Can we optimize TaxonomyFacetSumIntAssociations#sumValues? [LUCENE-10373]
- Dominant language
- Java
- Stars
- 3.6k
- Forks
- 1.4k
- Avg merge
- 2d 11h
- Merged PRs (30d)
- 88
Description
It looks like `TaxonomyFacetSumIntAssociations#sumValues` might also benefit from reduced `null` checks just like in #11386 (doing away with the use of the parent class `increment` method).
While we're at it, it would be nice to test moving to little endian encoding of the ord/value since we're trying to deprecate `BitUtil.VH_BE_INT`}.
The real challenge is that we don't have benchmark tasks for this class. Maybe we should add benchmarking first? I opened a `luceneutil` issue for this: https://github.com/mikemccand/luceneutil/issues/158
---
Migrated from [LUCENE-10373](https://issues.apache.org/jira/browse/LUCENE-10373) by Greg Miller (@gsmiller), updated Jan 18 2022
Attachments: [LUCENE-10373.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-10373/LUCENE-10373.patch)
Contributor guide
Research direction
Start at TaxonomyFacetSumIntAssociations#sumValues and compare the null-check optimization described in issue #11386, including the parent increment method. Review the linked luceneutil issue about benchmark tasks before evaluating little-endian ord/value encoding; done means measured benchmark results and a justified implementation decision.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- java
- Domain
- performance, search
- Issue type
- Refactor
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100