Jim Ferenczi

LUCENE-10059: Additional fix to handle n_best backtrace

4

This commit handles possible empty backtrace when using the nbest option. The general case (best path) was fixed in a previous commit so this change is a follow up to...

LUCENE-10086: Fix an AssertionError when KoreanTokenizer tries to backtrace from and to the same position

3

This change is a follow up of https://issues.apache.org/jira/browse/LUCENE-10059 for the Korean tokenizer. #11124

Allow rescorer with field collapsing

3

This change adds the support for rescoring collapsed documents. The rescoring is applied on the top document per group on each shard. Closes #27243

Fix cluster level dense vector stats

2

The cluster level dense vector stats returns the total number of dense vector indices globally including the replicas. This commit fixes the total to only include the value count of...

Refine indexing pressure accounting in semantic bulk inference filter

3

In #125517, we estimated that inference results would double the document _source size since they are pooled by the bulk action. This PR reduces the memory needed to perform the...