Skip to content

Apache Lucene open-source search software

Use TernaryLongHeap in UpdateGraphsUtils.computeJoinSet() (#16153)

Prithvi Sca26239b3397
diff --git a/lucene/CHANGES.txt b/lucene/CHANGES.txt
index bc49af53e04..ec542061d6f 100644
--- a/lucene/CHANGES.txt
+++ b/lucene/CHANGES.txt
@@ -379,6 +379,8 @@ Improvements
Optimizations
---------------------
+* GITHUB#16153: Use TernaryLongHeap in UpdateGraphsUtils for faster HNSW graph merging. (Prithvi S)
+
* GITHUB#15861: Optimise PhraseScorer by short circuiting non competitive documents in TOP_SCORES mode. (Prithvi S)
* GITHUB#16154: Use Lucene's PriorityQueue in NearestNeighbor to avoid poll/offer overhead on the hot path. (Prithvi S)
diff --git a/lucene/core/src/java/org/apache/lucene/util/hnsw/UpdateGraphsUtils.java b/lucene/core/src/java/org/apache/lucene/util/hnsw/UpdateGraphsUtils.java
index 667ba37b7e3..ce103952d50 100644
--- a/lucene/core/src/java/org/apache/lucene/util/hnsw/UpdateGraphsUtils.java
+++ b/lucene/core/src/java/org/apache/lucene/util/hnsw/UpdateGraphsUtils.java
@@ -21,7 +21,7 @@ import static org.apache.lucene.search.DocIdSetIterator.NO_MORE_DOCS;
import java.io.IOException;
import org.apache.lucene.internal.hppc.IntHashSet;
-import org.apache.lucene.util.LongHeap;
+import org.apache.lucene.util.TernaryLongHeap;
/**
* Utility class for updating a big graph with smaller graphs. This is used during merging of
@@ -41,7 +41,7 @@ public class UpdateGraphsUtils {
public static IntHashSet computeJoinSet(HnswGraph graph) throws IOException {
int k; // coverage for the current node
int size = graph.size();
- LongHeap heap = new LongHeap(size);
+ TernaryLongHeap heap = new TernaryLongHeap(size);
IntHashSet j = new IntHashSet();
boolean[] stale = new boolean[size];
short[] counts = new short[size];