[SPARK-20680][SQL] Spark-sql do not support for creating table with void column datatype

python/pyspark/sql/types.py

@@ -116,6 +116,9 @@ class NullType(DataType):
 
     __metaclass__ = DataTypeSingleton
 
+    def simpleString(self):
+        return 'void'
+
 
 class AtomicType(DataType):
     """An internal type used to represent everything that is not

sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/parser/AstBuilder.scala

@@ -2211,6 +2211,7 @@ class AstBuilder(conf: SQLConf) extends SqlBaseBaseVisitor[AnyRef] with Logging
         DecimalType(precision.getText.toInt, 0)
       case ("decimal" | "dec" | "numeric", precision :: scale :: Nil) =>
         DecimalType(precision.getText.toInt, scale.getText.toInt)
+      case ("void", Nil) => NullType
       case ("interval", Nil) => CalendarIntervalType
       case (dt, params) =>
         val dtStr = if (params.nonEmpty) s"$dt(${params.mkString(",")})" else dt

sql/catalyst/src/main/scala/org/apache/spark/sql/types/NullType.scala

@@ -32,6 +32,11 @@ class NullType private() extends DataType {
   override def defaultSize: Int = 1
 
   private[spark] override def asNullable: NullType = this
+
+  /**
+   * Readable string representation for NULL type.
+   */
+  override def simpleString: String = "void"
 }
 
 /**

sql/catalyst/src/test/scala/org/apache/spark/sql/catalyst/parser/DataTypeParserSuite.scala

@@ -61,6 +61,7 @@ class DataTypeParserSuite extends SparkFunSuite {
   checkDataType("varchAr(20)", StringType)
   checkDataType("cHaR(27)", StringType)
   checkDataType("BINARY", BinaryType)
+  checkDataType("void", NullType)
   checkDataType("interval", CalendarIntervalType)
 
   checkDataType("array<doublE>", ArrayType(DoubleType, true))

sql/core/src/test/resources/sql-functions/sql-expression-schema.md

@@ -34,7 +34,7 @@
 | org.apache.spark.sql.catalyst.expressions.Ascii | ascii | SELECT ascii('222') | struct<ascii(222):int> |
 | org.apache.spark.sql.catalyst.expressions.Asin | asin | SELECT asin(0) | struct<ASIN(CAST(0 AS DOUBLE)):double> |
 | org.apache.spark.sql.catalyst.expressions.Asinh | asinh | SELECT asinh(0) | struct<ASINH(CAST(0 AS DOUBLE)):double> |
-| org.apache.spark.sql.catalyst.expressions.AssertTrue | assert_true | SELECT assert_true(0 < 1) | struct<assert_true((0 < 1)):null> |
+| org.apache.spark.sql.catalyst.expressions.AssertTrue | assert_true | SELECT assert_true(0 < 1) | struct<assert_true((0 < 1)):void> |
 | org.apache.spark.sql.catalyst.expressions.Atan | atan | SELECT atan(0) | struct<ATAN(CAST(0 AS DOUBLE)):double> |
 | org.apache.spark.sql.catalyst.expressions.Atan2 | atan2 | SELECT atan2(0, 0) | struct<ATAN2(CAST(0 AS DOUBLE), CAST(0 AS DOUBLE)):double> |
 | org.apache.spark.sql.catalyst.expressions.Atanh | atanh | SELECT atanh(0) | struct<ATANH(CAST(0 AS DOUBLE)):double> |

sql/core/src/test/resources/sql-tests/results/ansi/literals.sql.out

@@ -5,7 +5,7 @@
 -- !query
 select null, Null, nUll
 -- !query schema
-struct<NULL:null,NULL:null,NULL:null>
+struct<NULL:void,NULL:void,NULL:void>
 -- !query output
 NULL	NULL	NULL
 

sql/core/src/test/resources/sql-tests/results/inline-table.sql.out

@@ -49,7 +49,7 @@ two	2
 -- !query
 select * from values ("one", null), ("two", null) as data(a, b)
 -- !query schema
-struct<a:string,b:null>
+struct<a:string,b:void>
 -- !query output
 one	NULL
 two	NULL

sql/core/src/test/resources/sql-tests/results/literals.sql.out

@@ -5,7 +5,7 @@
 -- !query
 select null, Null, nUll
 -- !query schema
-struct<NULL:null,NULL:null,NULL:null>
+struct<NULL:void,NULL:void,NULL:void>
 -- !query output
 NULL	NULL	NULL
 

sql/core/src/test/resources/sql-tests/results/misc-functions.sql.out

@@ -7,7 +7,7 @@ select typeof(null)
 -- !query schema
 struct<typeof(NULL):string>
 -- !query output
-null
+void
 
 
 -- !query

sql/core/src/test/resources/sql-tests/results/postgreSQL/select.sql.out

@@ -308,15 +308,15 @@ struct<1:int>
 -- !query
 select foo.* from (select null) as foo
 -- !query schema
-struct<NULL:null>
+struct<NULL:void>
 -- !query output
 NULL
 
 
 -- !query
 select foo.* from (select 'xyzzy',1,null) as foo
 -- !query schema
-struct<xyzzy:string,1:int,NULL:null>
+struct<xyzzy:string,1:int,NULL:void>
 -- !query output
 xyzzy	1	NULL
 

sql/core/src/test/resources/sql-tests/results/sql-compatibility-functions.sql.out

@@ -5,7 +5,7 @@
 -- !query
 SELECT ifnull(null, 'x'), ifnull('y', 'x'), ifnull(null, null)
 -- !query schema
-struct<ifnull(NULL, x):string,ifnull(y, x):string,ifnull(NULL, NULL):null>
+struct<ifnull(NULL, x):string,ifnull(y, x):string,ifnull(NULL, NULL):void>
 -- !query output
 x	y	NULL
 
@@ -21,15 +21,15 @@ NULL	x
 -- !query
 SELECT nvl(null, 'x'), nvl('y', 'x'), nvl(null, null)
 -- !query schema
-struct<nvl(NULL, x):string,nvl(y, x):string,nvl(NULL, NULL):null>
+struct<nvl(NULL, x):string,nvl(y, x):string,nvl(NULL, NULL):void>
 -- !query output
 x	y	NULL
 
 
 -- !query
 SELECT nvl2(null, 'x', 'y'), nvl2('n', 'x', 'y'), nvl2(null, null, null)
 -- !query schema
-struct<nvl2(NULL, x, y):string,nvl2(n, x, y):string,nvl2(NULL, NULL, NULL):null>
+struct<nvl2(NULL, x, y):string,nvl2(n, x, y):string,nvl2(NULL, NULL, NULL):void>
 -- !query output
 y	x	NULL
 

sql/core/src/test/resources/sql-tests/results/udf/udf-inline-table.sql.out

@@ -49,7 +49,7 @@ two	2
 -- !query
 select udf(a), b from values ("one", null), ("two", null) as data(a, b)
 -- !query schema
-struct<CAST(udf(cast(a as string)) AS STRING):string,b:null>
+struct<CAST(udf(cast(a as string)) AS STRING):string,b:void>
 -- !query output
 one	NULL
 two	NULL

sql/core/src/test/scala/org/apache/spark/sql/FileBasedDataSourceSuite.scala

@@ -403,7 +403,7 @@ class FileBasedDataSourceSuite extends QueryTest
         ""
       }
       def errorMessage(format: String): String = {
-        s"$format data source does not support null data type."
+        s"$format data source does not support void data type."
       }
       withSQLConf(SQLConf.USE_V1_SOURCE_LIST.key -> useV1List) {
         withTempDir { dir =>

sql/hive/src/test/scala/org/apache/spark/sql/hive/execution/HiveDDLSuite.scala

@@ -2309,6 +2309,22 @@ class HiveDDLSuite
     }
   }
 
+  test("SPARK-20680: Spark-sql do not support for void column datatype of view") {
+    withTable("t") {
+      withView("tabNullType") {
+        val client =
+          spark.sharedState.externalCatalog.unwrapped.asInstanceOf[HiveExternalCatalog].client
+        client.runSqlHive("CREATE TABLE t (t1 int)")
+        client.runSqlHive("INSERT INTO t VALUES (3)")
+        client.runSqlHive("CREATE VIEW tabNullType AS SELECT NULL AS col FROM t")
+        checkAnswer(spark.table("tabNullType"), Row(null))
+        // table description shows "void" representation for NULL type.
+        val desc = spark.sql("DESC tabNullType").collect().toSeq
+        assert(desc.contains(Row("col", "void", null)))
+      }
+    }
+  }
+
   test("SPARK-21216: join with a streaming DataFrame") {
     import org.apache.spark.sql.execution.streaming.MemoryStream
     import testImplicits._

sql/hive/src/test/scala/org/apache/spark/sql/hive/orc/HiveOrcSourceSuite.scala

@@ -121,7 +121,7 @@ class HiveOrcSourceSuite extends OrcSuite with TestHiveSingleton {
       msg = intercept[AnalysisException] {
         sql("select null").write.mode("overwrite").orc(orcDir)
       }.getMessage
-      assert(msg.contains("ORC data source does not support null data type."))
+      assert(msg.contains("ORC data source does not support void data type."))
 
       msg = intercept[AnalysisException] {
         spark.udf.register("testType", () => new IntervalData())