Add reset for vector

.github/workflows/links.yml

@@ -16,7 +16,7 @@ jobs:
         id: lychee
         uses: lycheeverse/lychee-action@master
         with:
-          args: --accept=200,403,429  "**/*.html" "**/*.md" "**/*.txt" "**/*.json"
+          args: --accept=200,403,429  **/*.html **/*.md **/*.txt **/*.json
         env:
           GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
       - name: Fail if there were link errors

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

@@ -432,6 +432,9 @@ protected void parseCreateField(ParseContext context, int dimension) throws IOEx
 
             vector.add(value);
             context.parser().nextToken();
+        } else if (token == XContentParser.Token.VALUE_NULL) {
+            context.path().remove();
+            return;
         }
 
         if (dimension != vector.size()) {

src/test/java/org/opensearch/knn/KNNRestTestCase.java

@@ -79,6 +79,8 @@
 public class KNNRestTestCase extends ODFERestTestCase {
     public static final String INDEX_NAME = "test_index";
     public static final String FIELD_NAME = "test_field";
+    private static final String DOCUMENT_FIELD_SOURCE = "_source";
+    private static final String DOCUMENT_FIELD_FOUND = "found";
 
     @AfterClass
     public static void dumpCoverage() throws IOException, MalformedObjectNameException {
@@ -486,6 +488,28 @@ protected void deleteKnnDoc(String index, String docId) throws IOException {
             RestStatus.fromCode(response.getStatusLine().getStatusCode()));
     }
 
+    /**
+     * Retrieve document by index and document id
+     */
+    protected Map<String, Object> getKnnDoc(final String index, final String docId) throws IOException {
+        final Request request = new Request(
+                "GET",
+                "/" + index + "/_doc/" + docId
+        );
+        final Response response = client().performRequest(request);
+
+        final Map<String, Object> responseMap =
+                createParser(XContentType.JSON.xContent(), EntityUtils.toString(response.getEntity())).map();
+
+        assertNotNull(responseMap);
+        assertTrue((Boolean) responseMap.get(DOCUMENT_FIELD_FOUND));
+        assertNotNull(responseMap.get(DOCUMENT_FIELD_SOURCE));
+
+        final Map<String, Object> docMap = (Map<String, Object>) responseMap.get(DOCUMENT_FIELD_SOURCE);
+
+        return docMap;
+    }
+
     /**
      * Utility to update  settings
      */

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

@@ -399,4 +399,53 @@ public void testExistsQuery() throws IOException {
 
         assertEquals(3, parseTotalSearchHits(EntityUtils.toString(response.getEntity())));
     }
+
+    public void testIndexingVectorValidation_updateVectorWithNull() throws Exception {
+        Settings settings = Settings.builder()
+                .put(getKNNDefaultIndexSettings())
+                .build();
+
+        createKnnIndex(INDEX_NAME, settings, createKnnIndexMapping(FIELD_NAME, 4));
+
+        // valid case with 4 dimension
+        final Float[] vectorForDocumentOne = {6.0f, 7.0f, 8.0f, 9.0f};
+        final String docOneId = "1";
+        addKnnDoc(INDEX_NAME, docOneId, FIELD_NAME, vectorForDocumentOne);
+
+        final Float[] vectorForDocumentTwo = {2.0f, 1.0f, 3.8f, 2.5f};
+        final String docTwoId = "2";
+        addKnnDoc(INDEX_NAME, docTwoId, FIELD_NAME, vectorForDocumentTwo);
+
+        //checking that one of two documents in retrievable based on closer similarity to query vector
+        int k = 1;
+        float[] queryVector = {5.0f, 6.0f, 7.0f, 10.0f};
+        final KNNQueryBuilder knnQueryBuilder = new KNNQueryBuilder(FIELD_NAME, queryVector, k);
+        final Response response = searchKNNIndex(INDEX_NAME, knnQueryBuilder, k);
+        final List<KNNResult> results = parseSearchResponse(EntityUtils.toString(response.getEntity()), FIELD_NAME);
+        assertEquals(1, results.size());
+        assertEquals(docOneId, results.get(0).getDocId());
+
+        //retrieving document by id
+        final Map<String, Object> knnDocMap = getKnnDoc(INDEX_NAME, docOneId);
+        assertNotNull(knnDocMap.get(FIELD_NAME));
+        final Float[] vectorInDocument = ((List<Double>) knnDocMap.get(FIELD_NAME)).stream()
+                .map(Double::floatValue).toArray(Float[]::new);
+        assertEquals(vectorForDocumentOne.length, vectorInDocument.length);
+        assertArrayEquals(vectorForDocumentOne, vectorInDocument);
+
+        // update vector value to null
+        updateKnnDoc(INDEX_NAME, docOneId, FIELD_NAME, null);
+
+        //retrieving updated document by id, vector should be null
+        final Map<String, Object> knnDocMapUpdated = getKnnDoc(INDEX_NAME, docOneId);
+        assertNull(knnDocMapUpdated.get(FIELD_NAME));
+
+        //checking that first document one is no longer discoverable by knn plugin
+        //expected result is the document two which had originally less similarity score
+        final Response updatedResponse = searchKNNIndex(INDEX_NAME, knnQueryBuilder, k);
+        final List<KNNResult> updatedResults =
+                parseSearchResponse(EntityUtils.toString(updatedResponse.getEntity()), FIELD_NAME);
+        assertEquals(1, updatedResults.size());
+        assertEquals(docTwoId, updatedResults.get(0).getDocId());
+    }
 }