apache · xushiyan · May 6, 2022 · May 5, 2022 · May 5, 2022
diff --git a/.github/workflows/bot.yml b/.github/workflows/bot.yml
@@ -59,3 +59,11 @@ jobs:
         if: ${{ !endsWith(env.SPARK_PROFILE, '3.2') }} # skip test spark 3.2 before hadoop upgrade to 3.x
         run:
           mvn test -Punit-tests -D"$SCALA_PROFILE" -D"$SPARK_PROFILE" -D"$FLINK_PROFILE" -DfailIfNoTests=false -pl hudi-examples/hudi-examples-flink,hudi-examples/hudi-examples-java,hudi-examples/hudi-examples-spark
+      - name: Spark SQL Test
+        env:
+          SCALA_PROFILE: ${{ matrix.scalaProfile }}
+          SPARK_PROFILE: ${{ matrix.sparkProfile }}
+          FLINK_PROFILE: ${{ matrix.flinkProfile }}
+        if: ${{ !endsWith(env.SPARK_PROFILE, '2.4') }} # skip test spark 2.4 as it's covered by Azure CI
+        run:
+          mvn test -Punit-tests -D"$SCALA_PROFILE" -D"$SPARK_PROFILE" -D"$FLINK_PROFILE" '-Dtest=org.apache.spark.sql.hudi.Test*' -pl hudi-spark-datasource/hudi-spark
diff --git a/...rk-datasource/hudi-spark/src/test/scala/org/apache/hudi/functional/TestSqlStatement.scala b/...rk-datasource/hudi-spark/src/test/scala/org/apache/hudi/functional/TestSqlStatement.scala
@@ -18,9 +18,9 @@
 package org.apache.hudi.functional
 
 import org.apache.hudi.common.util.FileIOUtils
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 
-class TestSqlStatement extends TestHoodieSqlBase {
+class TestSqlStatement extends HoodieSparkSqlTestBase {
   val STATE_INIT = 0
   val STATE_SKIP_COMMENT = 1
   val STATE_FINISH_COMMENT = 2

diff --git a/...src/test/scala/org/apache/spark/sql/execution/benchmark/SpaceCurveOptimizeBenchmark.scala b/...src/test/scala/org/apache/spark/sql/execution/benchmark/SpaceCurveOptimizeBenchmark.scala
@@ -23,15 +23,15 @@ import org.apache.hudi.ColumnStatsIndexHelper.buildColumnStatsTableFor
 import org.apache.hudi.config.HoodieClusteringConfig.LayoutOptimizationStrategy
 import org.apache.hudi.sort.SpaceCurveSortingHelper
 import org.apache.spark.sql.DataFrame
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 import org.apache.spark.sql.types.{IntegerType, StructField}
 import org.junit.jupiter.api.{Disabled, Tag, Test}
 
 import scala.collection.JavaConversions._
 import scala.util.Random
 
 @Tag("functional")
-object SpaceCurveOptimizeBenchmark extends TestHoodieSqlBase {
+object SpaceCurveOptimizeBenchmark extends HoodieSparkSqlTestBase {
 
   def evalSkippingPercent(tableName: String, co1: String, co2: String, value1: Int, value2: Int): Unit= {
     val sourceTableDF = spark.sql(s"select * from ${tableName}")

diff --git a/...he/spark/sql/hudi/TestHoodieSqlBase.scala → ...ark/sql/hudi/HoodieSparkSqlTestBase.scala b/...he/spark/sql/hudi/TestHoodieSqlBase.scala → ...ark/sql/hudi/HoodieSparkSqlTestBase.scala
@@ -31,7 +31,7 @@ import org.scalatest.{BeforeAndAfterAll, FunSuite, Tag}
 import java.io.File
 import java.util.TimeZone
 
-class TestHoodieSqlBase extends FunSuite with BeforeAndAfterAll {
+class HoodieSparkSqlTestBase extends FunSuite with BeforeAndAfterAll {
   org.apache.log4j.Logger.getRootLogger.setLevel(Level.WARN)
 
   private lazy val sparkWareHouse = {

diff --git a/...spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestAlterTable.scala b/...spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestAlterTable.scala
@@ -22,7 +22,7 @@ import org.apache.hudi.common.table.HoodieTableMetaClient
 import org.apache.spark.sql.catalyst.TableIdentifier
 import org.apache.spark.sql.types.{LongType, StructField, StructType}
 
-class TestAlterTable extends TestHoodieSqlBase {
+class TestAlterTable extends HoodieSparkSqlTestBase {
 
   test("Test Alter Table") {
     withTempDir { tmp =>

diff --git a/...rce/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestAlterTableDropPartition.scala b/...rce/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestAlterTableDropPartition.scala
@@ -23,7 +23,7 @@ import org.apache.hudi.config.HoodieWriteConfig
 import org.apache.hudi.keygen.{ComplexKeyGenerator, SimpleKeyGenerator}
 import org.apache.spark.sql.SaveMode
 
-class TestAlterTableDropPartition extends TestHoodieSqlBase {
+class TestAlterTableDropPartition extends HoodieSparkSqlTestBase {
 
   test("Drop non-partitioned table") {
     val tableName = generateTableName

diff --git a/...-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestCompactionTable.scala b/...-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestCompactionTable.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.hudi
 
-class TestCompactionTable extends TestHoodieSqlBase {
+class TestCompactionTable extends HoodieSparkSqlTestBase {
 
   test("Test compaction table") {
     withTempDir {tmp =>

diff --git a/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestCreateTable.scala b/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestCreateTable.scala
@@ -30,7 +30,7 @@ import org.apache.spark.sql.types._
 
 import scala.collection.JavaConverters._
 
-class TestCreateTable extends TestHoodieSqlBase {
+class TestCreateTable extends HoodieSparkSqlTestBase {
 
   test("Test Create Managed Hoodie Table") {
     val databaseName = "hudi_database"

diff --git a/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestDeleteTable.scala b/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestDeleteTable.scala
@@ -22,7 +22,7 @@ import org.apache.hudi.config.HoodieWriteConfig
 import org.apache.hudi.keygen.SimpleKeyGenerator
 import org.apache.spark.sql.SaveMode
 
-class TestDeleteTable extends TestHoodieSqlBase {
+class TestDeleteTable extends HoodieSparkSqlTestBase {
 
   test("Test Delete Table") {
     withTempDir { tmp =>

diff --git a/...-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestDropTable.scala b/...-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestDropTable.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.hudi
 
-class TestDropTable extends TestHoodieSqlBase {
+class TestDropTable extends HoodieSparkSqlTestBase {
 
   test("Test Drop Table") {
     withTempDir { tmp =>

diff --git a/...tasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestHoodieOptionConfig.scala b/...tasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestHoodieOptionConfig.scala
@@ -19,27 +19,13 @@ package org.apache.spark.sql.hudi
 
 import org.apache.hudi.common.model.{DefaultHoodieRecordPayload, OverwriteWithLatestAvroPayload}
 import org.apache.hudi.common.table.HoodieTableConfig
-import org.apache.hudi.testutils.HoodieClientTestBase
-
-import org.apache.spark.sql.SparkSession
+import org.apache.hudi.testutils.SparkClientFunctionalTestHarness
 import org.apache.spark.sql.types._
-
 import org.junit.jupiter.api.Assertions.assertTrue
-import org.junit.jupiter.api.{BeforeEach, Test}
-
+import org.junit.jupiter.api.Test
 import org.scalatest.Matchers.intercept
 
-class TestHoodieOptionConfig extends HoodieClientTestBase {
-
-  var spark: SparkSession = _
-
-  /**
-   * Setup method running before each test.
-   */
-  @BeforeEach override def setUp() {
-    initSparkContexts()
-    spark = sqlContext.sparkSession
-  }
+class TestHoodieOptionConfig extends SparkClientFunctionalTestHarness {
 
   @Test
   def testWithDefaultSqlOptions(): Unit = {

diff --git a/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestInsertTable.scala b/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestInsertTable.scala
@@ -26,7 +26,7 @@ import org.apache.spark.sql.SaveMode
 
 import java.io.File
 
-class TestInsertTable extends TestHoodieSqlBase {
+class TestInsertTable extends HoodieSparkSqlTestBase {
 
   test("Test Insert Into") {
     withTempDir { tmp =>

diff --git a/...ource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestMergeIntoLogOnlyTable.scala b/...ource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestMergeIntoLogOnlyTable.scala
@@ -19,7 +19,7 @@ package org.apache.spark.sql.hudi
 
 import org.apache.hudi.testutils.DataSourceTestUtils
 
-class TestMergeIntoLogOnlyTable extends TestHoodieSqlBase {
+class TestMergeIntoLogOnlyTable extends HoodieSparkSqlTestBase {
 
   test("Test Query Log Only MOR Table") {
     withTempDir { tmp =>

diff --git a/...k-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestMergeIntoTable.scala b/...k-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestMergeIntoTable.scala
@@ -20,7 +20,7 @@ package org.apache.spark.sql.hudi
 import org.apache.hudi.{DataSourceReadOptions, HoodieDataSourceHelpers}
 import org.apache.hudi.common.fs.FSUtils
 
-class TestMergeIntoTable extends TestHoodieSqlBase {
+class TestMergeIntoTable extends HoodieSparkSqlTestBase {
 
   test("Test MergeInto Basic") {
     withTempDir { tmp =>

diff --git a/...-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestMergeIntoTable2.scala b/...-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestMergeIntoTable2.scala
@@ -21,7 +21,7 @@ import org.apache.hudi.HoodieSparkUtils
 import org.apache.hudi.common.table.HoodieTableMetaClient
 import org.apache.spark.sql.Row
 
-class TestMergeIntoTable2 extends TestHoodieSqlBase {
+class TestMergeIntoTable2 extends HoodieSparkSqlTestBase {
 
   test("Test MergeInto for MOR table 2") {
     withTempDir { tmp =>

diff --git a/...e/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestPartialUpdateForMergeInto.scala b/...e/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestPartialUpdateForMergeInto.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.hudi
 
-class TestPartialUpdateForMergeInto extends TestHoodieSqlBase {
+class TestPartialUpdateForMergeInto extends HoodieSparkSqlTestBase {
 
   test("Test Partial Update") {
     withTempDir { tmp =>

diff --git a/...k-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestShowPartitions.scala b/...k-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestShowPartitions.scala
@@ -19,7 +19,7 @@ package org.apache.spark.sql.hudi
 
 import org.apache.spark.sql.Row
 
-class TestShowPartitions extends TestHoodieSqlBase {
+class TestShowPartitions extends HoodieSparkSqlTestBase {
 
   test("Test Show Non Partitioned Table's Partitions") {
     val tableName = generateTableName

diff --git a/...-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestSpark3DDL.scala b/...-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestSpark3DDL.scala
@@ -27,7 +27,7 @@ import org.apache.spark.sql.{DataFrame, Row, SaveMode, SparkSession}
 import scala.collection.JavaConversions._
 import scala.collection.JavaConverters._
 
-class TestSpark3DDL extends TestHoodieSqlBase {
+class TestSpark3DDL extends HoodieSparkSqlTestBase {
 
   def createTestResult(tableName: String): Array[Row] = {
     spark.sql(s"select * from ${tableName} order by id")

diff --git a/hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestSqlConf.scala b/hudi-spark-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestSqlConf.scala
@@ -28,7 +28,7 @@ import java.nio.file.{Files, Paths}
 
 import org.scalatest.BeforeAndAfter
 
-class TestSqlConf extends TestHoodieSqlBase with BeforeAndAfter {
+class TestSqlConf extends HoodieSparkSqlTestBase with BeforeAndAfter {
 
   def setEnv(key: String, value: String): String = {
     val field = System.getenv().getClass.getDeclaredField("m")

diff --git a/...-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestTimeTravelTable.scala b/...-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestTimeTravelTable.scala
@@ -20,7 +20,7 @@ package org.apache.spark.sql.hudi
 import org.apache.hudi.HoodieSparkUtils
 import org.apache.hudi.common.table.HoodieTableMetaClient
 
-class TestTimeTravelTable extends TestHoodieSqlBase {
+class TestTimeTravelTable extends HoodieSparkSqlTestBase {
   test("Test Insert and Update Record with time travel") {
     if (HoodieSparkUtils.gteqSpark3_2) {
       withTempDir { tmp =>

diff --git a/...rk-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestTruncateTable.scala b/...rk-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestTruncateTable.scala
@@ -23,7 +23,7 @@ import org.apache.hudi.config.HoodieWriteConfig
 import org.apache.hudi.keygen.{ComplexKeyGenerator, SimpleKeyGenerator}
 import org.apache.spark.sql.SaveMode
 
-class TestTruncateTable extends TestHoodieSqlBase {
+class TestTruncateTable extends HoodieSparkSqlTestBase {
 
   test("Test Truncate non-partitioned Table") {
     Seq("cow", "mor").foreach { tableType =>

diff --git a/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestUpdateTable.scala b/...park-datasource/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/TestUpdateTable.scala
@@ -17,7 +17,7 @@
 
 package org.apache.spark.sql.hudi
 
-class TestUpdateTable extends TestHoodieSqlBase {
+class TestUpdateTable extends HoodieSparkSqlTestBase {
 
   test("Test Update Table") {
     withTempDir { tmp =>

diff --git a/...hudi-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCallCommandParser.scala b/...hudi-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCallCommandParser.scala
@@ -21,13 +21,13 @@ import com.google.common.collect.ImmutableList
 import org.apache.hudi.HoodieSparkUtils
 import org.apache.spark.sql.catalyst.expressions.Literal
 import org.apache.spark.sql.catalyst.plans.logical.{CallCommand, NamedArgument, PositionalArgument}
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 import org.apache.spark.sql.types.{DataType, DataTypes}
 
 import java.math.BigDecimal
 import scala.collection.JavaConverters
 
-class TestCallCommandParser extends TestHoodieSqlBase {
+class TestCallCommandParser extends HoodieSparkSqlTestBase {
   private val parser = spark.sessionState.sqlParser
 
   test("Test Call Produce with Positional Arguments") {

diff --git a/...rce/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCallProcedure.scala b/...rce/hudi-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCallProcedure.scala
@@ -17,9 +17,9 @@
 
 package org.apache.spark.sql.hudi.procedure
 
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 
-class TestCallProcedure extends TestHoodieSqlBase {
+class TestCallProcedure extends HoodieSparkSqlTestBase {
 
   test("Test Call show_commits Procedure") {
     withTempDir { tmp =>

diff --git a/...di-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestClusteringProcedure.scala b/...di-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestClusteringProcedure.scala
@@ -24,11 +24,11 @@ import org.apache.hudi.common.table.timeline.{HoodieActiveTimeline, HoodieTimeli
 import org.apache.hudi.common.util.{Option => HOption}
 import org.apache.hudi.{HoodieCLIUtils, HoodieDataSourceHelpers}
 
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 
 import scala.collection.JavaConverters.asScalaIteratorConverter
 
-class TestClusteringProcedure extends TestHoodieSqlBase {
+class TestClusteringProcedure extends HoodieSparkSqlTestBase {
 
   test("Test Call run_clustering Procedure By Table") {
     withTempDir { tmp =>

diff --git a/...di-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCompactionProcedure.scala b/...di-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestCompactionProcedure.scala
@@ -19,9 +19,9 @@
 
 package org.apache.spark.sql.hudi.procedure
 
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 
-class TestCompactionProcedure extends TestHoodieSqlBase {
+class TestCompactionProcedure extends HoodieSparkSqlTestBase {
 
   test("Test Call run_compaction Procedure by Table") {
     withTempDir { tmp =>

diff --git a/...di-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestSavepointsProcedure.scala b/...di-spark/src/test/scala/org/apache/spark/sql/hudi/procedure/TestSavepointsProcedure.scala
@@ -17,9 +17,9 @@
 
 package org.apache.spark.sql.hudi.procedure
 
-import org.apache.spark.sql.hudi.TestHoodieSqlBase
+import org.apache.spark.sql.hudi.HoodieSparkSqlTestBase
 
-class TestSavepointsProcedure extends TestHoodieSqlBase {
+class TestSavepointsProcedure extends HoodieSparkSqlTestBase {
 
   test("Test Call create_savepoints Procedure") {
     withTempDir { tmp =>