Add a TopDocsCollector which does not sort by score [LUCENE-8880]
- Dominant language
- Java
- Stars
- 3.6k
- Forks
- 1.4k
- Avg merge
- 2d 11h
- Merged PRs (30d)
- 88
Description
We assume that the user cares about the underlying hits being ordered by score. This Jira explores adding a collector which does not make this guarantee, thus not using priority queue as the collection data structure. This should help with large hits case, where the heap’s rebalancing can become a bottleneck
---
Migrated from [LUCENE-8880](https://issues.apache.org/jira/browse/LUCENE-8880) by Atri Sharma (@atris)
Contributor guide
Research direction
No file or test is identified in the issue. Start by locating TopDocsCollector and the existing collector tests, then determine the expected behavior for collecting hits without score ordering or a priority queue. Done means the new collector's behavior and performance trade-offs are covered by tests.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- java
- Domain
- search
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 30/100