[GitHub] [spark] dongjoon-hyun commented on a diff in pull request #39933: [SPARK-42377][CONNECT][TESTS] Test framework for Spark Connect Scala Client

["dongjoon-hyun (via GitHub)" <gi...@apache.org>]

dongjoon-hyun commented on code in PR #39933:
URL: https://github.com/apache/spark/pull/39933#discussion_r1103878136


##########
connector/connect/client/jvm/src/test/scala/org/apache/spark/sql/PlanGenerationTestSuite.scala:
##########
@@ -0,0 +1,252 @@
+/*
+ * 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
+
+import java.nio.file.{Files, Path}
+
+import scala.collection.mutable
+import scala.util.{Failure, Success, Try}
+
+import com.google.protobuf.util.JsonFormat
+import io.grpc.inprocess.InProcessChannelBuilder
+import org.scalatest.BeforeAndAfterAll
+import org.scalatest.funsuite.{AnyFunSuite => ConnectFunSuite} // scalastyle:ignore funsuite
+
+import org.apache.spark.connect.proto
+import org.apache.spark.internal.Logging
+import org.apache.spark.sql.{functions => fn}
+import org.apache.spark.sql.connect.client.SparkConnectClient
+import org.apache.spark.sql.types.StructType
+
+// scalastyle:off
+/**
+ * Test the plans generated by the client. This serves two purposes:
+ *
+ *   1. Make sure the generated plan matches our expectations. The generated JSON file can be used
+ *      for this during review.
+ *   1. Make sure the generated plans are stable. Changes to the generated plans should be rare.
+ *      The generated plan is compared to the (previously) generated proto file; the test fails
+ *      when they are different.
+ *
+ * If you need to re-generate the golden files, you need to set the SPARK_GENERATE_GOLDEN_FILES=1
+ * environment variable before running this test, e.g.:
+ * {{{
+ *   SPARK_GENERATE_GOLDEN_FILES=1 build/sbt "connect-client-jvm/testOnly org.apache.spark.sql.PlanGenerationTestSuite"
+ * }}}
+ *
+ * Note that the plan protos are used as the input for the `ProtoToParsedPlanTestSuite` in the
+ * `connector/connect/server` module
+ */
+// scalastyle:on
+class PlanGenerationTestSuite extends ConnectFunSuite with BeforeAndAfterAll with Logging {
+
+  // Borrowed from SparkFunSuite
+  private val regenerateGoldenFiles: Boolean = System.getenv("SPARK_GENERATE_GOLDEN_FILES") == "1"
+
+  // Borrowed from SparkFunSuite
+  private def getWorkspaceFilePath(first: String, more: String*): Path = {
+    if (!(sys.props.contains("spark.test.home") || sys.env.contains("SPARK_HOME"))) {
+      fail("spark.test.home or SPARK_HOME is not set.")
+    }
+    val sparkHome = sys.props.getOrElse("spark.test.home", sys.env("SPARK_HOME"))
+    java.nio.file.Paths.get(sparkHome, first +: more: _*)
+  }
+
+  protected val baseResourcePath: Path = {
+    getWorkspaceFilePath(
+      "connector",
+      "connect",
+      "common",
+      "src",
+      "test",
+      "resources",
+      "query-tests").toAbsolutePath
+  }
+
+  protected val queryFilePath: Path = baseResourcePath.resolve("queries")
+
+  private val printer = JsonFormat.printer()
+
+  private var session: SparkSession = _
+
+  override protected def beforeAll(): Unit = {
+    super.beforeAll()
+    val client = new SparkConnectClient(
+      proto.UserContext.newBuilder().build(),
+      InProcessChannelBuilder.forName("/dev/null").build())
+    val builder = SparkSession.builder().client(client)
+    session = builder.build()
+  }
+
+  override protected def afterAll(): Unit = {
+    session.close()
+    super.afterAll()
+  }
+
+  private def test(name: String)(f: => Dataset[_]): Unit = super.test(name) {
+    val actual = f.plan.getRoot
+    val goldenFile = queryFilePath.resolve(name.replace(' ', '_') + ".proto.bin")
+    Try(readRelation(goldenFile)) match {
+      case Success(expected) if expected == actual =>
+      // Ok!
+      case Success(_) if regenerateGoldenFiles =>
+        logInfo("Rewriting Golden File")
+        writeGoldenFile(goldenFile, actual)
+      case Success(expected) =>
+        fail(s"""
+             |Expected and actual plans do not match:
+             |
+             |=== Expected Plan ===
+             |$expected
+             |
+             |=== Actual Plan ===
+             |$actual
+             |""".stripMargin)
+      case Failure(_) if regenerateGoldenFiles =>
+        logInfo("Writing Golden File")
+        writeGoldenFile(goldenFile, actual)
+      case Failure(_) =>
+        fail(
+          "No golden file found. Please re-run this test with the " +
+            "SPARK_GENERATE_GOLDEN_FILES=1 environment variable set")
+    }
+  }
+
+  private def readRelation(path: Path): proto.Relation = {
+    val input = Files.newInputStream(path)
+    try proto.Relation.parseFrom(input)
+    finally {
+      input.close()
+    }
+  }
+
+  private def writeGoldenFile(path: Path, relation: proto.Relation): Unit = {
+    val output = Files.newOutputStream(path)
+    try relation.writeTo(output)
+    finally {
+      output.close()
+    }
+    // Write the json file for verification.
+    val jsonPath =
+      path.getParent.resolve(path.getFileName.toString.stripSuffix(".proto.bin") + ".json")
+    val writer = Files.newBufferedWriter(jsonPath)
+    try writer.write(printer.print(relation))
+    finally {
+      writer.close()
+    }
+  }
+
+  private val simpleSchema = new StructType()
+    .add("id", "long")
+    .add("a", "int")
+    .add("b", "double")
+
+  private val simpleSchemaString = simpleSchema.catalogString
+
+  // We manually construct a simple empty data frame.
+  private def simple = session.newDataset { builder =>
+    // TODO API is not consistent. Now we have two different ways of working with schemas!
+    builder.getLocalRelationBuilder.setSchema(simpleSchemaString)
+  }
+
+  private def select(cs: Column*): DataFrame = simple.select(cs: _*)
+
+  /* Spark Session API */
+  test("sql") {
+    session.sql("select 1")
+  }
+
+  test("range") {
+    session.range(1, 10, 1, 2)
+  }
+
+  /* Dataset API */
+  test("select") {
+    simple.select(fn.col("id"))
+  }
+
+  test("limit") {
+    simple.limit(10)
+  }
+
+  test("filter") {
+    simple.filter(fn.col("id") === fn.lit(10L))
+  }
+
+  /* Column API */
+  test("column by name") {
+    select(fn.col("b"))
+  }
+
+  test("column add") {
+    select(fn.col("a") + fn.col("b"))
+  }
+
+  test("column alias") {
+    select(fn.col("a").name("b"))
+  }
+
+  test("column equals") {
+    select(fn.col("a") === fn.col("b"))
+  }
+
+  /* Function API */
+  test("function col") {
+    select(fn.col("id"))
+  }
+
+  test("function udf") {
+    // This test might be a bit tricky if different JVM
+    // versions are used to generate the golden files.

Review Comment:
   As mentioned here. Not only JVM, but also Scala seems to be affected.



-- 
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.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

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