diff options
-rw-r--r-- | python/pyspark/sql/functions.py | 5 | ||||
-rw-r--r-- | python/pyspark/sql/tests.py | 7 |
2 files changed, 12 insertions, 0 deletions
diff --git a/python/pyspark/sql/functions.py b/python/pyspark/sql/functions.py index 5213a3c358..4f4ae10892 100644 --- a/python/pyspark/sql/functions.py +++ b/python/pyspark/sql/functions.py @@ -1864,6 +1864,11 @@ class UserDefinedFunction(object): .. versionadded:: 1.3 """ def __init__(self, func, returnType, name=None): + if not callable(func): + raise TypeError( + "Not a function or callable (__call__ is not defined): " + "{0}".format(type(func))) + self.func = func self.returnType = ( returnType if isinstance(returnType, DataType) diff --git a/python/pyspark/sql/tests.py b/python/pyspark/sql/tests.py index d9d03337ff..73721674f6 100644 --- a/python/pyspark/sql/tests.py +++ b/python/pyspark/sql/tests.py @@ -504,6 +504,13 @@ class SQLTests(ReusedPySparkTestCase): self.assertTupleEqual(expected, actual) + def test_udf_shouldnt_accept_noncallable_object(self): + from pyspark.sql.functions import UserDefinedFunction + from pyspark.sql.types import StringType + + non_callable = None + self.assertRaises(TypeError, UserDefinedFunction, non_callable, StringType()) + def test_basic_functions(self): rdd = self.sc.parallelize(['{"foo":"bar"}', '{"foo":"baz"}']) df = self.spark.read.json(rdd) |