cloud-fan commented on a change in pull request #33183:
URL: https://github.com/apache/spark/pull/33183#discussion_r664062979
##########
File path:
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/NestedColumnAliasingSuite.scala
##########
@@ -738,6 +738,31 @@ class NestedColumnAliasingSuite extends SchemaPruningTest {
val optimized = Optimize.execute(query)
comparePlans(optimized, query)
}
+
+ test("SPARK-35972: When replace ExtractValue we should use semanticEquals") {
Review comment:
let's make the test name more general: `NestedColumnAliasing should
consider semantic equality`
##########
File path:
sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/NestedColumnAliasingSuite.scala
##########
@@ -738,6 +738,31 @@ class NestedColumnAliasingSuite extends SchemaPruningTest {
val optimized = Optimize.execute(query)
comparePlans(optimized, query)
}
+
+ test("SPARK-35972: NestedColumnAliasing should consider semantic equality") {
+ val dataType = new StructType()
+ .add(StructField("itemid", StringType))
+ .add(StructField("search_params", StructType(Seq(
+ StructField("col1", StringType),
+ StructField("col2", StringType)
+ ))))
+ val relation = LocalRelation('struct_data.struct(dataType))
+ val plan = relation
+ .repartition(100)
+ .select(
+ GetStructField('struct_data, 1, None).as("value"),
Review comment:
This is the key to trigger this bug. Ideally, in SQL query, nested
columns should result to `GetStructField` with non-None `name`. But there are
places that can create `GetStructField` with None `name`, such as
`UnresolvedStar.expand`, Dataset encoder stuff, etc.
It's better to have an end-to-end test, but I'm fine with this UT.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
