[VL] Enable right and anti join in smj (apache#6449)

jinchengchenghh · Jul 15, 2024 · a8dd354 · a8dd354
1 parent de26785
commit a8dd354
Show file tree

Hide file tree

Showing 2 changed files with 65 additions and 1 deletion.
diff --git a/backends-velox/src/test/scala/org/apache/gluten/execution/TestOperator.scala b/backends-velox/src/test/scala/org/apache/gluten/execution/TestOperator.scala
@@ -1442,6 +1442,68 @@ class TestOperator extends VeloxWholeStageTransformerSuite with AdaptiveSparkPla
     }
   }
 
+  test("test sort merge join") {
+    withTable("t1", "t2") {
+      sql("""
+            |create table t1 using parquet as
+            |select cast(id as int) as c1, cast(id as string) c2 from range(100)
+            |""".stripMargin)
+      sql("""
+            |create table t2 using parquet as
+            |select cast(id as int) as c1, cast(id as string) c2 from range(100) order by c1 desc;
+            |""".stripMargin)
+      withSQLConf("spark.gluten.sql.columnar.forceShuffledHashJoin" -> "false") {
+        runQueryAndCompare(
+          """
+            |select * from t1 inner join t2 on t1.c1 = t2.c1 and t1.c1 > 50;
+            |""".stripMargin
+        ) {
+          checkGlutenOperatorMatch[SortMergeJoinExecTransformer]
+        }
+      }
+
+      withSQLConf("spark.gluten.sql.columnar.forceShuffledHashJoin" -> "false") {
+        runQueryAndCompare(
+          """
+            |select * from t1 left join t2 on t1.c1 = t2.c1 and t1.c1 > 50;
+            |""".stripMargin
+        ) {
+          checkGlutenOperatorMatch[SortMergeJoinExecTransformer]
+        }
+      }
+
+      withSQLConf("spark.gluten.sql.columnar.forceShuffledHashJoin" -> "false") {
+        runQueryAndCompare(
+          """
+            |select * from t1 left semi join t2 on t1.c1 = t2.c1 and t1.c1 > 50;
+            |""".stripMargin
+        ) {
+          checkGlutenOperatorMatch[SortMergeJoinExecTransformer]
+        }
+      }
+
+      withSQLConf("spark.gluten.sql.columnar.forceShuffledHashJoin" -> "false") {
+        runQueryAndCompare(
+          """
+            |select * from t1 right join t2 on t1.c1 = t2.c1 and t1.c1 > 50;
+            |""".stripMargin
+        ) {
+          checkGlutenOperatorMatch[SortMergeJoinExecTransformer]
+        }
+      }
+
+      withSQLConf("spark.gluten.sql.columnar.forceShuffledHashJoin" -> "false") {
+        runQueryAndCompare(
+          """
+            |select * from t1 left anti join t2 on t1.c1 = t2.c1 and t1.c1 > 50;
+            |""".stripMargin
+        ) {
+          checkGlutenOperatorMatch[SortMergeJoinExecTransformer]
+        }
+      }
+    }
+  }
+
   test("Fix incorrect path by decode") {
     val c = "?.+<_>|/"
     val path = rootPath + "/test +?.+<_>|"

diff --git a/cpp/velox/substrait/SubstraitToVeloxPlanValidator.cc b/cpp/velox/substrait/SubstraitToVeloxPlanValidator.cc
@@ -906,11 +906,13 @@ bool SubstraitToVeloxPlanValidator::validate(const ::substrait::JoinRel& joinRel
     switch (joinRel.type()) {
       case ::substrait::JoinRel_JoinType_JOIN_TYPE_INNER:
       case ::substrait::JoinRel_JoinType_JOIN_TYPE_LEFT:
+      case ::substrait::JoinRel_JoinType_JOIN_TYPE_RIGHT:
       case ::substrait::JoinRel_JoinType_JOIN_TYPE_LEFT_SEMI:
       case ::substrait::JoinRel_JoinType_JOIN_TYPE_RIGHT_SEMI:
+      case ::substrait::JoinRel_JoinType_JOIN_TYPE_ANTI:
         break;
       default:
-        LOG_VALIDATION_MSG("Sort merge join only support inner, left, left semi and right semi join.");
+        LOG_VALIDATION_MSG("Sort merge join type is not supported: " + std::to_string(joinRel.type()));
         return false;
     }
   }