[SPARK-6564][SQL] SQLContext.emptyDataFrame should contain 0 row, not 1 row

Author: Reynold Xin <rxin@databricks.com>

Closes #5226 from rxin/empty-df and squashes the following commits:

1306d88 [Reynold Xin] Proper fix.
e135bb9 [Reynold Xin] [SPARK-6564][SQL] SQLContext.emptyDataFrame should contain 0 rows, not 1 row.

3 files changed, 8 insertions, 3 deletions

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala b/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala
index e59cf9b9e0..b8100782ec 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/SQLContext.scala
@@ -31,7 +31,7 @@ import org.apache.spark.rdd.RDD
 import org.apache.spark.sql.catalyst.analysis._
 import org.apache.spark.sql.catalyst.expressions._
 import org.apache.spark.sql.catalyst.optimizer.{DefaultOptimizer, Optimizer}
-import org.apache.spark.sql.catalyst.plans.logical.{LocalRelation, LogicalPlan, NoRelation}
+import org.apache.spark.sql.catalyst.plans.logical.{LocalRelation, LogicalPlan, OneRowRelation}
 import org.apache.spark.sql.catalyst.rules.RuleExecutor
 import org.apache.spark.sql.catalyst.{ScalaReflection, expressions}
 import org.apache.spark.sql.execution.{Filter, _}
@@ -177,7 +177,7 @@ class SQLContext(@transient val sparkContext: SparkContext)
    */
   @Experimental
   @transient
-  lazy val emptyDataFrame = DataFrame(this, NoRelation)
+  lazy val emptyDataFrame: DataFrame = createDataFrame(sparkContext.emptyRDD[Row], StructType(Nil))
 
   /**
    * A collection of methods for registering user-defined functions (UDF).
diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala
index 2b581152e5..f754fa770d 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkStrategies.scala
@@ -296,7 +296,7 @@ private[sql] abstract class SparkStrategies extends QueryPlanner[SparkPlan] {
         execution.Intersect(planLater(left), planLater(right)) :: Nil
       case logical.Generate(generator, join, outer, _, child) =>
         execution.Generate(generator, join = join, outer = outer, planLater(child)) :: Nil
-      case logical.NoRelation =>
+      case logical.OneRowRelation =>
         execution.PhysicalRDD(Nil, singleRowRdd) :: Nil
       case logical.Repartition(expressions, child) =>
         execution.Exchange(HashPartitioning(expressions, numPartitions), planLater(child)) :: Nil
diff --git a/sql/core/src/test/scala/org/apache/spark/sql/DataFrameSuite.scala b/sql/core/src/test/scala/org/apache/spark/sql/DataFrameSuite.scala
index 5f03805d70..6761d996fd 100644
--- a/sql/core/src/test/scala/org/apache/spark/sql/DataFrameSuite.scala
+++ b/sql/core/src/test/scala/org/apache/spark/sql/DataFrameSuite.scala
@@ -84,6 +84,11 @@ class DataFrameSuite extends QueryTest {
       testData.collect().toSeq)
   }
 
+  test("empty data frame") {
+    assert(TestSQLContext.emptyDataFrame.columns.toSeq === Seq.empty[String])
+    assert(TestSQLContext.emptyDataFrame.count() === 0)
+  }
+
   test("head and take") {
     assert(testData.take(2) === testData.collect().take(2))
     assert(testData.head(2) === testData.collect().take(2))