Github user viirya commented on a diff in the pull request:
https://github.com/apache/spark/pull/20276#discussion_r162018508
--- Diff: sql/core/src/main/scala/org/apache/spark/sql/Dataset.scala ---
@@ -1234,11 +1234,24 @@ class Dataset[T] private[sql](
if (sqlContext.conf.supportQuotedRegexColumnName) {
colRegex(colName)
} else {
- val expr = resolve(colName)
- Column(expr)
+ createCol(colName)
}
}
+ private def createCol(name: String): Column = {
+ val expr = resolve(name) transform {
+ case a: AttributeReference =>
+ // Associate the returned `AttributeReference` with the
`AnalysisBarrier` of this Dataset,
+ // by putting the barrier id into `AttributeReference.metadata`.
This information is only
+ // used to disambiguate the attributes in join condition when
resolving self-join and
+ // de-duplicating the right side plan.
--- End diff --
Shall we clarify that this metadata will be removed after analysis?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
