[CALCITE-5289] Assertion failure in MultiJoinOptimizeBushyRule

Signed-off-by: Mihai Budiu <mbudiu@feldera.com>

Author: mihaibudiu

core/src/main/java/org/apache/calcite/rel/rules/MultiJoinOptimizeBushyRule.java

@@ -107,6 +107,17 @@ public MultiJoinOptimizeBushyRule(RelFactories.JoinFactory joinFactory,
     final RelMetadataQuery mq = call.getMetadataQuery();
 
     final LoptMultiJoin multiJoin = new LoptMultiJoin(multiJoinRel);
+    for (int i = 0; i < multiJoin.getNumJoinFactors(); i++) {
+      ImmutableBitSet outerJoinFactors = multiJoin.getOuterJoinFactors(i);
+      if (outerJoinFactors == null) {
+        continue;
+      }
+      if (!outerJoinFactors.isEmpty()) {
+        // Refuse to apply this rule to a multijoin with outer joins,
+        // since this rule cannot handle outer joins.
+        return;
+      }
+    }
 
     final List<Vertex> vertexes = new ArrayList<>();
     int x = 0;

core/src/test/java/org/apache/calcite/test/RelOptRulesTest.java

@@ -3383,6 +3383,15 @@ private void checkPushJoinThroughUnionOnRightDoesNotMatchSemiOrAntiJoin(JoinRelT
         .check();
   }
 
+  /** Test case for <a href="https://issues.apache.org/jira/browse/CALCITE-5289">
+   * [CALCITE-5289] Assertion failure in MultiJoinOptimizeBushyRule</a>. */
+  @Test void testBushyJoinRule() {
+    final String sql = "select emp.ename from emp LEFT JOIN emp AS emp1 on emp.ename = emp1.ename";
+    sql(sql).withPreRule(CoreRules.JOIN_TO_MULTI_JOIN)
+        .withRule(CoreRules.MULTI_JOIN_OPTIMIZE_BUSHY)
+        .checkUnchanged();
+  }
+
   @Test void testConvertMultiJoinRule() {
     final String sql = "select e1.ename from emp e1, dept d, emp e2\n"
         + "where e1.deptno = d.deptno and d.deptno = e2.deptno";

core/src/test/resources/org/apache/calcite/test/RelOptRulesTest.xml

@@ -1374,6 +1374,19 @@ LogicalAggregate(group=[{}], EXPR$0=[SUM($0)], EXPR$1=[COUNT($0)], EXPR$2=[BIT_O
   LogicalAggregate(group=[{0}])
     LogicalProject(DEPTNO=[$7])
       LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+]]>
+    </Resource>
+  </TestCase>
+  <TestCase name="testBushyJoinRule">
+    <Resource name="sql">
+      <![CDATA[select emp.ename from emp LEFT JOIN emp AS emp1 on emp.ename = emp1.ename]]>
+    </Resource>
+    <Resource name="planBefore">
+      <![CDATA[
+LogicalProject(ENAME=[$1])
+  MultiJoin(joinFilter=[true], isFullOuterJoin=[false], joinTypes=[[INNER, LEFT]], outerJoinConditions=[[NULL, =($1, $10)]], projFields=[[ALL, ALL]])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
+    LogicalTableScan(table=[[CATALOG, SALES, EMP]])
 ]]>
     </Resource>
   </TestCase>