Major Description
BoundAttribute issue will occur after optimization by BooleanSimplification and PushFoldableIntoBranches. How to reproduce this issue:
spark.sql("CREATE TABLE t1 (a INT, b INT) USING parquet") spark.sql("CREATE TABLE t2 (a INT, b INT) USING parquet") spark.sql( """ |SELECT cnt, | NOT ( buyer_id ) AS buyer_id2 |FROM (SELECT t1.a IS NOT NULL AS buyer_id, | Count(*) AS cnt | FROM t1 | INNER JOIN t2 | ON t1.a = t2.a | GROUP BY 1) t |""".stripMargin).collect()
Couldn't find a#4 in [CASE WHEN isnotnull(a#4) THEN 1 ELSE 2 END#10,count(1)#3L] java.lang.IllegalStateException: Couldn't find a#4 in [CASE WHEN isnotnull(a#4) THEN 1 ELSE 2 END#10,count(1)#3L] at org.apache.spark.sql.catalyst.expressions.BindReferences$$anonfun$bindReference$1.applyOrElse(BoundAttribute.scala:80) at org.apache.spark.sql.catalyst.expressions.BindReferences$$anonfun$bindReference$1.applyOrElse(BoundAttribute.scala:73) at org.apache.spark.sql.catalyst.trees.TreeNode.$anonfun$transformDown$1(TreeNode.scala:316) at org.apache.spark.sql.catalyst.trees.CurrentOrigin$.withOrigin(TreeNode.scala:72) at org.apache.spark.sql.catalyst.trees.TreeNode.transformDown(TreeNode.scala:316) at org.apache.spark.sql.catalyst.trees.TreeNode.$anonfun$transformDown$3(TreeNode.scala:321) at org.apache.spark.sql.catalyst.trees.TreeNode.$anonfun$mapChildren$1(TreeNode.scala:406) at org.apache.spark.sql.catalyst.trees.TreeNode.mapProductIterator(TreeNode.scala:242) at org.apache.spark.sql.catalyst.trees.TreeNode.mapChildren(TreeNode.scala:404) at org.apache.spark.sql.catalyst.trees.TreeNode.mapChildren(TreeNode.scala:357)
Another case:
spark.sql(
"""
|SELECT cnt,
| CASE WHEN ( buyer_id = 2 AND cnt > 3 ) THEN 2 ELSE 3 END AS buyer_id2
|FROM (SELECT CASE WHEN t1.a IS NOT NULL THEN 1 ELSE 2 END AS buyer_id, Count(*) AS cnt
| FROM t1 INNER JOIN t2 ON t1.a = t2.a
| GROUP BY 1) t
|""".stripMargin).collect()
