feat(mol): add MOL type/function and PATTERN index

sdk-core/src/main/java/io/milvus/common/clientenum/FunctionType.java

@@ -23,7 +23,8 @@ public enum FunctionType {
     UNKNOWN("Unknown", 0), // in milvus-proto, the name is "Unknown"
     BM25("BM25", 1), // Added missing name parameter
     TEXTEMBEDDING("TextEmbedding", 2), // in milvus-proto, the name is "TextEmbedding"
-    RERANK("RERANK", 3); // Added missing name parameter
+    RERANK("RERANK", 3), // Added missing name parameter
+    MOL_FINGERPRINT("MolFingerprint", 5);
 
     private final String name;
     private final int code;

sdk-core/src/main/java/io/milvus/param/IndexType.java

@@ -50,6 +50,10 @@ public enum IndexType {
 
     // Only for varchar type field
     TRIE("Trie", 100),
+    // Only for geometry type field
+    RTREE(120),
+    // Only for mol type field
+    PATTERN(121),
     // Only for scalar type field
     STL_SORT(200), // only for numeric type field
     INVERTED(201), // works for all scalar fields except JSON type field

sdk-core/src/main/java/io/milvus/param/ParamUtils.java

@@ -270,6 +270,7 @@ public static void checkFieldData(FieldType fieldSchema, List<?> values, boolean
             case String:
             case Geometry:
             case Timestamptz:
+            case Mol:
                 for (Object value : values) {
                     if (checkNullableFieldData(fieldSchema, value, verifyElementType)) {
                         continue;
@@ -423,6 +424,7 @@ public static Object checkFieldValue(String fieldName, DataType dataType, DataTy
             case String:
             case Geometry:
             case Timestamptz:
+            case Mol:
                 if (!(value.isJsonPrimitive())) {
                     throw new ParamException(String.format(errMsgs.get(dataType), fieldName));
                 }
@@ -1414,6 +1416,11 @@ public static ScalarField genScalarField(DataType dataType, DataType elementType
                 GeometryWktArray wktArray = GeometryWktArray.newBuilder().addAllData(strings).build();
                 return ScalarField.newBuilder().setGeometryWktData(wktArray).build();
             }
+            case Mol: {
+                List<String> strings = objects.stream().map(p -> (p == null) ? null : (String) p).collect(Collectors.toList());
+                MolSmilesArray smilesArray = MolSmilesArray.newBuilder().addAllData(strings).build();
+                return ScalarField.newBuilder().setMolSmilesData(smilesArray).build();
+            }
             case JSON: {
                 List<ByteString> byteStrings = objects.stream().map(p -> (p == null) ? null : ByteString.copyFromUtf8(p.toString()))
                         .collect(Collectors.toList());
@@ -1555,6 +1562,7 @@ public static ValueField objectToValueField(Object obj, DataType dataType) {
             case String:
             case Geometry:
             case Timestamptz:
+            case Mol:
                 if (obj instanceof String) {
                     return builder.setStringData((String) obj).build();
                 }
@@ -1593,6 +1601,7 @@ public static Object valueFieldToObject(ValueField value, DataType dataType) {
             case String:
             case Geometry:
             case Timestamptz:
+            case Mol:
                 return value.getStringData();
             case JSON:
                 return JsonUtils.fromJson(value.getStringData(), JsonObject.class);

sdk-core/src/main/java/io/milvus/response/FieldDataWrapper.java

@@ -180,6 +180,8 @@ public long getRowCount() throws IllegalResponseException {
                 return fieldData.getScalars().getStringData().getDataCount();
             case Geometry:
                 return fieldData.getScalars().getGeometryWktData().getDataCount();
+            case Mol:
+                return fieldData.getScalars().getMolSmilesData().getDataCount();
             case JSON:
                 return fieldData.getScalars().getJsonData().getDataCount();
             case Array:
@@ -254,6 +256,7 @@ private List<?> getFieldDataInternal() throws IllegalResponseException {
             case String:
             case Geometry:
             case Timestamptz:
+            case Mol:
             case JSON:
                 return getScalarData(dt, fieldData.getScalars(), fieldData.getValidDataList());
             case ArrayOfStruct:
@@ -357,6 +360,10 @@ private List<?> getScalarData(DataType dt, ScalarField scalar, List<Boolean> val
                 ProtocolStringList protoGeoList = scalar.getGeometryWktData().getDataList();
                 return setNoneData(protoGeoList.subList(0, protoGeoList.size()), validData);
             }
+            case Mol: {
+                ProtocolStringList protoMolList = scalar.getMolSmilesData().getDataList();
+                return setNoneData(protoMolList.subList(0, protoMolList.size()), validData);
+            }
             case JSON: {
                 List<ByteString> dataList = scalar.getJsonData().getDataList();
                 return dataList.stream().map(ByteString::toStringUtf8).collect(Collectors.toList());

sdk-core/src/main/java/io/milvus/v2/common/DataType.java

@@ -38,6 +38,7 @@ public enum DataType {
     JSON(23),
     Geometry(24),
     Timestamptz(26),
+    Mol(27),
 
     BinaryVector(100),
     FloatVector(101),

sdk-core/src/main/java/io/milvus/v2/common/IndexParam.java

@@ -203,6 +203,8 @@ public enum IndexType {
 
         // Only for geometry type field
         RTREE(120),
+        // Only for mol type field
+        PATTERN(121),
 
         // Only for scalar type field
         STL_SORT(200), // only for numeric type field

sdk-core/src/main/java/io/milvus/v2/service/vector/request/InsertReq.java

@@ -30,7 +30,7 @@ public class InsertReq {
      * Sets the row data to insert. The rows list cannot be empty.
      * <p>
      * Internal class for insert data.
-     * If dataType is Bool/Int8/Int16/Int32/Int64/Float/Double/Varchar/Geometry/Timestamptz, use JsonObject.addProperty(key, value) to input;
+     * If dataType is Bool/Int8/Int16/Int32/Int64/Float/Double/Varchar/Geometry/Timestamptz/Mol, use JsonObject.addProperty(key, value) to input;
      * If dataType is FloatVector, use JsonObject.add(key, gson.toJsonTree(List[Float]) to input;
      * If dataType is BinaryVector/Float16Vector/BFloat16Vector/Int8Vector, use JsonObject.add(key, gson.toJsonTree(byte[])) to input;
      * If dataType is SparseFloatVector, use JsonObject.add(key, gson.toJsonTree(SortedMap[Long, Float])) to input;

sdk-core/src/main/java/io/milvus/v2/service/vector/request/UpsertReq.java

@@ -28,7 +28,7 @@ public class UpsertReq {
      * Sets the row data to insert. The rows list cannot be empty.
      * <p>
      * Internal class for insert data.
-     * If dataType is Bool/Int8/Int16/Int32/Int64/Float/Double/Varchar/Geometry/Timestamptz, use JsonObject.addProperty(key, value) to input;
+     * If dataType is Bool/Int8/Int16/Int32/Int64/Float/Double/Varchar/Geometry/Timestamptz/Mol, use JsonObject.addProperty(key, value) to input;
      * If dataType is FloatVector, use JsonObject.add(key, gson.toJsonTree(List[Float]) to input;
      * If dataType is BinaryVector/Float16Vector/BFloat16Vector, use JsonObject.add(key, gson.toJsonTree(byte[])) to input;
      * If dataType is SparseFloatVector, use JsonObject.add(key, gson.toJsonTree(SortedMap[Long, Float])) to input;

sdk-core/src/test/java/io/milvus/client/MilvusServiceClientTest.java

@@ -2866,6 +2866,13 @@ void testFieldDataWrapper() {
         }
         testScalarField(ScalarField.newBuilder().setStringData(strBuilder).build(),
                 DataType.VarChar, dim);
+
+        MolSmilesArray.Builder molBuilder = MolSmilesArray.newBuilder();
+        for (long i = 0; i < dim; ++i) {
+            molBuilder.addData("C" + i);
+        }
+        testScalarField(ScalarField.newBuilder().setMolSmilesData(molBuilder).build(),
+                DataType.Mol, dim);
     }
 
     @Test