Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
Expand Up @@ -102,6 +102,16 @@
* }</pre>
*
* <p>
* AWS OpenSearch Serverless usage example:
* </p>
* <pre>{@code
* OpenSearchVectorStore vectorStore = OpenSearchVectorStore.builder(openSearchClient, embeddingModel)
* .initializeSchema(true)
* .manageDocumentIds(false) // Required for AWS OpenSearch Serverless
* .build();
* }</pre>
*
* <p>
* Advanced configuration example:
* </p>
* <pre>{@code
Expand Down Expand Up @@ -137,6 +147,7 @@
* @author Christian Tzolov
* @author Thomas Vitale
* @author inpink
* @author Sanghun Lee
* @since 1.0.0
*/
public class OpenSearchVectorStore extends AbstractObservationVectorStore implements InitializingBean {
Expand Down Expand Up @@ -174,6 +185,8 @@ public class OpenSearchVectorStore extends AbstractObservationVectorStore implem

private final int dimensions;

private final boolean manageDocumentIds;
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Shouldn't this be true by default to let Spring AI managing the documentIDs and for backwards compatibility?.


/**
* Creates a new OpenSearchVectorStore using the builder pattern.
* @param builder The configured builder instance
Expand All @@ -193,6 +206,7 @@ protected OpenSearchVectorStore(Builder builder) {
this.initializeSchema = builder.initializeSchema;
this.useApproximateKnn = builder.useApproximateKnn;
this.dimensions = builder.dimensions;
this.manageDocumentIds = builder.manageDocumentIds;
}

/**
Expand All @@ -216,14 +230,27 @@ public void doAdd(List<Document> documents) {
for (Document document : documents) {
OpenSearchDocument openSearchDocument = new OpenSearchDocument(document.getId(), document.getText(),
document.getMetadata(), embedding.get(documents.indexOf(document)));
bulkRequestBuilder.operations(op -> op
.index(idx -> idx.index(this.index).id(openSearchDocument.id()).document(openSearchDocument)));

// Conditionally set document ID based on manageDocumentIds flag
if (this.manageDocumentIds) {
bulkRequestBuilder.operations(op -> op
.index(idx -> idx.index(this.index).id(openSearchDocument.id()).document(openSearchDocument)));
}
else {
bulkRequestBuilder
.operations(op -> op.index(idx -> idx.index(this.index).document(openSearchDocument)));
}
}
bulkRequest(bulkRequestBuilder.build());
}

@Override
public void doDelete(List<String> idList) {
if (!this.manageDocumentIds) {
logger.warn("Document ID management is disabled. Delete operations may not work as expected "
+ "since document IDs are auto-generated by OpenSearch. Consider using filter-based deletion instead.");
}

BulkRequest.Builder bulkRequestBuilder = new BulkRequest.Builder();
for (String id : idList) {
bulkRequestBuilder.operations(op -> op.delete(idx -> idx.index(this.index).id(id)));
Expand Down Expand Up @@ -481,6 +508,8 @@ public static class Builder extends AbstractVectorStoreBuilder<Builder> {

private int dimensions = 1536;

private boolean manageDocumentIds = true;

/**
* Sets the OpenSearch client.
* @param openSearchClient The OpenSearch client to use
Expand Down Expand Up @@ -585,6 +614,28 @@ public Builder dimensions(int dimensions) {
return this;
}

/**
* Sets whether to manage document IDs during indexing operations.
* <p>
* When set to {@code true} (default), document IDs will be explicitly set during
* indexing operations. When set to {@code false}, OpenSearch will auto-generate
* document IDs, which is required for AWS OpenSearch Serverless vector search
* collections.
* </p>
* <p>
* Note: When document ID management is disabled, the {@link #doDelete(List)}
* method may not work as expected since document IDs are auto-generated by
* OpenSearch.
* </p>
* @param manageDocumentIds true to manage document IDs (default), false to let
* OpenSearch auto-generate IDs
* @return The builder instance
*/
public Builder manageDocumentIds(boolean manageDocumentIds) {
this.manageDocumentIds = manageDocumentIds;
return this;
}

/**
* Builds a new OpenSearchVectorStore instance with the configured properties.
* @return A new OpenSearchVectorStore instance
Expand Down
Original file line number Diff line number Diff line change
@@ -0,0 +1,206 @@
/*
* Copyright 2023-2025 the original author or authors.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* https://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/

package org.springframework.ai.vectorstore.opensearch;

import java.io.IOException;
import java.util.List;
import java.util.Map;

import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.DisplayName;
import org.junit.jupiter.api.Test;
import org.junit.jupiter.api.extension.ExtendWith;
import org.junit.jupiter.params.ParameterizedTest;
import org.junit.jupiter.params.provider.ValueSource;
import org.mockito.ArgumentCaptor;
import org.mockito.Mock;
import org.mockito.junit.jupiter.MockitoExtension;
import org.opensearch.client.opensearch.OpenSearchClient;
import org.opensearch.client.opensearch.core.BulkRequest;
import org.opensearch.client.opensearch.core.BulkResponse;

import org.springframework.ai.document.Document;
import org.springframework.ai.embedding.EmbeddingModel;

import static org.assertj.core.api.Assertions.assertThat;
import static org.assertj.core.api.Assertions.assertThatThrownBy;
import static org.mockito.ArgumentMatchers.any;
import static org.mockito.Mockito.lenient;
import static org.mockito.Mockito.verify;
import static org.mockito.Mockito.when;

/**
* Unit tests for OpenSearchVectorStore.doAdd() method.
*
* Focuses on testing the manageDocumentIds functionality and document ID handling.
*/
@ExtendWith(MockitoExtension.class)
@DisplayName("OpenSearchVectorStore.doAdd() Tests")
class OpenSearchVectorStoreTest {

@Mock
private OpenSearchClient mockOpenSearchClient;

@Mock
private EmbeddingModel mockEmbeddingModel;

@Mock
private BulkResponse mockBulkResponse;

@BeforeEach
void setUp() throws IOException {
// Use lenient to avoid UnnecessaryStubbingException
lenient().when(this.mockEmbeddingModel.dimensions()).thenReturn(3);
lenient().when(this.mockOpenSearchClient.bulk(any(BulkRequest.class))).thenReturn(this.mockBulkResponse);
lenient().when(this.mockBulkResponse.errors()).thenReturn(false);
}

@ParameterizedTest(name = "manageDocumentIds={0}")
@ValueSource(booleans = { true, false })
@DisplayName("Should handle document ID management setting correctly")
void shouldHandleDocumentIdManagementSetting(boolean manageDocumentIds) throws IOException {
// Given
when(this.mockEmbeddingModel.embed(any(), any(), any()))
.thenReturn(List.of(new float[] { 0.1f, 0.2f, 0.3f }, new float[] { 0.4f, 0.5f, 0.6f }));

OpenSearchVectorStore vectorStore = createVectorStore(manageDocumentIds);
List<Document> documents = List.of(new Document("doc1", "content1", Map.of()),
new Document("doc2", "content2", Map.of()));

// When
vectorStore.add(documents);

// Then
BulkRequest capturedRequest = captureBulkRequest();
assertThat(capturedRequest.operations()).hasSize(2);

verifyDocumentIdHandling(capturedRequest, manageDocumentIds);
}

@Test
@DisplayName("Should handle single document correctly")
void shouldHandleSingleDocumentCorrectly() throws IOException {
// Given
when(this.mockEmbeddingModel.embed(any(), any(), any())).thenReturn(List.of(new float[] { 0.1f, 0.2f, 0.3f }));

OpenSearchVectorStore vectorStore = createVectorStore(true);
Document document = new Document("test-id", "test content", Map.of("key", "value"));

// When
vectorStore.add(List.of(document));

// Then
BulkRequest capturedRequest = captureBulkRequest();
var operation = capturedRequest.operations().get(0);

assertThat(operation.isIndex()).isTrue();
assertThat(operation.index().id()).isEqualTo("test-id");
assertThat(operation.index().document()).isNotNull();
}

@Test
@DisplayName("Should handle multiple documents with explicit IDs")
void shouldHandleMultipleDocumentsWithExplicitIds() throws IOException {
// Given
when(this.mockEmbeddingModel.embed(any(), any(), any())).thenReturn(List.of(new float[] { 0.1f, 0.2f, 0.3f },
new float[] { 0.4f, 0.5f, 0.6f }, new float[] { 0.7f, 0.8f, 0.9f }));

OpenSearchVectorStore vectorStore = createVectorStore(true);
List<Document> documents = List.of(new Document("doc1", "content1", Map.of()),
new Document("doc2", "content2", Map.of()), new Document("doc3", "content3", Map.of()));

// When
vectorStore.add(documents);

// Then
BulkRequest capturedRequest = captureBulkRequest();
assertThat(capturedRequest.operations()).hasSize(3);

for (int i = 0; i < 3; i++) {
var operation = capturedRequest.operations().get(i);
assertThat(operation.isIndex()).isTrue();
assertThat(operation.index().id()).isEqualTo("doc" + (i + 1));
}
}

@Test
@DisplayName("Should handle multiple documents without explicit IDs")
void shouldHandleMultipleDocumentsWithoutExplicitIds() throws IOException {
// Given
when(this.mockEmbeddingModel.embed(any(), any(), any()))
.thenReturn(List.of(new float[] { 0.1f, 0.2f, 0.3f }, new float[] { 0.4f, 0.5f, 0.6f }));

OpenSearchVectorStore vectorStore = createVectorStore(false);
List<Document> documents = List.of(new Document("doc1", "content1", Map.of()),
new Document("doc2", "content2", Map.of()));

// When
vectorStore.add(documents);

// Then
BulkRequest capturedRequest = captureBulkRequest();
assertThat(capturedRequest.operations()).hasSize(2);

for (var operation : capturedRequest.operations()) {
assertThat(operation.isIndex()).isTrue();
assertThat(operation.index().id()).isNull();
}
}

@Test
@DisplayName("Should handle embedding model error")
void shouldHandleEmbeddingModelError() {
// Given
when(this.mockEmbeddingModel.embed(any(), any(), any())).thenThrow(new RuntimeException("Embedding failed"));

OpenSearchVectorStore vectorStore = createVectorStore(true);
List<Document> documents = List.of(new Document("doc1", "content", Map.of()));

// When & Then
assertThatThrownBy(() -> vectorStore.add(documents)).isInstanceOf(RuntimeException.class)
.hasMessageContaining("Embedding failed");
}

// Helper methods

private OpenSearchVectorStore createVectorStore(boolean manageDocumentIds) {
return OpenSearchVectorStore.builder(this.mockOpenSearchClient, this.mockEmbeddingModel)
.manageDocumentIds(manageDocumentIds)
.build();
}

private BulkRequest captureBulkRequest() throws IOException {
ArgumentCaptor<BulkRequest> captor = ArgumentCaptor.forClass(BulkRequest.class);
verify(this.mockOpenSearchClient).bulk(captor.capture());
return captor.getValue();
}

private void verifyDocumentIdHandling(BulkRequest request, boolean shouldHaveExplicitIds) {
for (int i = 0; i < request.operations().size(); i++) {
var operation = request.operations().get(i);
assertThat(operation.isIndex()).isTrue();

if (shouldHaveExplicitIds) {
assertThat(operation.index().id()).isEqualTo("doc" + (i + 1));
}
else {
assertThat(operation.index().id()).isNull();
}
}
}

}