GitLab

Add support for searching Catalog data using Elasticsearch.

Indexing

1. Entities should be automatically added and removed from ES (on JPA Save/Remove)
2. ES index must be configured to index entity data and all relevant related data.

Search

1. Develop individual API endpoints for datasets, descriptors, descriptor lists
2. API search endpoint will search ES and return matching entities

Search results

Search results returned by the API are grouped by "entity": descriptor terms, crops, institutes, etc. and ordered by relevance to the search.

The selected hits are then applied as filter values to the filtering API. For that to work we need to know how various entities relate to what we are searching.

Example: Partners

Partners database accepts the following filters:

{
 "shortName": [ "string1", "string2" ],
 "name": { StringFilter },
 "description": { StringFilter },
 "wiewsCodes": [ "WIEWS1", "WIEWS2", "WIEWS3" ]
}

When the Search API returns matching records from the "FAO WIEWS Institute codes" vocabulary, these hits can only be applied to the wiewsCode Partner filter.

Example: Dataset

Dataset filtering accepts:

{
 "owner": { PartnerFilter },
 "accessionIdentifier": { AccessionIdentifierFilter },
 "descriptor": { DescriptorFilter },
 "crop": [ "crop1", "crop2" ],
 ...
}

Hits from Crop entity can be applied to crop filter, hits from Descriptor entity can be applied to descriptor.id filter.