[CALCITE-3539] EnumerableDefaults#nestedLoopJoin returns duplicates f…

…or JoinType.SEMI
apache · Nov 30, 2019 · 80f411d · 80f411d
1 parent 67e9367
commit 80f411d
Show file tree

Hide file tree

Showing 3 changed files with 76 additions and 17 deletions.
diff --git a/core/src/main/java/org/apache/calcite/adapter/enumerable/EnumerableNestedLoopJoin.java b/core/src/main/java/org/apache/calcite/adapter/enumerable/EnumerableNestedLoopJoin.java
@@ -25,6 +25,7 @@
 import org.apache.calcite.plan.RelTraitSet;
 import org.apache.calcite.rel.RelCollationTraitDef;
 import org.apache.calcite.rel.RelNode;
+import org.apache.calcite.rel.RelNodes;
 import org.apache.calcite.rel.core.CorrelationId;
 import org.apache.calcite.rel.core.Join;
 import org.apache.calcite.rel.core.JoinRelType;
@@ -89,11 +90,15 @@ public static EnumerableNestedLoopJoin create(
     // and have the same cost. To make the results stable between versions of
     // the planner, make one of the versions slightly more expensive.
     switch (joinType) {
+    case SEMI:
+    case ANTI:
+      // SEMI and ANTI join cannot be flipped
+      break;
     case RIGHT:
       rowCount = RelMdUtil.addEpsilon(rowCount);
       break;
     default:
-      if (left.getId() > right.getId()) {
+      if (RelNodes.COMPARATOR.compare(left, right) > 0) {
         rowCount = RelMdUtil.addEpsilon(rowCount);
       }
     }

diff --git a/core/src/test/java/org/apache/calcite/runtime/EnumerablesTest.java b/core/src/test/java/org/apache/calcite/runtime/EnumerablesTest.java
@@ -62,16 +62,39 @@ public class EnumerablesTest {
           + ", " + (v1 == null ? null : v1.name)
           + "}";
 
-  private static final Predicate2<Emp, Dept> EQUAL_DEPTNO =
+  private static final Predicate2<Emp, Dept> EMP_DEPT_EQUAL_DEPTNO =
       (e, d) -> e.deptno == d.deptno;
+  private static final Predicate2<Dept, Emp> DEPT_EMP_EQUAL_DEPTNO =
+      (d, e) -> d.deptno == e.deptno;
 
-  @Test public void testSemiJoin() {
+  @Test public void testSemiJoinEmp() {
     assertThat(
         EnumerableDefaults.semiJoin(EMPS, DEPTS, e -> e.deptno, d -> d.deptno,
             Functions.identityComparer()).toList().toString(),
         equalTo("[Emp(20, Theodore), Emp(20, Sebastian)]"));
   }
 
+  @Test public void testSemiJoinDept() {
+    assertThat(
+        EnumerableDefaults.semiJoin(DEPTS, EMPS, d -> d.deptno, e -> e.deptno,
+            Functions.identityComparer()).toList().toString(),
+        equalTo("[Dept(20, Sales)]"));
+  }
+
+  @Test public void testAntiJoinEmp() {
+    assertThat(
+        EnumerableDefaults.antiJoin(EMPS, DEPTS, e -> e.deptno, d -> d.deptno,
+            Functions.identityComparer()).toList().toString(),
+        equalTo("[Emp(10, Fred), Emp(30, Joe)]"));
+  }
+
+  @Test public void testAntiJoinDept() {
+    assertThat(
+        EnumerableDefaults.antiJoin(DEPTS, EMPS, d -> d.deptno, e -> e.deptno,
+            Functions.identityComparer()).toList().toString(),
+        equalTo("[Dept(15, Marketing)]"));
+  }
+
   @Test public void testMergeJoin() {
     assertThat(
         EnumerableDefaults.mergeJoin(
@@ -154,61 +177,89 @@ private static <T extends Comparable<T>> Enumerable<T> intersect(
         Functions.identitySelector(), (v0, v1) -> v0, false, false);
   }
 
-  @Test public void testThetaJoin() {
+  @Test public void testNestedLoopJoin() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.INNER).toList().toString(),
         equalTo("[{Theodore, 20, 20, Sales}, {Sebastian, 20, 20, Sales}]"));
   }
 
-  @Test public void testThetaLeftJoin() {
+  @Test public void testNestedLoopLeftJoin() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.LEFT).toList().toString(),
         equalTo("[{Fred, 10, null, null}, {Theodore, 20, 20, Sales}, "
             + "{Sebastian, 20, 20, Sales}, {Joe, 30, null, null}]"));
   }
 
-  @Test public void testThetaRightJoin() {
+  @Test public void testNestedLoopRightJoin() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.RIGHT).toList().toString(),
         equalTo("[{Theodore, 20, 20, Sales}, {Sebastian, 20, 20, Sales}, "
             + "{null, null, 15, Marketing}]"));
   }
 
-  @Test public void testThetaFullJoin() {
+  @Test public void testNestedLoopFullJoin() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.FULL).toList().toString(),
         equalTo("[{Fred, 10, null, null}, {Theodore, 20, 20, Sales}, "
             + "{Sebastian, 20, 20, Sales}, {Joe, 30, null, null}, "
             + "{null, null, 15, Marketing}]"));
   }
 
-  @Test public void testThetaFullJoinLeftEmpty() {
+  @Test public void testNestedLoopFullJoinLeftEmpty() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS.take(0), DEPTS, EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS.take(0), DEPTS, EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.FULL)
             .orderBy(Functions.identitySelector()).toList().toString(),
         equalTo("[{null, null, 15, Marketing}, {null, null, 20, Sales}]"));
   }
 
-  @Test public void testThetaFullJoinRightEmpty() {
+  @Test public void testNestedLoopFullJoinRightEmpty() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS.take(0), EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS.take(0), EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.FULL).toList().toString(),
         equalTo("[{Fred, 10, null, null}, {Theodore, 20, null, null}, "
             + "{Sebastian, 20, null, null}, {Joe, 30, null, null}]"));
   }
 
-  @Test public void testThetaFullJoinBothEmpty() {
+  @Test public void testNestedLoopFullJoinBothEmpty() {
     assertThat(
-        EnumerableDefaults.nestedLoopJoin(EMPS.take(0), DEPTS.take(0), EQUAL_DEPTNO,
+        EnumerableDefaults.nestedLoopJoin(EMPS.take(0), DEPTS.take(0), EMP_DEPT_EQUAL_DEPTNO,
             EMP_DEPT_TO_STRING, JoinType.FULL).toList().toString(),
         equalTo("[]"));
   }
 
+  @Test public void testNestedLoopSemiJoinEmp() {
+    assertThat(
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EMP_DEPT_EQUAL_DEPTNO,
+            (e, d) -> e.toString(), JoinType.SEMI).toList().toString(),
+        equalTo("[Emp(20, Theodore), Emp(20, Sebastian)]"));
+  }
+
+  @Test public void testNestedLoopSemiJoinDept() {
+    assertThat(
+        EnumerableDefaults.nestedLoopJoin(DEPTS, EMPS, DEPT_EMP_EQUAL_DEPTNO,
+            (d, e) -> d.toString(), JoinType.SEMI).toList().toString(),
+        equalTo("[Dept(20, Sales)]"));
+  }
+
+  @Test public void testNestedLoopAntiJoinEmp() {
+    assertThat(
+        EnumerableDefaults.nestedLoopJoin(EMPS, DEPTS, EMP_DEPT_EQUAL_DEPTNO,
+            (e, d) -> e.toString(), JoinType.ANTI).toList().toString(),
+        equalTo("[Emp(10, Fred), Emp(30, Joe)]"));
+  }
+
+  @Test public void testNestedLoopAntiJoinDept() {
+    assertThat(
+        EnumerableDefaults.nestedLoopJoin(DEPTS, EMPS, DEPT_EMP_EQUAL_DEPTNO,
+            (d, e) -> d.toString(), JoinType.ANTI).toList().toString(),
+        equalTo("[Dept(15, Marketing)]"));
+  }
+
   @Test
   @Ignore // TODO fix this
   public void testMatch() {

diff --git a/linq4j/src/main/java/org/apache/calcite/linq4j/EnumerableDefaults.java b/linq4j/src/main/java/org/apache/calcite/linq4j/EnumerableDefaults.java
@@ -1831,6 +1831,9 @@ public static <TSource, TInner, TResult> Enumerable<TResult> nestedLoopJoin(
               rightUnmatched.remove(right);
             }
             result.add(resultSelector.apply(left, right));
+            if (joinType == JoinType.SEMI) {
+              break;
+            }
           }
         }
       }
-Original file line number
+Diff line change
@@ Expand Up @@
                   rightUnmatched.remove(right);
                 }
                 result.add(resultSelector.apply(left, right));
+                if (joinType == JoinType.SEMI) {
+                  break;
+                }
               }
             }
           }
@@ Expand Down @@