Removes Array based vector serialization

Previously, we had serialized vectors via an array method. This was
inefficient and remove in the #253 PR and launced in 1.3.0. With that,
we no longer serialized new segment data via the array based serializer.
Now that it is 3.0, because the oldest index that can be upgraded is
from 1.3.0, we no longer need to handle array based serialization. So,
we can remove all of it.

Signed-off-by: John Mazanec <jmazane@amazon.com>

src/main/java/org/opensearch/knn/index/VectorDataType.java

@@ -12,8 +12,8 @@
 import org.apache.lucene.document.KnnFloatVectorField;
 import org.apache.lucene.index.VectorSimilarityFunction;
 import org.apache.lucene.util.BytesRef;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 import org.opensearch.knn.index.codec.util.KNNVectorSerializer;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
 import org.opensearch.knn.index.memory.NativeMemoryAllocation;
 import org.opensearch.knn.jni.JNICommons;
 import org.opensearch.knn.training.BinaryTrainingDataConsumer;
@@ -105,7 +105,7 @@ public FieldType createKnnVectorFieldType(int dimension, KNNVectorSimilarityFunc
 
         @Override
         public float[] getVectorFromBytesRef(BytesRef binaryValue) {
-            final KNNVectorSerializer vectorSerializer = KNNVectorSerializerFactory.getSerializerByBytesRef(binaryValue);
+            final KNNVectorSerializer vectorSerializer = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE;
             return vectorSerializer.byteToFloatArray(binaryValue);
         }
 

src/main/java/org/opensearch/knn/index/VectorField.java

@@ -8,15 +8,15 @@
 import org.apache.lucene.document.Field;
 import org.apache.lucene.index.IndexableFieldType;
 import org.apache.lucene.util.BytesRef;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 import org.opensearch.knn.index.codec.util.KNNVectorSerializer;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
 
 public class VectorField extends Field {
 
     public VectorField(String name, float[] value, IndexableFieldType type) {
         super(name, new BytesRef(), type);
         try {
-            final KNNVectorSerializer vectorSerializer = KNNVectorSerializerFactory.getDefaultSerializer();
+            final KNNVectorSerializer vectorSerializer = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE;
             final byte[] floatToByte = vectorSerializer.floatToByteArray(value);
             this.setBytesValue(floatToByte);
         } catch (Exception e) {

src/main/java/org/opensearch/knn/index/codec/util/KNNVectorAsCollectionOfFloatsSerializer.java

@@ -17,6 +17,8 @@
 public class KNNVectorAsCollectionOfFloatsSerializer implements KNNVectorSerializer {
     private static final int BYTES_IN_FLOAT = 4;
 
+    public static final KNNVectorAsCollectionOfFloatsSerializer INSTANCE = new KNNVectorAsCollectionOfFloatsSerializer();
+
     @Override
     public byte[] floatToByteArray(float[] input) {
         final ByteBuffer bb = ByteBuffer.allocate(input.length * BYTES_IN_FLOAT).order(ByteOrder.BIG_ENDIAN);

src/main/java/org/opensearch/knn/index/mapper/KNNVectorFieldMapperUtil.java

@@ -24,7 +24,7 @@
 import org.opensearch.knn.index.KnnCircuitBreakerException;
 import org.opensearch.knn.index.SpaceType;
 import org.opensearch.knn.index.VectorDataType;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 import org.opensearch.knn.index.engine.KNNEngine;
 import org.opensearch.knn.index.engine.KNNMethodContext;
 import org.opensearch.knn.index.engine.MethodComponentContext;
@@ -74,7 +74,7 @@ public static StoredField createStoredFieldForByteVector(String name, byte[] vec
      * @param vector vector to be added to stored field
      */
     public static StoredField createStoredFieldForFloatVector(String name, float[] vector) {
-        return new StoredField(name, KNNVectorSerializerFactory.getDefaultSerializer().floatToByteArray(vector));
+        return new StoredField(name, KNNVectorAsCollectionOfFloatsSerializer.INSTANCE.floatToByteArray(vector));
     }
 
     /**

src/main/java/org/opensearch/knn/index/vectorvalues/VectorValueExtractorStrategy.java

@@ -13,8 +13,8 @@
 import org.apache.lucene.util.ArrayUtil;
 import org.apache.lucene.util.BytesRef;
 import org.opensearch.knn.index.VectorDataType;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 import org.opensearch.knn.index.codec.util.KNNVectorSerializer;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
 
 import java.io.IOException;
 
@@ -121,7 +121,7 @@ private <T> T extractFromKnnVectorValues(
         }
 
         private float[] getFloatVectorFromByteRef(final BytesRef bytesRef) {
-            final KNNVectorSerializer vectorSerializer = KNNVectorSerializerFactory.getSerializerByBytesRef(bytesRef);
+            final KNNVectorSerializer vectorSerializer = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE;
             return vectorSerializer.byteToFloatArray(bytesRef);
         }
     }

src/test/java/org/opensearch/knn/index/KNNVectorDVLeafFieldDataTests.java

@@ -18,7 +18,7 @@
 import org.apache.lucene.store.Directory;
 import org.opensearch.index.fielddata.ScriptDocValues;
 import org.junit.Before;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 
 import java.io.IOException;
 
@@ -43,7 +43,7 @@ private void createKNNVectorDocument(Directory directory) throws IOException {
         IndexWriterConfig conf = newIndexWriterConfig(new MockAnalyzer(random()));
         IndexWriter writer = new IndexWriter(directory, conf);
         Document knnDocument = new Document();
-        byte[] vectorBinary = KNNVectorSerializerFactory.getDefaultSerializer().floatToByteArray(new float[] { 1.0f, 2.0f });
+        byte[] vectorBinary = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE.floatToByteArray(new float[] { 1.0f, 2.0f });
         knnDocument.add(new BinaryDocValuesField(MOCK_INDEX_FIELD_NAME, new BytesRef(vectorBinary)));
         knnDocument.add(new NumericDocValuesField(MOCK_NUMERIC_INDEX_FIELD_NAME, 1000));
         writer.addDocument(knnDocument);

src/test/java/org/opensearch/knn/index/KNNVectorScriptDocValuesTests.java

@@ -19,7 +19,7 @@
 import org.junit.Before;
 import org.junit.After;
 import org.junit.Test;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 
 import java.io.IOException;
 
@@ -167,7 +167,7 @@ private void createKNNVectorDocument(Directory directory, Class<?> valuesClass)
         Field field;
 
         if (BinaryDocValues.class.equals(valuesClass)) {
-            byte[] vectorBinary = KNNVectorSerializerFactory.getDefaultSerializer().floatToByteArray(SAMPLE_VECTOR_DATA);
+            byte[] vectorBinary = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE.floatToByteArray(SAMPLE_VECTOR_DATA);
             field = new BinaryDocValuesField(MOCK_INDEX_FIELD_NAME, new BytesRef(vectorBinary));
         } else if (ByteVectorValues.class.equals(valuesClass)) {
             field = new KnnByteVectorField(MOCK_INDEX_FIELD_NAME, SAMPLE_BYTE_VECTOR_DATA);

src/test/java/org/opensearch/knn/index/VectorDataTypeTests.java

@@ -17,7 +17,7 @@
 import org.apache.lucene.util.BytesRef;
 import org.junit.Assert;
 import org.opensearch.knn.KNNTestCase;
-import org.opensearch.knn.index.codec.util.KNNVectorSerializerFactory;
+import org.opensearch.knn.index.codec.util.KNNVectorAsCollectionOfFloatsSerializer;
 
 import java.io.IOException;
 
@@ -82,7 +82,7 @@ private void createKNNFloatVectorDocument(Directory directory) throws IOExceptio
         IndexWriterConfig conf = newIndexWriterConfig(new MockAnalyzer(random()));
         IndexWriter writer = new IndexWriter(directory, conf);
         Document knnDocument = new Document();
-        BytesRef bytesRef = new BytesRef(KNNVectorSerializerFactory.getDefaultSerializer().floatToByteArray(SAMPLE_FLOAT_VECTOR_DATA));
+        BytesRef bytesRef = new BytesRef(KNNVectorAsCollectionOfFloatsSerializer.INSTANCE.floatToByteArray(SAMPLE_FLOAT_VECTOR_DATA));
         knnDocument.add(new BinaryDocValuesField(MOCK_FLOAT_INDEX_FIELD_NAME, bytesRef));
         writer.addDocument(knnDocument);
         writer.commit();

src/test/java/org/opensearch/knn/index/codec/util/KNNVectorSerializerTests.java

@@ -10,73 +10,13 @@
 
 import java.io.ByteArrayOutputStream;
 import java.io.DataOutputStream;
-import java.io.ObjectOutputStream;
 import java.util.Random;
 import java.util.stream.IntStream;
 
 public class KNNVectorSerializerTests extends KNNTestCase {
 
     Random random = new Random();
 
-    public void testVectorSerializerFactory() throws Exception {
-        // check that default serializer can work with array of floats
-        // setup
-        final float[] vector = getArrayOfRandomFloats(20);
-        final ByteArrayOutputStream bas = new ByteArrayOutputStream();
-        final DataOutputStream ds = new DataOutputStream(bas);
-        for (float f : vector)
-            ds.writeFloat(f);
-        final BytesRef vectorAsCollectionOfFloats = new BytesRef(bas.toByteArray());
-        final KNNVectorSerializer defaultSerializer = KNNVectorSerializerFactory.getDefaultSerializer();
-        assertNotNull(defaultSerializer);
-
-        final float[] actualDeserializedVector = defaultSerializer.byteToFloatArray(vectorAsCollectionOfFloats);
-        assertNotNull(actualDeserializedVector);
-        assertArrayEquals(vector, actualDeserializedVector, 0.1f);
-
-        final KNNVectorSerializer arraySerializer = KNNVectorSerializerFactory.getSerializerBySerializationMode(SerializationMode.ARRAY);
-        assertNotNull(arraySerializer);
-
-        final KNNVectorSerializer collectionOfFloatsSerializer = KNNVectorSerializerFactory.getSerializerBySerializationMode(
-            SerializationMode.COLLECTION_OF_FLOATS
-        );
-        assertNotNull(collectionOfFloatsSerializer);
-    }
-
-    public void testVectorSerializerFactory_throwExceptionForBytesWithUnsupportedDataType() throws Exception {
-        // prepare array of chars that is not supported by serializer factory. expected behavior is to fail
-        final char[] arrayOfChars = new char[] { 'a', 'b', 'c' };
-        final ByteArrayOutputStream bas = new ByteArrayOutputStream();
-        final DataOutputStream ds = new DataOutputStream(bas);
-        for (char ch : arrayOfChars)
-            ds.writeChar(ch);
-        final BytesRef vectorAsCollectionOfChars = new BytesRef(bas.toByteArray());
-
-        expectThrows(RuntimeException.class, () -> KNNVectorSerializerFactory.getSerializerByBytesRef(vectorAsCollectionOfChars));
-    }
-
-    public void testVectorAsArraySerializer() throws Exception {
-        final float[] vector = getArrayOfRandomFloats(20);
-
-        final ByteArrayOutputStream byteStream = new ByteArrayOutputStream();
-        final ObjectOutputStream objectStream = new ObjectOutputStream(byteStream);
-        objectStream.writeObject(vector);
-        final BytesRef serializedVector = new BytesRef(byteStream.toByteArray());
-        final KNNVectorSerializer vectorSerializer = KNNVectorSerializerFactory.getSerializerByBytesRef(serializedVector);
-
-        // testing serialization
-        final byte[] actualSerializedVector = vectorSerializer.floatToByteArray(vector);
-
-        assertNotNull(actualSerializedVector);
-        assertArrayEquals(serializedVector.bytes, actualSerializedVector);
-
-        // testing deserialization
-        final float[] actualDeserializedVector = vectorSerializer.byteToFloatArray(serializedVector);
-
-        assertNotNull(actualDeserializedVector);
-        assertArrayEquals(vector, actualDeserializedVector, 0.1f);
-    }
-
     public void testVectorAsCollectionOfFloatsSerializer() throws Exception {
         // setup
         final float[] vector = getArrayOfRandomFloats(20);
@@ -86,7 +26,7 @@ public void testVectorAsCollectionOfFloatsSerializer() throws Exception {
         for (float f : vector)
             ds.writeFloat(f);
         final BytesRef vectorAsCollectionOfFloats = new BytesRef(bas.toByteArray());
-        final KNNVectorSerializer vectorSerializer = KNNVectorSerializerFactory.getSerializerByBytesRef(vectorAsCollectionOfFloats);
+        final KNNVectorSerializer vectorSerializer = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE;
 
         // testing serialization
         final byte[] actualSerializedVector = vectorSerializer.floatToByteArray(vector);
@@ -104,16 +44,14 @@ public void testVectorAsCollectionOfFloatsSerializer() throws Exception {
     public void testVectorSerializer_whenVectorBytesOffset_thenSuccess() {
         final float[] vector = getArrayOfRandomFloats(20);
         int offset = randomInt(4);
-        for (SerializationMode serializationMode : SerializationMode.values()) {
-            final KNNVectorSerializer vectorSerializer = KNNVectorSerializerFactory.getSerializerBySerializationMode(serializationMode);
-            assertNotNull(vectorSerializer);
-            byte[] bytes = vectorSerializer.floatToByteArray(vector);
-            byte[] bytesWithOffset = new byte[bytes.length + 2 * offset];
-            System.arraycopy(bytes, 0, bytesWithOffset, offset, bytes.length);
-            BytesRef serializedVector = new BytesRef(bytesWithOffset, offset, bytes.length);
-            float[] deserializedVector = vectorSerializer.byteToFloatArray(serializedVector);
-            assertArrayEquals(vector, deserializedVector, 0.1f);
-        }
+        final KNNVectorSerializer vectorSerializer = KNNVectorAsCollectionOfFloatsSerializer.INSTANCE;
+        assertNotNull(vectorSerializer);
+        byte[] bytes = vectorSerializer.floatToByteArray(vector);
+        byte[] bytesWithOffset = new byte[bytes.length + 2 * offset];
+        System.arraycopy(bytes, 0, bytesWithOffset, offset, bytes.length);
+        BytesRef serializedVector = new BytesRef(bytesWithOffset, offset, bytes.length);
+        float[] deserializedVector = vectorSerializer.byteToFloatArray(serializedVector);
+        assertArrayEquals(vector, deserializedVector, 0.1f);
     }
 
     private float[] getArrayOfRandomFloats(int arrayLength) {