[
https://issues.apache.org/jira/browse/SPARK-60030?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Yang Jie updated SPARK-60030:
-----------------------------
Description:
DECLARE VARIABLE fails with an internal error when the default contains a
Python UDF:
DECLARE OR REPLACE VARIABLE v STRING DEFAULT py_udf(1)
[INTERNAL_ERROR] ... No plan for CreateVariable default(cast(cast(pythonUDF0#1
as int) as string), sql='py_udf(1)'), true
+- BatchEvalPython [py_udf(1)], [pythonUDF0#1]
+- ResolvedIdentifier ..., session.v
ExtractPythonUDFs treats the ResolvedIdentifier children of CreateVariable as
inputs: a Python UDF without references is a subset of every child's output, so
it wraps the child in a BatchEvalPython, and V2CommandStrategy no longer plans
the command. `nullif(py_udf(1), '3')` fails the same way. `SELECT py_udf(1)`
and `SET VAR v = py_udf(1)` work.
Raised in https://github.com/apache/spark/pull/59209#discussion_r4199845698
> DECLARE VARIABLE with a Python UDF in the default fails with No plan for
> CreateVariable
> ---------------------------------------------------------------------------------------
>
> Key: SPARK-60030
> URL: https://issues.apache.org/jira/browse/SPARK-60030
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 5.0.0
> Reporter: Yang Jie
> Priority: Major
>
> DECLARE VARIABLE fails with an internal error when the default contains a
> Python UDF:
> DECLARE OR REPLACE VARIABLE v STRING DEFAULT py_udf(1)
> [INTERNAL_ERROR] ... No plan for CreateVariable
> default(cast(cast(pythonUDF0#1 as int) as string), sql='py_udf(1)'), true
> +- BatchEvalPython [py_udf(1)], [pythonUDF0#1]
> +- ResolvedIdentifier ..., session.v
> ExtractPythonUDFs treats the ResolvedIdentifier children of CreateVariable as
> inputs: a Python UDF without references is a subset of every child's output,
> so it wraps the child in a BatchEvalPython, and V2CommandStrategy no longer
> plans the command. `nullif(py_udf(1), '3')` fails the same way. `SELECT
> py_udf(1)` and `SET VAR v = py_udf(1)` work.
> Raised in https://github.com/apache/spark/pull/59209#discussion_r4199845698
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
