Skip to content

Apache Lucene open-source search software

LUCENE-10408: Write doc IDs of KNN vectors as ints rather than vints. (#708)

Adrien Grand44d7d962ae42
diff --git a/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsFormat.java b/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsFormat.java
index a4f21b44ba7..274da2ccd58 100644
--- a/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsFormat.java
+++ b/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsFormat.java
@@ -69,13 +69,13 @@ import org.apache.lucene.util.hnsw.HnswGraph;
* <li><b>[int]</b> the number of documents having values for this field
* <li><b>[int8]</b> if equals to -1, dense – all documents have values for a field. If equals to
* 0, sparse – some documents missing values.
- * <li><b>array[vint]</b> for sparse case, the docids of documents having vectors, in order
+ * <li><b>array[int]</b> for sparse case, the docids of documents having vectors, in order
* <li><b>[int]</b> the maximum number of connections (neigbours) that each node can have
* <li><b>[int]</b> number of levels in the graph
* <li>Graph nodes by level. For each level
* <ul>
* <li><b>[int]</b> the number of nodes on this level
- * <li><b>array[vint]</b> for levels greater than 0 list of nodes on this level, stored as
+ * <li><b>array[int]</b> for levels greater than 0 list of nodes on this level, stored as
* the the level 0th nodes ordinals.
* </ul>
* </ul>
diff --git a/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsReader.java b/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsReader.java
index 014414882e7..84f5006a856 100644
--- a/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsReader.java
+++ b/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsReader.java
@@ -347,7 +347,7 @@ public final class Lucene91HnswVectorsReader extends KnnVectorsReader {
// as not all docs have vector values, fill a mapping from dense vector ordinals to docIds
ordToDoc = new int[size];
for (int i = 0; i < size; i++) {
- int doc = input.readVInt();
+ int doc = input.readInt();
ordToDoc[i] = doc;
}
}
@@ -366,7 +366,7 @@ public final class Lucene91HnswVectorsReader extends KnnVectorsReader {
} else {
nodesByLevel[level] = new int[numNodesOnLevel];
for (int i = 0; i < numNodesOnLevel; i++) {
- nodesByLevel[level][i] = input.readVInt();
+ nodesByLevel[level][i] = input.readInt();
}
}
}
diff --git a/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsWriter.java b/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsWriter.java
index c6ca1d59eee..6d42a5af321 100644
--- a/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsWriter.java
+++ b/lucene/core/src/java/org/apache/lucene/codecs/lucene91/Lucene91HnswVectorsWriter.java
@@ -213,7 +213,7 @@ public final class Lucene91HnswVectorsWriter extends KnnVectorsWriter {
meta.writeByte((byte) 0); // sparse marker, some documents don't have vector values
DocIdSetIterator iter = docsWithField.iterator();
for (int doc = iter.nextDoc(); doc != DocIdSetIterator.NO_MORE_DOCS; doc = iter.nextDoc()) {
- meta.writeVInt(doc);
+ meta.writeInt(doc);
}
}
@@ -229,7 +229,7 @@ public final class Lucene91HnswVectorsWriter extends KnnVectorsWriter {
if (level > 0) {
while (nodesOnLevel.hasNext()) {
int node = nodesOnLevel.nextInt();
- meta.writeVInt(node); // list of nodes on a level
+ meta.writeInt(node); // list of nodes on a level
}
}
}