Optimize code

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

@@ -7,8 +7,8 @@ SELECT * FROM testdata LIMIT 2
 -- !query schema
 struct<key:int,value:string>
 -- !query output
-1	1
-2	2
+51	51
+52	52
 
 
 -- !query
@@ -34,17 +34,17 @@ SELECT * FROM testdata LIMIT 2 + 1
 -- !query schema
 struct<key:int,value:string>
 -- !query output
-1	1
-2	2
-3	3
+51	51
+52	52
+53	53
 
 
 -- !query
 SELECT * FROM testdata LIMIT CAST(1 AS int)
 -- !query schema
 struct<key:int,value:string>
 -- !query output
-1	1
+51	51
 
 
 -- !query
@@ -70,7 +70,7 @@ SELECT * FROM testdata LIMIT CAST(1 AS INT)
 -- !query schema
 struct<key:int,value:string>
 -- !query output
-1	1
+51	51
 
 
 -- !query
@@ -88,7 +88,7 @@ SELECT * FROM testdata LIMIT key > 3
 struct<>
 -- !query output
 org.apache.spark.sql.AnalysisException
-The limit expression must evaluate to a constant value, but got (testdata.`key` > 3);
+The limit expression must evaluate to a constant value, but got (spark_catalog.default.testdata.`key` > 3);
 
 
 -- !query

sql/core/src/test/resources/sql-tests/results/show-tables.sql.out

@@ -63,31 +63,19 @@ SHOW TABLES
 -- !query schema
 struct<database:string,tableName:string,isTemporary:boolean>
 -- !query output
-aggtest
-arraydata
-mapdata
-onek
 show_t1
 show_t2
 show_t3
-tenk1
-testdata
 
 
 -- !query
 SHOW TABLES IN showdb
 -- !query schema
 struct<database:string,tableName:string,isTemporary:boolean>
 -- !query output
-aggtest
-arraydata
-mapdata
-onek
 show_t1
 show_t2
 show_t3
-tenk1
-testdata
 
 
 -- !query

sql/core/src/test/resources/sql-tests/results/show-views.sql.out

@@ -63,12 +63,6 @@ SHOW VIEWS
 -- !query schema
 struct<namespace:string,viewName:string,isTemporary:boolean>
 -- !query output
-aggtest
-arraydata
-mapdata
-onek
-tenk1
-testdata
 view_1
 view_2
 view_4
@@ -79,12 +73,6 @@ SHOW VIEWS FROM showdb
 -- !query schema
 struct<namespace:string,viewName:string,isTemporary:boolean>
 -- !query output
-aggtest
-arraydata
-mapdata
-onek
-tenk1
-testdata
 view_1
 view_2
 view_4
@@ -95,12 +83,6 @@ SHOW VIEWS IN showdb
 -- !query schema
 struct<namespace:string,viewName:string,isTemporary:boolean>
 -- !query output
-aggtest
-arraydata
-mapdata
-onek
-tenk1
-testdata
 view_1
 view_2
 view_4
@@ -111,12 +93,6 @@ SHOW VIEWS IN global_temp
 -- !query schema
 struct<namespace:string,viewName:string,isTemporary:boolean>
 -- !query output
-aggtest
-arraydata
-mapdata
-onek
-tenk1
-testdata
 view_3
 view_4
 

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

@@ -363,7 +363,6 @@ class SQLQueryTestSuite extends QueryTest with SharedSparkSession {
     // Create a local SparkSession to have stronger isolation between different test cases.
     // This does not isolate catalog changes.
     val localSparkSession = spark.newSession()
-    loadTestData(localSparkSession)
 
     testCase match {
       case udfTestCase: UDFTest =>
@@ -575,27 +574,36 @@ class SQLQueryTestSuite extends QueryTest with SharedSparkSession {
   private def loadTestData(session: SparkSession): Unit = {
     import session.implicits._
 
-    (1 to 100).map(i => (i, i.toString)).toDF("key", "value").createOrReplaceTempView("testdata")
+    (1 to 100).map(i => (i, i.toString)).toDF("key", "value")
+      .write
+      .format("parquet")
+      .saveAsTable("testdata")
 
     ((Seq(1, 2, 3), Seq(Seq(1, 2, 3))) :: (Seq(2, 3, 4), Seq(Seq(2, 3, 4))) :: Nil)
       .toDF("arraycol", "nestedarraycol")
-      .createOrReplaceTempView("arraydata")
+      .write
+      .format("parquet")
+      .saveAsTable("arraydata")
 
     (Tuple1(Map(1 -> "a1", 2 -> "b1", 3 -> "c1", 4 -> "d1", 5 -> "e1")) ::
       Tuple1(Map(1 -> "a2", 2 -> "b2", 3 -> "c2", 4 -> "d2")) ::
       Tuple1(Map(1 -> "a3", 2 -> "b3", 3 -> "c3")) ::
       Tuple1(Map(1 -> "a4", 2 -> "b4")) ::
       Tuple1(Map(1 -> "a5")) :: Nil)
       .toDF("mapcol")
-      .createOrReplaceTempView("mapdata")
+      .write
+      .format("parquet")
+      .saveAsTable("mapdata")
 
     session
       .read
       .format("csv")
       .options(Map("delimiter" -> "\t", "header" -> "false"))
       .schema("a int, b float")
       .load(testFile("test-data/postgresql/agg.data"))
-      .createOrReplaceTempView("aggtest")
+      .write
+      .format("parquet")
+      .saveAsTable("aggtest")
 
     session
       .read
@@ -621,7 +629,9 @@ class SQLQueryTestSuite extends QueryTest with SharedSparkSession {
           |string4 string
         """.stripMargin)
       .load(testFile("test-data/postgresql/onek.data"))
-      .createOrReplaceTempView("onek")
+      .write
+      .format("parquet")
+      .saveAsTable("onek")
 
     session
       .read
@@ -647,14 +657,26 @@ class SQLQueryTestSuite extends QueryTest with SharedSparkSession {
           |string4 string
         """.stripMargin)
       .load(testFile("test-data/postgresql/tenk.data"))
-      .createOrReplaceTempView("tenk1")
+      .write
+      .format("parquet")
+      .saveAsTable("tenk1")
+  }
+
+  private def unloadTestData(session: SparkSession): Unit = {
+    session.sql("DROP TABLE IF EXISTS testdata")
+    session.sql("DROP TABLE IF EXISTS arraydata")
+    session.sql("DROP TABLE IF EXISTS mapdata")
+    session.sql("DROP TABLE IF EXISTS aggtest")
+    session.sql("DROP TABLE IF EXISTS onek")
+    session.sql("DROP TABLE IF EXISTS tenk1")
   }
 
   private val originalTimeZone = TimeZone.getDefault
   private val originalLocale = Locale.getDefault
 
   override def beforeAll(): Unit = {
     super.beforeAll()
+    loadTestData(spark)
     // Timezone is fixed to America/Los_Angeles for those timezone sensitive tests (timestamp_*)
     TimeZone.setDefault(TimeZone.getTimeZone("America/Los_Angeles"))
     // Add Locale setting
@@ -668,6 +690,7 @@ class SQLQueryTestSuite extends QueryTest with SharedSparkSession {
     try {
       TimeZone.setDefault(originalTimeZone)
       Locale.setDefault(originalLocale)
+      unloadTestData(spark)
 
       // For debugging dump some statistics about how much time was spent in various optimizer rules
       logWarning(RuleExecutor.dumpTimeSpent())