cloud-fan commented on a change in pull request #29585:
URL: https://github.com/apache/spark/pull/29585#discussion_r490097964
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/analysis/Analyzer.scala
##########
@@ -2575,13 +2580,13 @@ class Analyzer(
case ne: NamedExpression =>
// If a named expression is not in regularExpressions, add it to
// extractedExprBuffer and replace it with an AttributeReference.
+ val attr = ne.toAttribute
val missingExpr =
- AttributeSet(Seq(expr)) -- (regularExpressions ++
extractedExprBuffer)
+ AttributeSet(Seq(attr)) -- (regularExpressions ++
extractedExprBuffer)
if (missingExpr.nonEmpty) {
extractedExprBuffer += ne
}
- // alias will be cleaned in the rule CleanupAliases
- ne
+ attr
Review comment:
Can you give a simpler example to explain this change? It seems like if
the project list is `rank(a + b as c) over ...`, we should keep attribute `c`
in the window operator, instead of `a + b as c`.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
