fix

zml1206 · Jan 14, 2025 · 23f0b46 · 23f0b46
1 parent 20c31db
commit 23f0b46
Show file tree

Hide file tree

Showing 4 changed files with 24 additions and 10 deletions.
diff --git a/...talyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteWithExpression.scala b/...talyst/src/main/scala/org/apache/spark/sql/catalyst/optimizer/RewriteWithExpression.scala
@@ -162,7 +162,7 @@ object RewriteWithExpression extends Rule[LogicalPlan] {
               } else {
                 val alias = originalAttr match {
                   case Some(a) =>
-                    Alias(child, a.name)(a.exprId)
+                    Alias(child, a.name)(a.exprId, a.qualifier, Option(a.metadata))
                   case _ =>
                     val aliasName = if (SQLConf.get.getConf(SQLConf.USE_COMMON_EXPR_ID_FOR_ALIAS)) {
                       s"_common_expr_${id.id}"

diff --git a/...catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/FilterPushdownSuite.scala b/...catalyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/FilterPushdownSuite.scala
@@ -148,6 +148,23 @@ class FilterPushdownSuite extends PlanTest {
     comparePlans(optimized, originalQuery)
   }
 
+  test("can't push without rewrite") {
+    val originalQuery =
+      testRelation
+        .select($"a" + $"b" as "e")
+        .where($"e" === 1)
+        .analyze
+
+    val optimized = Optimize.execute(originalQuery.analyze)
+    val correctAnswer =
+      testRelation
+        .where($"a" + $"b" === 1)
+        .select($"a" + $"b" as "e")
+        .analyze
+
+    comparePlans(optimized, correctAnswer)
+  }
+
   test("nondeterministic: can always push down filter through project with deterministic field") {
     val originalQuery = testRelation
       .select($"a")

diff --git a/...talyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizerLoggingSuite.scala b/...talyst/src/test/scala/org/apache/spark/sql/catalyst/optimizer/OptimizerLoggingSuite.scala
@@ -32,7 +32,7 @@ class OptimizerLoggingSuite extends PlanTest {
   object Optimize extends RuleExecutor[LogicalPlan] {
     val batches =
       Batch("Optimizer Batch", FixedPoint(100),
-        PushPredicateThroughNonJoin) ::
+        PushPredicateThroughNonJoin, ColumnPruning, CollapseProject) ::
       Batch("Batch Has No Effect", Once,
         ColumnPruning) :: Nil
   }
@@ -60,9 +60,9 @@ class OptimizerLoggingSuite extends PlanTest {
     (ruleOrBatch => logMessages.exists(_.contains(ruleOrBatch))))
     assert(events.forall(_.getLevel == expectedLevel))
     val expectedMetrics = Seq(
-      "Total number of runs: 7",
+      "Total number of runs: 15",
       "Total time:",
-      "Total number of effective runs: 3",
+      "Total number of effective runs: 2",
       "Total time of effective runs:")
     assert(expectedMetrics.forall(metrics => logMessages.exists(_.contains(metrics))))
   }
@@ -113,17 +113,15 @@ class OptimizerLoggingSuite extends PlanTest {
         ColumnPruning.ruleName,
         CollapseProject.ruleName).reduce(_ + "," + _) ->
         Seq(PushPredicateThroughNonJoin.ruleName,
-          ColumnPruning.ruleName,
           CollapseProject.ruleName),
       Seq(PushPredicateThroughNonJoin.ruleName,
         ColumnPruning.ruleName).reduce(_ + "," + _) ->
-        Seq(PushPredicateThroughNonJoin.ruleName,
-          ColumnPruning.ruleName),
+        Seq(PushPredicateThroughNonJoin.ruleName),
       CollapseProject.ruleName ->
         Seq(CollapseProject.ruleName),
       Seq(ColumnPruning.ruleName,
         "DummyRule").reduce(_ + "," + _) ->
-        Seq(ColumnPruning.ruleName),
+        Seq(),
       "DummyRule" -> Seq(),
       "" -> Seq()
     )

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/OptimizeMetadataOnlyQuery.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/OptimizeMetadataOnlyQuery.scala
@@ -50,7 +50,7 @@ case class OptimizeMetadataOnlyQuery(catalog: SessionCatalog) extends Rule[Logic
       return plan
     }
 
-    val p = plan.transform {
+    plan.transform {
       case a @ Aggregate(_, aggExprs, child @ PhysicalOperation(
           projectList, filters, PartitionedRelation(partAttrs, rel)), _) =>
         // We only apply this optimization when only partitioned attributes are scanned.
@@ -86,7 +86,6 @@ case class OptimizeMetadataOnlyQuery(catalog: SessionCatalog) extends Rule[Logic
           a
         }
     }
-    p
   }
 
   /**