[CALCITE-6650] Optimize the IN sub-query and SOME sub-query by Metada…

…ta RowCount
chucheng92 · Nov 11, 2024 · f347fba · f347fba
1 parent edd6d11
commit f347fba
Show file tree

Hide file tree

Showing 4 changed files with 368 additions and 3 deletions.
diff --git a/core/src/main/java/org/apache/calcite/rel/rules/SubQueryRemoveRule.java b/core/src/main/java/org/apache/calcite/rel/rules/SubQueryRemoveRule.java
@@ -170,6 +170,13 @@ private static RexNode rewriteCollection(RexSubQuery e,
    */
   private static RexNode rewriteSome(RexSubQuery e, Set<CorrelationId> variablesSet,
       RelBuilder builder, int subQueryIndex) {
+    // If the sub-query is guaranteed to return 0 row, just return
+    // FALSE.
+    final RelMetadataQuery mq = e.rel.getCluster().getMetadataQuery();
+    final Double maxRowCount = mq.getMaxRowCount(e.rel);
+    if (maxRowCount != null && maxRowCount <= 0D) {
+      return builder.getRexBuilder().makeLiteral(Boolean.FALSE, e.getType(), true);
+    }
     // Most general case, where the left and right keys might have nulls, and
     // caller requires 3-valued logic return.
     //
@@ -459,11 +466,11 @@ private static RexNode rewriteExists(RexSubQuery e, Set<CorrelationId> variables
     // TRUE.
     final RelMetadataQuery mq = e.rel.getCluster().getMetadataQuery();
     final Double minRowCount = mq.getMinRowCount(e.rel);
-    if (minRowCount != null && minRowCount >= 1D) {
+    if (minRowCount != null && minRowCount > 0D) {
       return builder.literal(true);
     }
     final Double maxRowCount = mq.getMaxRowCount(e.rel);
-    if (maxRowCount != null && maxRowCount < 1D) {
+    if (maxRowCount != null && maxRowCount <= 0D) {
       return builder.literal(false);
     }
     builder.push(e.rel);
@@ -555,6 +562,13 @@ private static RexNode rewriteUnique(RexSubQuery e, RelBuilder builder) {
    */
   private static RexNode rewriteIn(RexSubQuery e, Set<CorrelationId> variablesSet,
       RelOptUtil.Logic logic, RelBuilder builder, int offset, int subQueryIndex) {
+    // If the sub-query is guaranteed to return 0 row, just return
+    // FALSE.
+    final RelMetadataQuery mq = e.rel.getCluster().getMetadataQuery();
+    final Double maxRowCount = mq.getMaxRowCount(e.rel);
+    if (maxRowCount != null && maxRowCount <= 0D) {
+      return builder.getRexBuilder().makeLiteral(Boolean.FALSE, e.getType(), true);
+    }
     // Most general case, where the left and right keys might have nulls, and
     // caller requires 3-valued logic return.
     //

diff --git a/core/src/test/java/org/apache/calcite/test/RelOptRulesTest.java b/core/src/test/java/org/apache/calcite/test/RelOptRulesTest.java
@@ -5049,6 +5049,52 @@ private void checkEmptyJoin(RelOptFixture f) {
     sql(sql).withSubQueryRules().check();
   }
 
+  /** Test case for
+   * <a href="https://issues.apache.org/jira/browse/CALCITE-6650">[CALCITE-6650]
+   * Optimize the IN sub-query and SOME sub-query
+   * by Metadata RowCount</a>. */
+  @Test void testInWithNoRowSubQuery() {
+    final String sql = "select * from dept as d\n"
+        + "where deptno in (\n"
+        + "  select deptno from emp e where false)";
+    sql(sql).withSubQueryRules().check();
+  }
+
+  @Test void testNotInWithNoRowSubQuery() {
+    final String sql = "select * from dept as d\n"
+        + "where deptno not in (\n"
+        + "  select deptno from emp e where false)";
+    sql(sql).withSubQueryRules().check();
+  }
+
+  @Test void testSomeWithGreaterThanNoRowSubQuery() {
+    final String sql = "select * from dept as d\n"
+        + "where deptno > some(\n"
+        + "  select deptno from emp e where false)";
+    sql(sql).withSubQueryRules().check();
+  }
+
+  @Test void testSomeWithLessThanOrEqualNoRowSubQuery() {
+    final String sql = "select * from dept as d\n"
+        + "where deptno <= some(\n"
+        + "  select deptno from emp e where false)";
+    sql(sql).withSubQueryRules().check();
+  }
+
+  @Test void testUniqueWithNoRowSubQuery() {
+    final String sql = "select * from dept as d\n"
+        + "where unique(\n"
+        + "  select deptno from emp e where false)";
+    sql(sql).withSubQueryRules().check();
+  }
+
+  @Test void testNotUniqueWithNoRowSubQuery() {
+    final String sql = "select * from dept as d\n"
+        + "where not unique(\n"
+        + "  select deptno from emp e where false)";
+    sql(sql).withSubQueryRules().check();
+  }
+
   /** Test case for
    * <a href="https://issues.apache.org/jira/browse/CALCITE-4848">[CALCITE-4848]
    * Adding a HAVING condition to a query with a dynamic parameter makes the result always empty

diff --git a/core/src/test/resources/org/apache/calcite/test/RelOptRulesTest.xml b/core/src/test/resources/org/apache/calcite/test/RelOptRulesTest.xml
@@ -5081,6 +5081,30 @@ LogicalUnion(all=[true])
     LogicalTableScan(table=[[CATALOG, SALES, EMP]])
   LogicalProject(EXPR$0=[LOWER($1)])
     LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testInWithNoRowSubQuery">
+    <Resource name="sql">
+      <![CDATA[select * from dept as d
+where deptno in (
+  select deptno from emp e where false)]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalFilter(condition=[IN($0, {
+LogicalProject(DEPTNO=[$7])
+  LogicalFilter(condition=[false])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+})])
+    LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+    <Resource name="planAfter">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalValues(tuples=[[]])
 ]]>
     </Resource>
   </TestCase>
@@ -7724,6 +7748,30 @@ LogicalProject(USER=[USER])
 LogicalAggregate(group=[{0}], EXPR$1=[SUM($1)])
   LogicalProject(NAME=[$1], DEPTNO=[$0])
     LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testNotInWithNoRowSubQuery">
+    <Resource name="sql">
+      <![CDATA[select * from dept as d
+where deptno not in (
+  select deptno from emp e where false)]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalFilter(condition=[NOT(IN($0, {
+LogicalProject(DEPTNO=[$7])
+  LogicalFilter(condition=[false])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+}))])
+    LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+    <Resource name="planAfter">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
 ]]>
     </Resource>
   </TestCase>
@@ -7762,6 +7810,30 @@ LogicalProject(NAME=[$1], DEPTNO=[$0], EXPR$2=[$2])
   LogicalFilter(condition=[=($0, 0)])
     LogicalWindow(window#0=[window(partition {1} aggs [COUNT()])])
       LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testNotUniqueWithNoRowSubQuery">
+    <Resource name="sql">
+      <![CDATA[select * from dept as d
+where not unique(
+  select deptno from emp e where false)]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalFilter(condition=[NOT(UNIQUE({
+LogicalProject(DEPTNO=[$7])
+  LogicalFilter(condition=[false])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+}))])
+    LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+    <Resource name="planAfter">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalValues(tuples=[[]])
 ]]>
     </Resource>
   </TestCase>
@@ -14508,6 +14580,54 @@ LogicalProject(EMPNO=[$0], ENAME=[$1], JOB=[$2], MGR=[$3], HIREDATE=[$4], SAL=[$
       LogicalAggregate(group=[{0}])
         LogicalProject(NAME=[$1])
           LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testSomeWithGreaterThanNoRowSubQuery">
+    <Resource name="sql">
+      <![CDATA[select * from dept as d
+where deptno > some(
+  select deptno from emp e where false)]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalFilter(condition=[> SOME($0, {
+LogicalProject(DEPTNO=[$7])
+  LogicalFilter(condition=[false])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+})])
+    LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+    <Resource name="planAfter">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalValues(tuples=[[]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testSomeWithLessThanOrEqualNoRowSubQuery">
+    <Resource name="sql">
+      <![CDATA[select * from dept as d
+where deptno <= some(
+  select deptno from emp e where false)]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalFilter(condition=[<= SOME($0, {
+LogicalProject(DEPTNO=[$7])
+  LogicalFilter(condition=[false])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+})])
+    LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+    <Resource name="planAfter">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalValues(tuples=[[]])
 ]]>
     </Resource>
   </TestCase>
@@ -16888,6 +17008,30 @@ LogicalUnion(all=[true])
     <Resource name="planAfter">
       <![CDATA[
 LogicalValues(tuples=[[{ 5 }, { 5 }]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testUniqueWithNoRowSubQuery">
+    <Resource name="sql">
+      <![CDATA[select * from dept as d
+where unique(
+  select deptno from emp e where false)]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalFilter(condition=[UNIQUE({
+LogicalProject(DEPTNO=[$7])
+  LogicalFilter(condition=[false])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+})])
+    LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
+]]>
+    </Resource>
+    <Resource name="planAfter">
+      <![CDATA[
+LogicalProject(DEPTNO=[$0], NAME=[$1])
+  LogicalTableScan(table=[[CATALOG, SALES, DEPT]])
 ]]>
     </Resource>
   </TestCase>