Consider removing CachingCollector [LUCENE-9361]
Open
legacy-jira-priority:Major
type:enhancement
- Dominant language
- Java
- Stars
- 3.6k
- Forks
- 1.4k
- Avg merge
- 2d 11h
- Merged PRs (30d)
- 88
Description
We have a caching collector implementation that is used by the grouping module to avoid running queries twice when doing two-phase grouping queries. This stores things in a flat array of int[maxDoc], and optionally stores scores as well in a parallel float[] array. Given that we have a much more efficient way of caching matching documents per-segment in the built-in QueryCache, should we consider removing this collector entirely and advise users to configure query caches appropriately instead?
---
Migrated from [LUCENE-9361](https://issues.apache.org/jira/browse/LUCENE-9361) by Alan Woodward (@romseygeek)
Contributor guide
Assessment
This issue has not been assessed yet.