lucene

lucene copied to clipboard

As the title said, we could use the skip functionality of the doc value sparse index to improve the performance of NumericLeafComparator.

iverase

Can IndexSearcher enable search concurrency by default? [LUCENE-10487]

2

IndexSearcher has the ability to search an index using multiple threads, but this behavior is an opt-in. Could we enable it by default so that:  - Users experience better latency...

asfimport

Spatial 3d performance improvements and refactorings

1

### Description (proxying for kwright) A major restructuring and massive performance improvement for spatial3d: basically the path pieces are now in a balanced b-tree and thus isWithin() operations are log(n)...

dweiss

"gradlew clean check" results in internal gradle error "Unable to make progress running work."

1

### Description As reported by Chris, running "gradlew clean check" on main results in this nasty (or its variations): ``` Unable to make progress running work. The following items are...

dweiss

Add higher quantization level for kNN vector search

1

### Description There have been multiple discussions around supporting higher vector quantization levels (e.g. more than scalar quantization can provide). We (elastic) have been doing some research around what the...

benwtrent

Use range optimizations for "slow" MultiTermQueries when terms happen to be contiguous

5

### Description This is a small optimization that treats "slow" multi-term queries (e.g., TermInSet, RegexpQuery, etc.) as ordinal ranges when the query terms create a contiguous range. This is likely...

gsmiller

tests fail due to large amount of output with highish tests.iters

2

### Description I've noticed that some tests fail when `tests.iters` is set to a highish value (like 20) because stdout grows above the allowed threshold. I assume this is something...

msokolov

Update QueryUtils to use CollectorManager

2

### Description Relates to https://github.com/apache/lucene/issues/12892 We introduce a new `SimpleCollectorWithLastReader` class to help keep context of LastReader per thread. This way calling `CollectorManager#reduce` can work with an executor. The diff...

harshavamsi

Add factory methods for IndexWriterConfig.

1

This adds factory methods for `IndexWriterConfig` that are optimized for a few different use-cases. Configuring an `IndexWriterConfig` is quite an expert task, the goal is to give sensible defaults for...

jpountz

Should EdgeNGramTokenizer's DEFAULT_MAX_GRAM_SIZE be ONE?

2

### Description From org.apache.lucene:lucene-analysis-common:9.11.1, the static variable `DEFAULT_MAX_GRAM_SIZE` of EdgeNGramTokenizer is ONE not TWO. Logically, the maximum n-gram size must be >= minGramSize and it's not a problem but **NOT...

YeonghyeonKO