maropu commented on a change in pull request #23783: [SPARK-26854][SQL] Support
ANY/SOME subquery
URL: https://github.com/apache/spark/pull/23783#discussion_r260575796
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/expressions/predicates.scala
##########
@@ -145,11 +145,19 @@ case class Not(child: Expression)
override def sql: String = s"(NOT ${child.sql})"
}
-/**
- * Evaluates to `true` if `values` are returned in `query`'s result set.
- */
-case class InSubquery(values: Seq[Expression], query: ListQuery)
- extends Predicate with Unevaluable {
+abstract class PredicateSubquery extends Predicate with Unevaluable {
+
+ def values: Seq[Expression]
+ def query: ListQuery
+ def genCmp: (Expression, Expression) => Expression
+ def symbol: String
+
+ @transient protected lazy val cmpSymbol: String = genCmp(value, query) match
{
+ case b: BinaryComparison =>
+ b.symbol
+ case _ @ Not(EqualTo(_, _)) =>
Review comment:
For example, we can define `NotAnySubquery` for that special case and add
`<>` for semi-join in `RewritePredicateSubquery`? (I'm not 100% sure about
which is better, so it'd be better to wait for other developer's suggestions)
----------------------------------------------------------------
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]
