JoshRosen commented on code in PR #48661: URL: https://github.com/apache/spark/pull/48661#discussion_r1821454712
########## sql/core/src/test/scala/org/apache/spark/sql/execution/InsertSortForLimitAndOffsetSuite.scala: ########## @@ -0,0 +1,120 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.spark.sql.execution + +import org.apache.spark.sql.QueryTest +import org.apache.spark.sql.execution.adaptive.AdaptiveSparkPlanHelper +import org.apache.spark.sql.internal.SQLConf +import org.apache.spark.sql.test.SharedSparkSession + +class InsertSortForLimitAndOffsetSuite extends QueryTest + with SharedSparkSession + with AdaptiveSparkPlanHelper { + import testImplicits._ + + private def assertHasTopKSort(plan: SparkPlan): Unit = { + assert(find(plan) { + case _: TakeOrderedAndProjectExec => true + case _ => false + }.isDefined) + } + + private def assertHasCollectLimitExec(plan: SparkPlan): Unit = { + assert(find(plan) { + case _: CollectLimitExec => true + case _ => false + }.isDefined) + } + + private def assertHasGlobalLimitExec(plan: SparkPlan): Unit = { + assert(find(plan) { + case _: GlobalLimitExec => true + case _ => false + }.isDefined) + } + + private def hasLocalSort(plan: SparkPlan): Boolean = { + find(plan) { + case GlobalLimitExec(_, s: SortExec, _) => !s.global + case _ => false + }.isDefined + } + + test("root LIMIT preserves data ordering with top-K sort") { + val df = spark.range(10).orderBy($"id" % 8).limit(2) + df.collect() + val physicalPlan = df.queryExecution.executedPlan + assertHasTopKSort(physicalPlan) + // Extra local sort is not needed for LIMIT with top-K sort optimization. + assert(!hasLocalSort(physicalPlan)) + } + + test("middle LIMIT preserves data ordering with top-K sort") { + val df = spark.range(10).orderBy($"id" % 8).limit(2).distinct() + df.collect() + val physicalPlan = df.queryExecution.executedPlan + assertHasTopKSort(physicalPlan) + // Extra local sort is not needed for LIMIT with top-K sort optimization. + assert(!hasLocalSort(physicalPlan)) Review Comment: For future readers: Conceptually, this isn't needed because the above plan uses TakeOrderedAndProjectExec.doExecute() which performs an unordered shuffle fetch but then does a local top-K rather than a simple limit: https://github.com/apache/spark/blob/dbd946986dab660d1a65c875f3b87f002c2e34d5/sql/core/src/main/scala/org/apache/spark/sql/execution/limit.scala#L374-L386 ########## sql/core/src/test/scala/org/apache/spark/sql/execution/InsertSortForLimitAndOffsetSuite.scala: ########## @@ -0,0 +1,120 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.spark.sql.execution + +import org.apache.spark.sql.QueryTest +import org.apache.spark.sql.execution.adaptive.AdaptiveSparkPlanHelper +import org.apache.spark.sql.internal.SQLConf +import org.apache.spark.sql.test.SharedSparkSession + +class InsertSortForLimitAndOffsetSuite extends QueryTest + with SharedSparkSession + with AdaptiveSparkPlanHelper { + import testImplicits._ + + private def assertHasTopKSort(plan: SparkPlan): Unit = { + assert(find(plan) { + case _: TakeOrderedAndProjectExec => true + case _ => false + }.isDefined) + } + + private def assertHasCollectLimitExec(plan: SparkPlan): Unit = { + assert(find(plan) { + case _: CollectLimitExec => true + case _ => false + }.isDefined) + } + + private def assertHasGlobalLimitExec(plan: SparkPlan): Unit = { + assert(find(plan) { + case _: GlobalLimitExec => true + case _ => false + }.isDefined) + } + + private def hasLocalSort(plan: SparkPlan): Boolean = { + find(plan) { + case GlobalLimitExec(_, s: SortExec, _) => !s.global + case _ => false + }.isDefined + } + + test("root LIMIT preserves data ordering with top-K sort") { + val df = spark.range(10).orderBy($"id" % 8).limit(2) + df.collect() + val physicalPlan = df.queryExecution.executedPlan + assertHasTopKSort(physicalPlan) + // Extra local sort is not needed for LIMIT with top-K sort optimization. + assert(!hasLocalSort(physicalPlan)) + } + + test("middle LIMIT preserves data ordering with top-K sort") { + val df = spark.range(10).orderBy($"id" % 8).limit(2).distinct() + df.collect() + val physicalPlan = df.queryExecution.executedPlan + assertHasTopKSort(physicalPlan) + // Extra local sort is not needed for LIMIT with top-K sort optimization. + assert(!hasLocalSort(physicalPlan)) Review Comment: For future readers: Conceptually, an extra local sort isn't needed in this case because the above plan uses TakeOrderedAndProjectExec.doExecute() which performs an unordered shuffle fetch but then does a local top-K rather than a simple limit: https://github.com/apache/spark/blob/dbd946986dab660d1a65c875f3b87f002c2e34d5/sql/core/src/main/scala/org/apache/spark/sql/execution/limit.scala#L374-L386 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
