fix

Author: zhengruifeng

python/pyspark/ml/tests/test_linalg.py

@@ -364,6 +364,9 @@ def test_unwrap_udt(self):
         ]
         self.assertEqual(results, expected)
 
+    def test_hashable(self):
+        _ = hash(VectorUDT())
+
 
 class MatrixUDTTests(MLlibTestCase):
     dm1 = DenseMatrix(3, 2, [0, 1, 4, 5, 9, 10])
@@ -394,6 +397,9 @@ def test_infer_schema(self):
             else:
                 raise ValueError("Expected a matrix but got type %r" % type(m))
 
+    def test_hashable(self):
+        _ = hash(MatrixUDT())
+
 
 if __name__ == "__main__":
     from pyspark.ml.tests.test_linalg import *  # noqa: F401

python/pyspark/sql/tests/test_types.py

@@ -2076,6 +2076,38 @@ def test_repr(self):
         for instance in instances:
             self.assertEqual(eval(repr(instance)), instance)
 
+    def test_hashable(self):
+        for dt in [
+            NullType(),
+            StringType(),
+            StringType("UTF8_BINARY"),
+            StringType("UTF8_LCASE"),
+            StringType("UNICODE"),
+            StringType("UNICODE_CI"),
+            CharType(10),
+            VarcharType(10),
+            BinaryType(),
+            BooleanType(),
+            DateType(),
+            TimeType(),
+            TimestampType(),
+            DecimalType(),
+            DoubleType(),
+            FloatType(),
+            ByteType(),
+            IntegerType(),
+            LongType(),
+            ShortType(),
+            CalendarIntervalType(),
+            ArrayType(StringType()),
+            MapType(StringType(), IntegerType()),
+            StructField("f1", StringType(), True),
+            StructType([StructField("f1", StringType(), True)]),
+            VariantType(),
+            ExamplePointUDT(),
+        ]:
+            _ = hash(dt)
+
     def test_daytime_interval_type_constructor(self):
         # SPARK-37277: Test constructors in day time interval.
         self.assertEqual(DayTimeIntervalType().simpleString(), "interval day to second")

python/pyspark/sql/types.py

@@ -1995,6 +1995,10 @@ def fromJson(cls, json: Dict[str, Any]) -> "UserDefinedType":
     def __eq__(self, other: Any) -> bool:
         return type(self) == type(other)
 
+    # __hash__ should be defined together with __eq__, otherwise it is not hashable
+    def __hash__(self) -> int:
+        return hash(str(self))
+
 
 class VariantVal:
     """