address comments from @marmbrus

Author: Davies Liu

sql/core/src/main/scala/org/apache/spark/sql/sources/DataSourceStrategy.scala

@@ -23,6 +23,7 @@ import org.apache.spark.sql.catalyst.expressions._
 import org.apache.spark.sql.catalyst.planning.PhysicalOperation
 import org.apache.spark.sql.catalyst.plans.logical
 import org.apache.spark.sql.catalyst.plans.logical.LogicalPlan
+import org.apache.spark.sql.execution.SparkPlan
 import org.apache.spark.sql.types.{UTF8String, StringType}
 import org.apache.spark.sql.{Row, Strategy, execution, sources}
 
@@ -114,9 +115,10 @@ private[sql] object DataSourceStrategy extends Strategy {
     }
   }
 
-  private[this] def createPhysicalRDD(relation: BaseRelation,
-                                      output: Seq[Attribute],
-                                      rdd: RDD[Row]) = {
+  private[this] def createPhysicalRDD(
+      relation: BaseRelation,
+      output: Seq[Attribute],
+      rdd: RDD[Row]): SparkPlan = {
     val converted = if (relation.needConversion) {
       execution.RDDConversions.rowToRowRdd(rdd, relation.schema)
     } else {

sql/core/src/main/scala/org/apache/spark/sql/sources/interfaces.scala

@@ -131,6 +131,9 @@ abstract class BaseRelation {
    * Whether does it need to convert the objects in Row to internal representation, for example:
    *  java.lang.String -> UTF8String
    *  java.lang.Decimal -> Decimal
+   *
+   * Note: The internal representation is not stable across releases and thus data sources outside
+   * of Spark SQL should leave this as true.
    */
   def needConversion: Boolean = true
 }