Skip to content

Apache Lucene open-source search software

Fix TestBpVectorReorderer#testIndexReorderDense (#14153)

Ignacio Veraa0adeb9fa060
diff --git a/lucene/misc/src/test/org/apache/lucene/misc/index/TestBpVectorReorderer.java b/lucene/misc/src/test/org/apache/lucene/misc/index/TestBpVectorReorderer.java
index e4398a76183..3b484ed2430 100644
--- a/lucene/misc/src/test/org/apache/lucene/misc/index/TestBpVectorReorderer.java
+++ b/lucene/misc/src/test/org/apache/lucene/misc/index/TestBpVectorReorderer.java
@@ -318,10 +318,7 @@ public class TestBpVectorReorderer extends LuceneTestCase {
public void testIndexReorderDense() throws Exception {
List<float[]> vectors = shuffleVectors(randomLinearVectors());
- // compute the expected ordering
- Sorter.DocMap expected =
- reorderer.computeValueMap(
- FloatVectorValues.fromFloats(vectors, 2), VectorSimilarityFunction.EUCLIDEAN, null);
+
Path tmpdir = createTempDir();
try (Directory dir = newFSDirectory(tmpdir)) {
// create an index with a single leaf
@@ -335,6 +332,28 @@ public class TestBpVectorReorderer extends LuceneTestCase {
}
writer.forceMerge(1);
}
+
+ // The docId of the documents might have changed due to merging. Compute a mapping from
+ // the stored id to the current docId and repopulate the vector list.
+ int[] storedIdToDocId = new int[vectors.size()];
+ vectors.clear();
+ try (IndexReader reader = DirectoryReader.open(dir)) {
+ LeafReader leafReader = getOnlyLeafReader(reader);
+ FloatVectorValues values = leafReader.getFloatVectorValues("f");
+ StoredFields storedFields = reader.storedFields();
+ KnnVectorValues.DocIndexIterator it = values.iterator();
+ while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
+ int storedId = Integer.parseInt(storedFields.document(it.docID()).get("id"));
+ vectors.add(values.vectorValue(it.index()).clone());
+ storedIdToDocId[storedId] = it.docID();
+ }
+ }
+
+ // compute the expected ordering
+ Sorter.DocMap expected =
+ reorderer.computeValueMap(
+ FloatVectorValues.fromFloats(vectors, 2), VectorSimilarityFunction.EUCLIDEAN, null);
+
int threadCount = random().nextInt(4) + 1;
threadCount = 1;
// reorder using the index reordering tool
@@ -355,12 +374,13 @@ public class TestBpVectorReorderer extends LuceneTestCase {
StoredFields storedFields = reader.storedFields();
KnnVectorValues.DocIndexIterator it = values.iterator();
while (it.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
- int storedId = Integer.parseInt(storedFields.document(it.docID()).get("id"));
- assertEquals(expected.oldToNew(storedId), newId);
+ int oldDocId =
+ storedIdToDocId[Integer.parseInt(storedFields.document(it.docID()).get("id"))];
+ assertEquals(expected.oldToNew(oldDocId), newId);
float[] expectedVector = vectors.get(expected.newToOld(it.docID()));
float[] actualVector = values.vectorValue(it.index());
assertArrayEquals(
- "values differ at index " + storedId + "->" + newId + " docid=" + it.docID(),
+ "values differ at index " + oldDocId + "->" + newId + " docid=" + it.docID(),
expectedVector,
actualVector,
0);