Marko Radoicic created SPARK-60715:
--------------------------------------
Summary: Report aliased pipe AGGREGATE GROUP BY ordinals as syntax
errors
Key: SPARK-60715
URL: https://issues.apache.org/jira/browse/SPARK-60715
Project: Spark
Issue Type: Bug
Components: SQL
Affects Versions: 4.4.0
Reporter: Marko Radoicic
Follow-up to SPARK-60239. With spark.sql.groupByOrdinal=true (the default),
pipe AGGREGATE rejects an aliased integer-literal grouping ordinal, such as
{{TABLE t |> AGGREGATE sum(x) AS s GROUP BY 2 AS grp}}. SPARK-60239 reported
this with UNSUPPORTED_FEATURE.PIPE_OPERATOR_AGGREGATE_UNSUPPORTED_CASE
("aliases on GROUP BY ordinals"). Regular GROUP BY rejects any alias on a
grouping expression with PARSE_SYNTAX_ERROR.
This ticket changes the pipe AGGREGATE error to PARSE_SYNTAX_ERROR (SQLSTATE
42601), with the alias source text as the error parameter, matching regular
GROUP BY. Bare ordinals and aliases on other grouping expressions are
unchanged. The legacy config spark.sql.legacy.allowAliasedPipeAggregateOrdinals
still applies.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
