Is there a recommended way to use CompositeAggregation with kt-search?

[nfa-dev]

Hi, I'm using kt-search (version 2.0.3) in a Spring Boot project and would like to use CompositeAggregation to paginate over a large result set.

I’ve read through the official documentation here:
https://jillesvangurp.github.io/kt-search/manual/Aggregations.html

But I couldn’t find any documentation or examples related to how to use CompositeAggregation, especially with paging using afterKey.

Why is this needed?

Composite aggregation is the recommended way in Elasticsearch to get accurate distinct field values for large sets,
but it's tricky to implement correctly without guidance on how to handle afterKey and buckets safely across pages.

Could you provide any guidance or a working example?

Thanks for the awesome DSL!

[jillesvangurp]

I don't currently have any support for that and it looks like it's a relatively new feature that I haven't used yet myself. You should be able by just accessing the JsonDsl map directly. Or you could create a PR that adds the support if you feel up to it.

BTW. Using this in combination with a large result set sounds like it could be expensive.