[SPARK-17244] Catalyst should not pushdown non-deterministic join conditions

## What changes were proposed in this pull request?

Given that non-deterministic expressions can be stateful, pushing them down the query plan during the optimization phase can cause incorrect behavior. This patch fixes that issue by explicitly disabling that.

## How was this patch tested?

A new test in `FilterPushdownSuite` that checks catalyst behavior for both deterministic and non-deterministic join conditions.

Author: Sameer Agarwal <sameerag@cs.berkeley.edu>

Closes #14815 from sameeragarwal/constraint-inputfile.

1 files changed, 14 insertions, 7 deletions

diff --git a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
index 82ad0fb5ee..5c8316189d 100644
--- a/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
+++ b/sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/Optimizer.scala
@@ -1379,18 +1379,25 @@ object EliminateOuterJoin extends Rule[LogicalPlan] with PredicateHelper {
  */
 object PushPredicateThroughJoin extends Rule[LogicalPlan] with PredicateHelper {
   /**
-   * Splits join condition expressions into three categories based on the attributes required
-   * to evaluate them.
+   * Splits join condition expressions or filter predicates (on a given join's output) into three
+   * categories based on the attributes required to evaluate them. Note that we explicitly exclude
+   * on-deterministic (i.e., stateful) condition expressions in canEvaluateInLeft or
+   * canEvaluateInRight to prevent pushing these predicates on either side of the join.
    *
    * @return (canEvaluateInLeft, canEvaluateInRight, haveToEvaluateInBoth)
    */
   private def split(condition: Seq[Expression], left: LogicalPlan, right: LogicalPlan) = {
+    // Note: In order to ensure correctness, it's important to not change the relative ordering of
+    // any deterministic expression that follows a non-deterministic expression. To achieve this,
+    // we only consider pushing down those expressions that precede the first non-deterministic
+    // expression in the condition.
+    val (pushDownCandidates, containingNonDeterministic) = condition.span(_.deterministic)
     val (leftEvaluateCondition, rest) =
-        condition.partition(_.references subsetOf left.outputSet)
+      pushDownCandidates.partition(_.references.subsetOf(left.outputSet))
     val (rightEvaluateCondition, commonCondition) =
-        rest.partition(_.references subsetOf right.outputSet)
+        rest.partition(expr => expr.references.subsetOf(right.outputSet))
 
-    (leftEvaluateCondition, rightEvaluateCondition, commonCondition)
+    (leftEvaluateCondition, rightEvaluateCondition, commonCondition ++ containingNonDeterministic)
   }
 
   def apply(plan: LogicalPlan): LogicalPlan = plan transform {
@@ -1441,7 +1448,7 @@ object PushPredicateThroughJoin extends Rule[LogicalPlan] with PredicateHelper {
       }
 
     // push down the join filter into sub query scanning if applicable
-    case f @ Join(left, right, joinType, joinCondition) =>
+    case j @ Join(left, right, joinType, joinCondition) =>
       val (leftJoinConditions, rightJoinConditions, commonJoinCondition) =
         split(joinCondition.map(splitConjunctivePredicates).getOrElse(Nil), left, right)
 
@@ -1471,7 +1478,7 @@ object PushPredicateThroughJoin extends Rule[LogicalPlan] with PredicateHelper {
           val newJoinCond = (leftJoinConditions ++ commonJoinCondition).reduceLeftOption(And)
 
           Join(newLeft, newRight, LeftOuter, newJoinCond)
-        case FullOuter => f
+        case FullOuter => j
         case NaturalJoin(_) => sys.error("Untransformed NaturalJoin node")
         case UsingJoin(_, _) => sys.error("Untransformed Using join node")
       }