Can one of the admins verify this patch?

I think the same problem is solved in: #10943

@hvanhovell and @jayadevanmurali We checked the source code change in the referenced PR@10943. That PR is resolving the similar issue in the drop scenario. But this current fix would be for the explicit select statements as stated in our bug.Also we double checked whether the fix @PR-10943 would resolve the select issue . But, it does not resolve this bug. So ,please let us know your comments.

@thomastechs

We should use df.select("a.c") to select a column with name "a.c".
The reason is that we can use df.select("path.to.column") to select a nested column, for example:

scala> case class A(inner: Int)
scala> val df = Seq((A(1), 2)).toDF("a", "b")
scala> df.select("a.inner").show()
+-----+
|inner|
+-----+
|    1|
+-----+

So, I think this is NOT a bug, current behavior is expected.

+1 for not a problem.