|
| 1 | +/* |
| 2 | + * Licensed to the Apache Software Foundation (ASF) under one or more |
| 3 | + * contributor license agreements. See the NOTICE file distributed with |
| 4 | + * this work for additional information regarding copyright ownership. |
| 5 | + * The ASF licenses this file to You under the Apache License, Version 2.0 |
| 6 | + * (the "License"); you may not use this file except in compliance with |
| 7 | + * the License. You may obtain a copy of the License at |
| 8 | + * |
| 9 | + * http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | + * |
| 11 | + * Unless required by applicable law or agreed to in writing, software |
| 12 | + * distributed under the License is distributed on an "AS IS" BASIS, |
| 13 | + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 14 | + * See the License for the specific language governing permissions and |
| 15 | + * limitations under the License. |
| 16 | + */ |
| 17 | + |
| 18 | +package org.apache.spark.sql.catalyst.optimizer |
| 19 | + |
| 20 | +import org.apache.spark.sql.catalyst.dsl.expressions._ |
| 21 | +import org.apache.spark.sql.catalyst.dsl.plans._ |
| 22 | +import org.apache.spark.sql.catalyst.expressions.{Alias, ScalarSubquery} |
| 23 | +import org.apache.spark.sql.catalyst.plans._ |
| 24 | +import org.apache.spark.sql.catalyst.plans.logical.{DomainJoin, LocalRelation, LogicalPlan, OneRowRelation, Project} |
| 25 | +import org.apache.spark.sql.catalyst.rules.RuleExecutor |
| 26 | +import org.apache.spark.sql.internal.SQLConf |
| 27 | + |
| 28 | +class OptimizeOneRowRelationSubquerySuite extends PlanTest { |
| 29 | + |
| 30 | + private var optimizeOneRowRelationSubqueryEnabled: Boolean = _ |
| 31 | + |
| 32 | + protected override def beforeAll(): Unit = { |
| 33 | + super.beforeAll() |
| 34 | + optimizeOneRowRelationSubqueryEnabled = |
| 35 | + SQLConf.get.getConf(SQLConf.OPTIMIZE_ONE_ROW_RELATION_SUBQUERY) |
| 36 | + SQLConf.get.setConf(SQLConf.OPTIMIZE_ONE_ROW_RELATION_SUBQUERY, true) |
| 37 | + } |
| 38 | + |
| 39 | + protected override def afterAll(): Unit = { |
| 40 | + SQLConf.get.setConf(SQLConf.OPTIMIZE_ONE_ROW_RELATION_SUBQUERY, |
| 41 | + optimizeOneRowRelationSubqueryEnabled) |
| 42 | + super.afterAll() |
| 43 | + } |
| 44 | + |
| 45 | + object Optimize extends RuleExecutor[LogicalPlan] { |
| 46 | + val batches = |
| 47 | + Batch("Subquery", Once, |
| 48 | + OptimizeOneRowRelationSubquery, |
| 49 | + PullupCorrelatedPredicates) :: Nil |
| 50 | + } |
| 51 | + |
| 52 | + private def assertHasDomainJoin(plan: LogicalPlan): Unit = { |
| 53 | + assert(plan.collectWithSubqueries { case d: DomainJoin => d }.nonEmpty, |
| 54 | + s"Plan does not contain DomainJoin:\n$plan") |
| 55 | + } |
| 56 | + |
| 57 | + val t0 = OneRowRelation() |
| 58 | + val a = 'a.int |
| 59 | + val b = 'b.int |
| 60 | + val t1 = LocalRelation(a, b) |
| 61 | + val t2 = LocalRelation('c.int, 'd.int) |
| 62 | + |
| 63 | + test("Optimize scalar subquery with a single project") { |
| 64 | + // SELECT (SELECT a) FROM t1 |
| 65 | + val query = t1.select(ScalarSubquery(t0.select('a)).as("sub")) |
| 66 | + val optimized = Optimize.execute(query.analyze) |
| 67 | + val correctAnswer = t1.select('a.as("sub")) |
| 68 | + comparePlans(optimized, correctAnswer.analyze) |
| 69 | + } |
| 70 | + |
| 71 | + test("Optimize lateral subquery with a single project") { |
| 72 | + Seq(Inner, LeftOuter, Cross).foreach { joinType => |
| 73 | + // SELECT * FROM t1 JOIN LATERAL (SELECT a, b) |
| 74 | + val query = t1.lateralJoin(t0.select('a, 'b), joinType, None) |
| 75 | + val optimized = Optimize.execute(query.analyze) |
| 76 | + val correctAnswer = t1.select('a, 'b, 'a.as("a"), 'b.as("b")) |
| 77 | + comparePlans(optimized, correctAnswer.analyze) |
| 78 | + } |
| 79 | + } |
| 80 | + |
| 81 | + test("Optimize subquery with subquery alias") { |
| 82 | + val inner = t0.select('a).as("t2") |
| 83 | + val query = t1.select(ScalarSubquery(inner).as("sub")) |
| 84 | + val optimized = Optimize.execute(query.analyze) |
| 85 | + val correctAnswer = t1.select('a.as("sub")) |
| 86 | + comparePlans(optimized, correctAnswer.analyze) |
| 87 | + } |
| 88 | + |
| 89 | + test("Optimize scalar subquery with multiple projects") { |
| 90 | + // SELECT (SELECT a1 + b1 FROM (SELECT a AS a1, b AS b1)) FROM t1 |
| 91 | + val inner = t0.select('a.as("a1"), 'b.as("b1")).select(('a1 + 'b1).as("c")) |
| 92 | + val query = t1.select(ScalarSubquery(inner).as("sub")) |
| 93 | + val optimized = Optimize.execute(query.analyze) |
| 94 | + val correctAnswer = Project(Alias(Alias(a + b, "c")(), "sub")() :: Nil, t1) |
| 95 | + comparePlans(optimized, correctAnswer) |
| 96 | + } |
| 97 | + |
| 98 | + test("Optimize lateral subquery with multiple projects") { |
| 99 | + Seq(Inner, LeftOuter, Cross).foreach { joinType => |
| 100 | + val inner = t0.select('a.as("a1"), 'b.as("b1")) |
| 101 | + .select(('a1 + 'b1).as("c1"), ('a1 - 'b1).as("c2")) |
| 102 | + val query = t1.lateralJoin(inner, joinType, None) |
| 103 | + val optimized = Optimize.execute(query.analyze) |
| 104 | + val correctAnswer = t1.select('a, 'b, ('a + 'b).as("c1"), ('a - 'b).as("c2")) |
| 105 | + comparePlans(optimized, correctAnswer.analyze) |
| 106 | + } |
| 107 | + } |
| 108 | + |
| 109 | + test("Optimize subquery with nested correlated subqueries") { |
| 110 | + // SELECT (SELECT (SELECT b) FROM (SELECT a AS b)) FROM t1 |
| 111 | + val inner = t0.select('a.as("b")).select(ScalarSubquery(t0.select('b)).as("s")) |
| 112 | + val query = t1.select(ScalarSubquery(inner).as("sub")) |
| 113 | + val optimized = Optimize.execute(query.analyze) |
| 114 | + val correctAnswer = Project(Alias(Alias(a, "s")(), "sub")() :: Nil, t1) |
| 115 | + comparePlans(optimized, correctAnswer) |
| 116 | + } |
| 117 | + |
| 118 | + test("Batch should be idempotent") { |
| 119 | + // SELECT (SELECT 1 WHERE a = a + 1) FROM t1 |
| 120 | + val inner = t0.select(1).where('a === 'a + 1) |
| 121 | + val query = t1.select(ScalarSubquery(inner).as("sub")) |
| 122 | + val optimized = Optimize.execute(query.analyze) |
| 123 | + val doubleOptimized = Optimize.execute(optimized) |
| 124 | + comparePlans(optimized, doubleOptimized, checkAnalysis = false) |
| 125 | + } |
| 126 | + |
| 127 | + test("Should not optimize scalar subquery with operators other than project") { |
| 128 | + // SELECT (SELECT a AS a1 WHERE a = 1) FROM t1 |
| 129 | + val inner = t0.where('a === 1).select('a.as("a1")) |
| 130 | + val query = t1.select(ScalarSubquery(inner).as("sub")) |
| 131 | + val optimized = Optimize.execute(query.analyze) |
| 132 | + assertHasDomainJoin(optimized) |
| 133 | + } |
| 134 | + |
| 135 | + test("Should not optimize subquery with non-deterministic expressions") { |
| 136 | + // SELECT (SELECT r FROM (SELECT a + rand() AS r)) FROM t1 |
| 137 | + val inner = t0.select(('a + rand(0)).as("r")).select('r) |
| 138 | + val query = t1.select(ScalarSubquery(inner).as("sub")) |
| 139 | + val optimized = Optimize.execute(query.analyze) |
| 140 | + assertHasDomainJoin(optimized) |
| 141 | + } |
| 142 | + |
| 143 | + test("Should not optimize lateral join with non-empty join conditions") { |
| 144 | + Seq(Inner, LeftOuter).foreach { joinType => |
| 145 | + // SELECT * FROM t1 JOIN LATERAL (SELECT a AS a1, b AS b1) ON a = b1 |
| 146 | + val query = t1.lateralJoin(t0.select('a.as("a1"), 'b.as("b1")), joinType, Some('a === 'b1)) |
| 147 | + val optimized = Optimize.execute(query.analyze) |
| 148 | + assertHasDomainJoin(optimized) |
| 149 | + } |
| 150 | + } |
| 151 | + |
| 152 | + test("Should not optimize subquery with nested subqueries") { |
| 153 | + // SELECT (SELECT (SELECT a WHERE a = 1) FROM (SELECT a AS a)) FROM t1 |
| 154 | + val inner = t0.select('a).where('a === 1) |
| 155 | + val subquery = t0.select('a.as("a")) |
| 156 | + .select(ScalarSubquery(inner).as("s")).select('s + 1) |
| 157 | + val query = t1.select(ScalarSubquery(subquery).as("sub")) |
| 158 | + val optimized = Optimize.execute(query.analyze) |
| 159 | + assertHasDomainJoin(optimized) |
| 160 | + } |
| 161 | +} |
0 commit comments