Aggregation updates

[StepanBrychta]

What does this change?

wellcomecollection/platform#5825

Changes how we handle aggregations in Elasticsearch queries, following on changes to the catalogue pipeline.

The API now supports two aggregation types:

• Labelled ID aggregations: Consist of an ID-based terms aggregation with a label-based sub-aggregation to retrieve all labels associated with a given ID.
• Label-only aggregations: Consist of a label-based terms aggregation with no sub-aggregation. Included to keep supporting label-based concept aggregations until we switch to ID-based aggregations/filters.

All existing production functionality should remain the same. From the frontend's perspective, the only change to the API involves renaming ID-based filters (genres.concepts→genres), but the affected filters are only used behind a feature flag.

Note

This branch has been rebased to #829.

Checklist

• Does this patch need a change to the documentation?
• Do you need to update the Catalogue API Swagger?: Add ID-based concept aggregations #5825 developers.wellcomecollection.org#44

How to test

How can we measure success?

Added support for ID-based aggregations without negative side effects and without losing support for label-based concept aggregations.

Have we considered potential risks?

[kenoir]

Is the use of .id in keeping with principle 1 in the API faceting RFC?

Filters are named by the JSON paths of the identified object that they filter or, if applied to an attribute other than the identifier, the path of that attribute

I would have expected the if filter to just be for example source.contributors.agent, omitting the ID?

[StepanBrychta]

Yes, this should now be fixed