Add publication resolver to person

[outten45]

Currently, the graphql schema has a minimal nested objects. This would be to add a call to the resolver to get the full publication with (hopefully) all of the pagination and faceting available.

[jimwood]

An example of this is working in the vstf-73-person-publication-resolver branch. Front-end work can use the 'person' resolver in this branch. This is not ready to merge pending some open questions:

This commit retrieves an accurate list of publications with batched queries:
9dba7cb

This one tries to address those concerns by using a single fetch based on 'syncIds':
aa251bc

This may not be accurate since syncIds may contain person ids for editors or other roles.

The question then is, can role specific id fields be added to the Document schema (authorIds, editorIds)?

Also for discussion, should we limit exposed resolvers for each type to a singular (ie person to fetch the EnhancedPerons with fully resolved associations) and a plural (ie people that allows query, sorting, filtering, faceting, paging on the nested schema) for simplicity? That would handle search/browse pages and entity pages. I started this:

35dd150

But ran out of time and Java knowledge to complete.

[ghost]

Instead of adding an indexed field of ids for each relationship, a filter can be implemented for the specific desired relationship. In this case, the documents collection is queried for all Solr documents matching the syncIds and filtered by author matching the person.

[ghost]

https://github.com/vivo-community/scholars-discovery/pull/112/files

[ghost]

Here is a related issue, #117.