-
Notifications
You must be signed in to change notification settings - Fork 0
Description
re3dataSearch resolver
parameters
-
limit to subjectArea
-
limit to repository type
-
search terms
-
Needs to be able to do paging (page number, current page, whatever the standard pattern we have for paging in graphQL)
Mutators
addRe3dataDoc -- this should take the standard re3data document we have defined in schema/ZOD (should be very much what we have in OpenSearch). It would upsert it (insert or update the current document with the same id)
removeRe3dataDoc -- remove the id from re3Data list.
connecting to opensearch
Is there an existing way that we're doing this through GraphQL? Seems like we should follow that pattern. Otherwise, I'm seeing libraries like "Searchkit" or I'm sure there is an OpenSearch library for Javascript that can be used inside resolvers.
Models
Do we really need to have models that model the search results? Perhaps we don't need models since OpenSearch already presents a model and I doubt we'll be doing granular changing of one field in an Opensearch record? Do we really need whatever caching? I assume that searches will be used the most and OpenSearch will be fast.