[GitHub] [spark] MaxGekk commented on a change in pull request #29339: [Spark-32512][SQL] add alter table add/drop partition command for datasourcev2

[GitBox <gi...@apache.org>]

MaxGekk commented on a change in pull request #29339:
URL: https://github.com/apache/spark/pull/29339#discussion_r493827163



##########
File path: sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/plans/logical/v2Commands.scala
##########
@@ -551,3 +552,31 @@ case class ShowFunctions(
     pattern: Option[String]) extends Command {
   override def children: Seq[LogicalPlan] = child.toSeq
 }
+
+/**
+ * The logical plan of the ALTER TABLE ADD PARTITION command that works for v2 tables.
+ *
+ * The syntax of this command is:
+ * {{{
+ *     ALTER TABLE table ADD [IF NOT EXISTS] PARTITION spec1 [LOCATION 'loc1']
+ *                                          PARTITION spec2 [LOCATION 'loc2']

Review comment:
       Is the second `PARTITION` mandatory? Can you use similar format as for `DROP`?
   ```
   ALTER TABLE table ADD [IF NOT EXISTS] PARTITION spec1 [LOCATION 'loc1'] [, PARTITION spec2 [LOCATION 'loc2'], ...];
   ```

##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/v2/DataSourceV2Strategy.scala
##########
@@ -276,6 +276,13 @@ class DataSourceV2Strategy(session: SparkSession) extends Strategy with Predicat
     case r @ ShowTableProperties(rt: ResolvedTable, propertyKey) =>
       ShowTablePropertiesExec(r.output, rt.table, propertyKey) :: Nil
 
+    case AlterTableAddPartition(table, parts, ignoreIfExists) =>
+      AlterTableAddPartitionExec(table, parts, ignoreIfExists) :: Nil
+
+    case AlterTableDropPartition(table, partIdents, ignoreIfNotExists) =>
+      AlterTableDropPartitionExec(
+        table, partIdents, ignoreIfNotExists) :: Nil

Review comment:
       ```suggestion
         AlterTableDropPartitionExec(table, partIdents, ignoreIfNotExists) :: Nil
   ```

##########
File path: sql/catalyst/src/test/scala/org/apache/spark/sql/connector/InMemoryPartitionTableCatalog.scala
##########
@@ -0,0 +1,46 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.connector
+import java.util

Review comment:
       nit: add a blank line before `import` 

##########
File path: sql/core/src/test/scala/org/apache/spark/sql/connector/DataSourceV2SQLSuite.scala
##########
@@ -2066,13 +2069,71 @@ class DataSourceV2SQLSuite
   }
 
   test("ALTER TABLE ADD PARTITION") {
-    val t = "testcat.ns1.ns2.tbl"
+    val t = "testpart.ns1.ns2.tbl"
     withTable(t) {
       spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
-      val e = intercept[AnalysisException] {
-        sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
-      }
-      assert(e.message.contains("ALTER TABLE ADD PARTITION is only supported with v1 tables"))
+      spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      partTable.clearPartitions()
+    }
+  }
+
+  test("ALTER TABLE ADD PARTITIONS") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(
+        s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc' PARTITION (id=2) LOCATION 'loc1'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(2))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))

Review comment:
       ```suggestion
         val partMetadata = partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
   ```

##########
File path: sql/core/src/test/scala/org/apache/spark/sql/connector/DataSourceV2SQLSuite.scala
##########
@@ -2087,14 +2148,52 @@ class DataSourceV2SQLSuite
     }
   }
 
+  test("ALTER TABLE DROP PARTITION") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
+      spark.sql(s"ALTER TABLE $t DROP PARTITION (id=1)")
+
+      val partTable =
+        catalog("testpart").asTableCatalog.loadTable(Identifier.of(Array("ns1", "ns2"), "tbl"))
+      assert(!partTable.asPartitionable.partitionExists(InternalRow.fromSeq(Seq(1))))
+    }
+  }
+
   test("ALTER TABLE DROP PARTITIONS") {
-    val t = "testcat.ns1.ns2.tbl"
+    val t = "testpart.ns1.ns2.tbl"
     withTable(t) {
       spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
-      val e = intercept[AnalysisException] {
-        sql(s"ALTER TABLE $t DROP PARTITION (id=1)")
-      }
-      assert(e.message.contains("ALTER TABLE DROP PARTITION is only supported with v1 tables"))
+      spark.sql(s"ALTER TABLE $t ADD IF NOT EXISTS PARTITION (id=1) LOCATION 'loc'" +
+        s" PARTITION (id=2) LOCATION 'loc1'")

Review comment:
       remove `s`

##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/v2/AlterTableAddPartitionExec.scala
##########
@@ -0,0 +1,62 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.datasources.v2
+
+import scala.collection.JavaConverters._
+
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.analysis.PartitionsAlreadyExistException
+import org.apache.spark.sql.catalyst.expressions.Attribute
+import org.apache.spark.sql.connector.catalog.{SupportsAtomicPartitionManagement, SupportsPartitionManagement}
+
+/**
+ * Physical plan node for adding partitions of table.
+ */
+case class AlterTableAddPartitionExec(
+    table: SupportsPartitionManagement,
+    partitions: Seq[(InternalRow, Map[String, String])],
+    ignoreIfExists: Boolean) extends V2CommandExec {
+  import DataSourceV2Implicits._
+
+  override def output: Seq[Attribute] = Seq.empty
+
+  override protected def run(): Seq[InternalRow] = {
+    val existsPartIdents = partitions.map(_._1).filter(table.partitionExists)

Review comment:
       Could you traverse `partitions ` only once:
   ```suggestion
       val (existsPartIdents, notExistsPartitions) = partitions
         .partition(p => table.partitionExists(p._1))
   ```

##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/v2/AlterTableDropPartitionExec.scala
##########
@@ -0,0 +1,57 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.datasources.v2
+
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.analysis.NoSuchPartitionsException
+import org.apache.spark.sql.catalyst.expressions.Attribute
+import org.apache.spark.sql.connector.catalog.{SupportsAtomicPartitionManagement, SupportsPartitionManagement}
+
+/**
+ * Physical plan node for dropping partitions of table.
+ */
+case class AlterTableDropPartitionExec(
+    table: SupportsPartitionManagement,
+    partIdents: Seq[InternalRow],
+    ignoreIfNotExists: Boolean) extends V2CommandExec {
+  import DataSourceV2Implicits._
+
+  override def output: Seq[Attribute] = Seq.empty
+
+  override protected def run(): Seq[InternalRow] = {
+    val notExistsPartIdents = partIdents.filterNot(table.partitionExists)

Review comment:
       ```suggestion
       val (existsPartitions, notExistsPartIdents) = partIdents.partition(table.partitionExists)
   ```

##########
File path: sql/catalyst/src/main/scala/org/apache/spark/sql/execution/datasources/v2/DataSourceV2Implicits.scala
##########
@@ -62,4 +85,42 @@ object DataSourceV2Implicits {
       new CaseInsensitiveStringMap(options.asJava)
     }
   }
+
+  implicit class TablePartitionSpecHelper(partSpec: TablePartitionSpec) {
+    def asPartitionIdentifier(partSchema: StructType): InternalRow = {
+      val conflictKeys = partSpec.keys.toSeq.diff(partSchema.map(_.name))
+      if (conflictKeys.nonEmpty) {
+        throw new AnalysisException(
+          s"Partition key ${conflictKeys.mkString(",")} not exists")

Review comment:
       ```suggestion
           throw new AnalysisException(s"Partition key ${conflictKeys.mkString(",")} not exists")
   ```

##########
File path: sql/catalyst/src/main/scala/org/apache/spark/sql/execution/datasources/v2/DataSourceV2Implicits.scala
##########
@@ -62,4 +85,42 @@ object DataSourceV2Implicits {
       new CaseInsensitiveStringMap(options.asJava)
     }
   }
+
+  implicit class TablePartitionSpecHelper(partSpec: TablePartitionSpec) {
+    def asPartitionIdentifier(partSchema: StructType): InternalRow = {
+      val conflictKeys = partSpec.keys.toSeq.diff(partSchema.map(_.name))
+      if (conflictKeys.nonEmpty) {
+        throw new AnalysisException(
+          s"Partition key ${conflictKeys.mkString(",")} not exists")
+      }
+
+      val partValues = partSchema.map { part =>
+        val partValue = partSpec.get(part.name).orNull
+        if (partValue == null) {
+          null
+        } else {
+          part.dataType match {

Review comment:
       How about other types like `DateType`. If you are not going to support them in this PR, please, open an JIRA and add `TODO` here.

##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/v2/AlterTableDropPartitionExec.scala
##########
@@ -0,0 +1,57 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.datasources.v2
+
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.analysis.NoSuchPartitionsException
+import org.apache.spark.sql.catalyst.expressions.Attribute
+import org.apache.spark.sql.connector.catalog.{SupportsAtomicPartitionManagement, SupportsPartitionManagement}
+
+/**
+ * Physical plan node for dropping partitions of table.
+ */
+case class AlterTableDropPartitionExec(
+    table: SupportsPartitionManagement,
+    partIdents: Seq[InternalRow],
+    ignoreIfNotExists: Boolean) extends V2CommandExec {
+  import DataSourceV2Implicits._
+
+  override def output: Seq[Attribute] = Seq.empty
+
+  override protected def run(): Seq[InternalRow] = {
+    val notExistsPartIdents = partIdents.filterNot(table.partitionExists)
+    if (notExistsPartIdents.nonEmpty && !ignoreIfNotExists) {
+      throw new NoSuchPartitionsException(
+        table.name(), notExistsPartIdents, table.partitionSchema())
+    }
+
+    val existsPartitions = partIdents.filterNot(notExistsPartIdents.contains)
+    existsPartitions match {
+      case Seq() => // Nothing will be done
+      case Seq(partIdent) =>
+        table.dropPartition(partIdent)
+      case Seq(_ *) if table.isInstanceOf[SupportsAtomicPartitionManagement] =>
+        table.asAtomicPartitionable
+          .dropPartitions(existsPartitions.toArray)

Review comment:
       nit:
   ```suggestion
           table.asAtomicPartitionable.dropPartitions(existsPartitions.toArray)
   ```

##########
File path: sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/v2/AlterTableAddPartitionExec.scala
##########
@@ -0,0 +1,62 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *    http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.sql.execution.datasources.v2
+
+import scala.collection.JavaConverters._
+
+import org.apache.spark.sql.catalyst.InternalRow
+import org.apache.spark.sql.catalyst.analysis.PartitionsAlreadyExistException
+import org.apache.spark.sql.catalyst.expressions.Attribute
+import org.apache.spark.sql.connector.catalog.{SupportsAtomicPartitionManagement, SupportsPartitionManagement}
+
+/**
+ * Physical plan node for adding partitions of table.
+ */
+case class AlterTableAddPartitionExec(
+    table: SupportsPartitionManagement,
+    partitions: Seq[(InternalRow, Map[String, String])],
+    ignoreIfExists: Boolean) extends V2CommandExec {
+  import DataSourceV2Implicits._
+
+  override def output: Seq[Attribute] = Seq.empty
+
+  override protected def run(): Seq[InternalRow] = {
+    val existsPartIdents = partitions.map(_._1).filter(table.partitionExists)
+    if (existsPartIdents.nonEmpty && !ignoreIfExists) {
+      throw new PartitionsAlreadyExistException(
+        table.name(), existsPartIdents, table.partitionSchema())
+    }
+
+    val notExistsPartitions =
+      partitions.filterNot(part => existsPartIdents.contains(part._1))
+    notExistsPartitions match {
+      case Seq() => // Nothing will be done
+      case Seq((partIdent, properties)) =>
+        table.createPartition(partIdent, properties.asJava)
+      case Seq(_ *) if table.isInstanceOf[SupportsAtomicPartitionManagement] =>
+        table.asAtomicPartitionable
+          .createPartitions(
+            notExistsPartitions.map(_._1).toArray,

Review comment:
       Is it possible to traverse `notExistsPartitions ` only once like:
   ```Scala
   val (idents, props) = notExistsPartitions.unzip
   ```

##########
File path: sql/core/src/test/scala/org/apache/spark/sql/connector/DataSourceV2SQLSuite.scala
##########
@@ -2066,13 +2069,71 @@ class DataSourceV2SQLSuite
   }
 
   test("ALTER TABLE ADD PARTITION") {
-    val t = "testcat.ns1.ns2.tbl"
+    val t = "testpart.ns1.ns2.tbl"
     withTable(t) {
       spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
-      val e = intercept[AnalysisException] {
-        sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
-      }
-      assert(e.message.contains("ALTER TABLE ADD PARTITION is only supported with v1 tables"))
+      spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))

Review comment:
       ```suggestion
         val partMetadata = partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
   ```

##########
File path: sql/core/src/test/scala/org/apache/spark/sql/connector/DataSourceV2SQLSuite.scala
##########
@@ -2066,13 +2069,71 @@ class DataSourceV2SQLSuite
   }
 
   test("ALTER TABLE ADD PARTITION") {
-    val t = "testcat.ns1.ns2.tbl"
+    val t = "testpart.ns1.ns2.tbl"
     withTable(t) {
       spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
-      val e = intercept[AnalysisException] {
-        sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
-      }
-      assert(e.message.contains("ALTER TABLE ADD PARTITION is only supported with v1 tables"))
+      spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      partTable.clearPartitions()
+    }
+  }
+
+  test("ALTER TABLE ADD PARTITIONS") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(
+        s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc' PARTITION (id=2) LOCATION 'loc1'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(2))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      val partMetadata1 =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(2)))
+      assert(partMetadata1.containsKey("location"))
+      assert(partMetadata1.get("location") == "loc1")
+
+      partTable.clearPartitions()
+    }
+  }
+
+  test("ALTER TABLE ADD PARTITIONS: partition already exists") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(
+        s"ALTER TABLE $t ADD PARTITION (id=2) LOCATION 'loc1'")
+
+      assertThrows[PartitionsAlreadyExistException](
+        spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'" +
+          s" PARTITION (id=2) LOCATION 'loc1'"))

Review comment:
       `s` is not needed here.

##########
File path: sql/core/src/test/scala/org/apache/spark/sql/connector/DataSourceV2SQLSuite.scala
##########
@@ -2066,13 +2069,71 @@ class DataSourceV2SQLSuite
   }
 
   test("ALTER TABLE ADD PARTITION") {
-    val t = "testcat.ns1.ns2.tbl"
+    val t = "testpart.ns1.ns2.tbl"
     withTable(t) {
       spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
-      val e = intercept[AnalysisException] {
-        sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
-      }
-      assert(e.message.contains("ALTER TABLE ADD PARTITION is only supported with v1 tables"))
+      spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      partTable.clearPartitions()
+    }
+  }
+
+  test("ALTER TABLE ADD PARTITIONS") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(
+        s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc' PARTITION (id=2) LOCATION 'loc1'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(2))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      val partMetadata1 =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(2)))
+      assert(partMetadata1.containsKey("location"))
+      assert(partMetadata1.get("location") == "loc1")
+
+      partTable.clearPartitions()
+    }
+  }
+
+  test("ALTER TABLE ADD PARTITIONS: partition already exists") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(
+        s"ALTER TABLE $t ADD PARTITION (id=2) LOCATION 'loc1'")
+
+      assertThrows[PartitionsAlreadyExistException](
+        spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'" +
+          s" PARTITION (id=2) LOCATION 'loc1'"))
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(!partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+
+      spark.sql(s"ALTER TABLE $t ADD IF NOT EXISTS PARTITION (id=1) LOCATION 'loc'" +
+        s" PARTITION (id=2) LOCATION 'loc1'")

Review comment:
       remove `s`

##########
File path: sql/core/src/test/scala/org/apache/spark/sql/connector/DataSourceV2SQLSuite.scala
##########
@@ -2066,13 +2069,71 @@ class DataSourceV2SQLSuite
   }
 
   test("ALTER TABLE ADD PARTITION") {
-    val t = "testcat.ns1.ns2.tbl"
+    val t = "testpart.ns1.ns2.tbl"
     withTable(t) {
       spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
-      val e = intercept[AnalysisException] {
-        sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
-      }
-      assert(e.message.contains("ALTER TABLE ADD PARTITION is only supported with v1 tables"))
+      spark.sql(s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      partTable.clearPartitions()
+    }
+  }
+
+  test("ALTER TABLE ADD PARTITIONS") {
+    val t = "testpart.ns1.ns2.tbl"
+    withTable(t) {
+      spark.sql(s"CREATE TABLE $t (id bigint, data string) USING foo PARTITIONED BY (id)")
+      spark.sql(
+        s"ALTER TABLE $t ADD PARTITION (id=1) LOCATION 'loc' PARTITION (id=2) LOCATION 'loc1'")
+
+      val partTable = catalog("testpart").asTableCatalog
+        .loadTable(Identifier.of(Array("ns1", "ns2"), "tbl")).asInstanceOf[InMemoryPartitionTable]
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(1))))
+      assert(partTable.partitionExists(InternalRow.fromSeq(Seq(2))))
+
+      val partMetadata =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(1)))
+      assert(partMetadata.containsKey("location"))
+      assert(partMetadata.get("location") == "loc")
+
+      val partMetadata1 =
+        partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(2)))

Review comment:
       ```suggestion
         val partMetadata1 = partTable.loadPartitionMetadata(InternalRow.fromSeq(Seq(2)))
   ```




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org