE should be fine as long as the prefix expansion is provided in the usual way.

[balhoff]

@matentzn dosdp-tools won't care about unknown keys, so you can add before dosdp-tools supports it.

[matentzn]

Problem could be davids Schema checker.. :P I can wait.

[dosumis]

YAML spec change now merged, so Python schema checker should work.

[balhoff]

I saw this comment in the schema:

We do not recommend that this list is instantiated in terms generated using a pattern, but where it is it should be instantiated as a set of annotation axioms using dc:contributor.

Does this mean there is no action needed in dosdp-tools in terms of generating axioms?

[matentzn]

The only thing i can think of is during the compilation of pattern.owl using this dosdp prototype tool. There I would like the dc:contributor tags! Not, for sure, on the compiled instances (think filesize)!

[balhoff]

That makes sense!