maryannxue commented on a change in pull request #23712: [SPARK-26798][SQL]
HandleNullInputsForUDF should trust nullability
URL: https://github.com/apache/spark/pull/23712#discussion_r255804674
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/ScalaUDF.scala
##########
@@ -31,9 +31,9 @@ import org.apache.spark.sql.types.{AbstractDataType,
DataType}
* null. Use boxed type or [[Option]] if you wanna do the
null-handling yourself.
* @param dataType Return type of function.
* @param children The input expressions of this UDF.
- * @param inputsNullSafe Whether the inputs are of non-primitive types or not
nullable. Null values
- * of Scala primitive types will be converted to the
type's default value and
- * lead to wrong results, thus need special handling
before calling the UDF.
+ * @param inputPrimitives Whether the inputs are of primitive types. Null
values of Java primitive
Review comment:
=> The analyzer should be aware of Scala primitive types so as to make the
UDF return null if there is any null input value of these types. On the other
hand, Java UDFs can only have boxed types, thus this parameter will always be
all false.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
