Handle case with nested list of objects (#477)

* Handle case with nested list of objects Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * fix validateEmbeddingsFieldValues Method Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * spotless formatting Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Onboard jenkins prod docker images on github actions (#483) Signed-off-by: Peter Zhu <zhujiaxi@amazon.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Update dependency org.json:json to v20231013 (#481) Co-authored-by: mend-for-github-com[bot] <50673670+mend-for-github-com[bot]@users.noreply.github.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * [Backport main manually][bug fix] Fix async actions are left in neural_sparse query (#438) (#479) * [bug fix] Fix async actions are left in neural_sparse query (#438) * add serialization and deserialization Signed-off-by: zhichao-aws <zhichaog@amazon.com> * hash, equals. + UT Signed-off-by: zhichao-aws <zhichaog@amazon.com> * tidy Signed-off-by: zhichao-aws <zhichaog@amazon.com> * add test Signed-off-by: zhichao-aws <zhichaog@amazon.com> --------- Signed-off-by: zhichao-aws <zhichaog@amazon.com> (cherry picked from commit 51e6c00) * rm max_token_score Signed-off-by: zhichao-aws <zhichaog@amazon.com> * add changelog Signed-off-by: zhichao-aws <zhichaog@amazon.com> * tidy Signed-off-by: zhichao-aws <zhichaog@amazon.com> --------- Signed-off-by: zhichao-aws <zhichaog@amazon.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Fixed exception for case when Hybrid query being wrapped into bool query (#490) * Adding null check for case when hybrid query wrapped into bool query Signed-off-by: Martin Gaievski <gaievski@amazon.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Fixed Hybrid query for cases when it's wrapped into other compound queries (#498) * Fixed nested field case Signed-off-by: Martin Gaievski <gaievski@amazon.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Added the github action to copy the attached issues label to PR. (#504) Signed-off-by: Navneet Verma <navneev@amazon.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Added support for jdk-21 (#500) * Added support for jdk-21 Signed-off-by: Martin Gaievski <gaievski@amazon.com> Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> * Add unit tests + small fixes Signed-off-by: krishy91 <crgkc.r@gmail.com> * fix indentation Signed-off-by: krishy91 <crgkc.r@gmail.com> * remove unused code + add 2nd level nesting test Signed-off-by: krishy91 <crgkc.r@gmail.com> * add integration test for list of nested objects Signed-off-by: krishy91 <crgkc.r@gmail.com> --------- Signed-off-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> Signed-off-by: Peter Zhu <zhujiaxi@amazon.com> Signed-off-by: zhichao-aws <zhichaog@amazon.com> Signed-off-by: Martin Gaievski <gaievski@amazon.com> Signed-off-by: Navneet Verma <navneev@amazon.com> Signed-off-by: krishy91 <crgkc.r@gmail.com> Co-authored-by: Gopala-Krishna.Char <gopala-krishna.char@intrafind.de> Co-authored-by: Peter Zhu <zhujiaxi@amazon.com> Co-authored-by: mend-for-github-com[bot] <50673670+mend-for-github-com[bot]@users.noreply.github.com> Co-authored-by: zhichao-aws <zhichaog@amazon.com> Co-authored-by: Martin Gaievski <gaievski@amazon.com> Co-authored-by: Navneet Verma <navneev@amazon.com> (cherry picked from commit ea49d3c)
opensearch-project · Mar 1, 2024 · 7bf5b91 · 7bf5b91
1 parent c422065
commit 7bf5b91
Show file tree

Hide file tree

Showing 5 changed files with 145 additions and 18 deletions.
diff --git a/src/main/java/org/opensearch/neuralsearch/processor/InferenceProcessor.java b/src/main/java/org/opensearch/neuralsearch/processor/InferenceProcessor.java
@@ -11,6 +11,7 @@
 import java.util.Objects;
 import java.util.function.BiConsumer;
 import java.util.function.Supplier;
+import java.util.stream.Collectors;
 import java.util.stream.IntStream;
 
 import org.apache.commons.lang3.StringUtils;
@@ -173,13 +174,28 @@ private void buildMapWithProcessorKeyAndOriginalValueForMapType(
         if (processorKey == null || sourceAndMetadataMap == null) return;
         if (processorKey instanceof Map) {
             Map<String, Object> next = new LinkedHashMap<>();
-            for (Map.Entry<String, Object> nestedFieldMapEntry : ((Map<String, Object>) processorKey).entrySet()) {
-                buildMapWithProcessorKeyAndOriginalValueForMapType(
-                    nestedFieldMapEntry.getKey(),
-                    nestedFieldMapEntry.getValue(),
-                    (Map<String, Object>) sourceAndMetadataMap.get(parentKey),
-                    next
-                );
+            if (sourceAndMetadataMap.get(parentKey) instanceof Map) {
+                for (Map.Entry<String, Object> nestedFieldMapEntry : ((Map<String, Object>) processorKey).entrySet()) {
+                    buildMapWithProcessorKeyAndOriginalValueForMapType(
+                        nestedFieldMapEntry.getKey(),
+                        nestedFieldMapEntry.getValue(),
+                        (Map<String, Object>) sourceAndMetadataMap.get(parentKey),
+                        next
+                    );
+                }
+            } else if (sourceAndMetadataMap.get(parentKey) instanceof List) {
+                for (Map.Entry<String, Object> nestedFieldMapEntry : ((Map<String, Object>) processorKey).entrySet()) {
+                    List<Map<String, Object>> list = (List<Map<String, Object>>) sourceAndMetadataMap.get(parentKey);
+                    List<Object> listOfStrings = list.stream().map(x -> x.get(nestedFieldMapEntry.getKey())).collect(Collectors.toList());
+                    Map<String, Object> map = new LinkedHashMap<>();
+                    map.put(nestedFieldMapEntry.getKey(), listOfStrings);
+                    buildMapWithProcessorKeyAndOriginalValueForMapType(
+                        nestedFieldMapEntry.getKey(),
+                        nestedFieldMapEntry.getValue(),
+                        map,
+                        next
+                    );
+                }
             }
             treeRes.put(parentKey, next);
         } else {
@@ -212,7 +228,7 @@ private void validateNestedTypeValue(String sourceKey, Object sourceValue, Suppl
         if (maxDepth > MapperService.INDEX_MAPPING_DEPTH_LIMIT_SETTING.get(environment.settings())) {
             throw new IllegalArgumentException("map type field [" + sourceKey + "] reached max depth limit, cannot process it");
         } else if ((List.class.isAssignableFrom(sourceValue.getClass()))) {
-            validateListTypeValue(sourceKey, sourceValue);
+            validateListTypeValue(sourceKey, sourceValue, maxDepthSupplier);
         } else if (Map.class.isAssignableFrom(sourceValue.getClass())) {
             ((Map) sourceValue).values()
                 .stream()
@@ -226,9 +242,11 @@ private void validateNestedTypeValue(String sourceKey, Object sourceValue, Suppl
     }
 
     @SuppressWarnings({ "rawtypes" })
-    private void validateListTypeValue(String sourceKey, Object sourceValue) {
+    private void validateListTypeValue(String sourceKey, Object sourceValue, Supplier<Integer> maxDepthSupplier) {
         for (Object value : (List) sourceValue) {
-            if (value == null) {
+            if (value instanceof Map) {
+                validateNestedTypeValue(sourceKey, value, () -> maxDepthSupplier.get() + 1);
+            } else if (value == null) {
                 throw new IllegalArgumentException("list type field [" + sourceKey + "] has null, cannot process it");
             } else if (!(value instanceof String)) {
                 throw new IllegalArgumentException("list type field [" + sourceKey + "] has non string value, cannot process it");
@@ -275,13 +293,20 @@ private void putNLPResultToSourceMapForMapType(
         if (processorKey == null || sourceAndMetadataMap == null || sourceValue == null) return;
         if (sourceValue instanceof Map) {
             for (Map.Entry<String, Object> inputNestedMapEntry : ((Map<String, Object>) sourceValue).entrySet()) {
-                putNLPResultToSourceMapForMapType(
-                    inputNestedMapEntry.getKey(),
-                    inputNestedMapEntry.getValue(),
-                    results,
-                    indexWrapper,
-                    (Map<String, Object>) sourceAndMetadataMap.get(processorKey)
-                );
+                if (sourceAndMetadataMap.get(processorKey) instanceof List) {
+                    // build nlp output for list of nested objects
+                    for (Map<String, Object> nestedElement : (List<Map<String, Object>>) sourceAndMetadataMap.get(processorKey)) {
+                        nestedElement.put(inputNestedMapEntry.getKey(), results.get(indexWrapper.index++));
+                    }
+                } else {
+                    putNLPResultToSourceMapForMapType(
+                        inputNestedMapEntry.getKey(),
+                        inputNestedMapEntry.getValue(),
+                        results,
+                        indexWrapper,
+                        (Map<String, Object>) sourceAndMetadataMap.get(processorKey)
+                    );
+                }
             }
         } else if (sourceValue instanceof String) {
             sourceAndMetadataMap.put(processorKey, results.get(indexWrapper.index++));

diff --git a/src/test/java/org/opensearch/neuralsearch/processor/TextEmbeddingProcessorIT.java b/src/test/java/org/opensearch/neuralsearch/processor/TextEmbeddingProcessorIT.java
@@ -70,7 +70,15 @@ private void ingestDocument() throws Exception {
             + "  \"favorites\": {\n"
             + "    \"game\": \"overwatch\",\n"
             + "    \"movie\": null\n"
-            + "  }\n"
+            + "  },\n"
+            + "  \"nested_passages\": [\n"
+            + "    {\n"
+            + "      \"text\": \"hello\"\n"
+            + "    },\n"
+            + "    {\n"
+            + "      \"text\": \"world\"\n"
+            + "    }\n"
+            + "  ]\n"
             + "}\n";
         Response response = makeRequest(
             client(),

diff --git a/src/test/java/org/opensearch/neuralsearch/processor/TextEmbeddingProcessorTests.java b/src/test/java/org/opensearch/neuralsearch/processor/TextEmbeddingProcessorTests.java
@@ -20,6 +20,7 @@
 import java.util.LinkedList;
 import java.util.List;
 import java.util.Map;
+import java.util.Arrays;
 import java.util.function.BiConsumer;
 import java.util.function.Supplier;
 
@@ -404,6 +405,35 @@ public void testBuildVectorOutput_withNestedMap_successful() {
         assertNotNull(actionGamesKnn);
     }
 
+    public void testBuildVectorOutput_withNestedList_successful() {
+        Map<String, Object> config = createNestedListConfiguration();
+        IngestDocument ingestDocument = createNestedListIngestDocument();
+        TextEmbeddingProcessor textEmbeddingProcessor = createInstanceWithNestedMapConfiguration(config);
+        Map<String, Object> knnMap = textEmbeddingProcessor.buildMapWithProcessorKeyAndOriginalValue(ingestDocument);
+        List<List<Float>> modelTensorList = createMockVectorResult();
+        textEmbeddingProcessor.buildNLPResult(knnMap, modelTensorList, ingestDocument.getSourceAndMetadata());
+        List<Map<String, Object>> nestedObj = (List<Map<String, Object>>) ingestDocument.getSourceAndMetadata().get("nestedField");
+        assertTrue(nestedObj.get(0).containsKey("vectorField"));
+        assertTrue(nestedObj.get(1).containsKey("vectorField"));
+        assertNotNull(nestedObj.get(0).get("vectorField"));
+        assertNotNull(nestedObj.get(1).get("vectorField"));
+    }
+
+    public void testBuildVectorOutput_withNestedList_Level2_successful() {
+        Map<String, Object> config = createNestedList2LevelConfiguration();
+        IngestDocument ingestDocument = create2LevelNestedListIngestDocument();
+        TextEmbeddingProcessor textEmbeddingProcessor = createInstanceWithNestedMapConfiguration(config);
+        Map<String, Object> knnMap = textEmbeddingProcessor.buildMapWithProcessorKeyAndOriginalValue(ingestDocument);
+        List<List<Float>> modelTensorList = createMockVectorResult();
+        textEmbeddingProcessor.buildNLPResult(knnMap, modelTensorList, ingestDocument.getSourceAndMetadata());
+        Map<String, Object> nestedLevel1 = (Map<String, Object>) ingestDocument.getSourceAndMetadata().get("nestedField");
+        List<Map<String, Object>> nestedObj = (List<Map<String, Object>>) nestedLevel1.get("nestedField");
+        assertTrue(nestedObj.get(0).containsKey("vectorField"));
+        assertTrue(nestedObj.get(1).containsKey("vectorField"));
+        assertNotNull(nestedObj.get(0).get("vectorField"));
+        assertNotNull(nestedObj.get(1).get("vectorField"));
+    }
+
     public void test_updateDocument_appendVectorFieldsToDocument_successful() {
         Map<String, Object> config = createPlainStringConfiguration();
         IngestDocument ingestDocument = createPlainIngestDocument();
@@ -520,4 +550,44 @@ private IngestDocument createNestedMapIngestDocument() {
         result.put("favorites", favorite);
         return new IngestDocument(result, new HashMap<>());
     }
+
+    private Map<String, Object> createNestedListConfiguration() {
+        Map<String, Object> nestedConfig = new HashMap<>();
+        nestedConfig.put("textField", "vectorField");
+        Map<String, Object> result = new HashMap<>();
+        result.put("nestedField", nestedConfig);
+        return result;
+    }
+
+    private Map<String, Object> createNestedList2LevelConfiguration() {
+        Map<String, Object> nestedConfig = new HashMap<>();
+        nestedConfig.put("textField", "vectorField");
+        Map<String, Object> nestConfigLevel1 = new HashMap<>();
+        nestConfigLevel1.put("nestedField", nestedConfig);
+        Map<String, Object> result = new HashMap<>();
+        result.put("nestedField", nestConfigLevel1);
+        return result;
+    }
+
+    private IngestDocument createNestedListIngestDocument() {
+        HashMap<String, Object> nestedObj1 = new HashMap<>();
+        nestedObj1.put("textField", "This is a text field");
+        HashMap<String, Object> nestedObj2 = new HashMap<>();
+        nestedObj2.put("textField", "This is another text field");
+        HashMap<String, Object> nestedList = new HashMap<>();
+        nestedList.put("nestedField", Arrays.asList(nestedObj1, nestedObj2));
+        return new IngestDocument(nestedList, new HashMap<>());
+    }
+
+    private IngestDocument create2LevelNestedListIngestDocument() {
+        HashMap<String, Object> nestedObj1 = new HashMap<>();
+        nestedObj1.put("textField", "This is a text field");
+        HashMap<String, Object> nestedObj2 = new HashMap<>();
+        nestedObj2.put("textField", "This is another text field");
+        HashMap<String, Object> nestedList = new HashMap<>();
+        nestedList.put("nestedField", Arrays.asList(nestedObj1, nestedObj2));
+        HashMap<String, Object> nestedList1 = new HashMap<>();
+        nestedList1.put("nestedField", nestedList);
+        return new IngestDocument(nestedList1, new HashMap<>());
+    }
 }
diff --git a/src/test/resources/processor/IndexMappings.json b/src/test/resources/processor/IndexMappings.json
@@ -83,6 +83,27 @@
       },
       "passage_text": {
         "type": "text"
+      },
+      "nested_passages": {
+        "type": "nested",
+        "properties": {
+          "text": {
+            "type": "text"
+          },
+          "embedding": {
+            "type": "knn_vector",
+            "dimension": 768,
+            "method": {
+              "name": "hnsw",
+              "space_type": "l2",
+              "engine": "lucene",
+              "parameters": {
+                "ef_construction": 128,
+                "m": 24
+              }
+            }
+          }
+        }
       }
     }
   }

diff --git a/src/test/resources/processor/PipelineConfiguration.json b/src/test/resources/processor/PipelineConfiguration.json
@@ -10,6 +10,9 @@
           "favorites": {
             "game": "game_knn",
             "movie": "movie_knn"
+          },
+          "nested_passages": {
+            "text": "embedding"
           }
         }
       }