added spark connect example which does not yet work with kotlin-spark…

…-api

buildSrc/src/main/kotlin/Dependencies.kt

@@ -4,6 +4,8 @@ object Dependencies : Dsl<Dependencies> {
     inline val scalaLibrary get() = "org.scala-lang:scala-library:${Versions.scala}"
     inline val kotlinxHtml get() = "org.jetbrains.kotlinx:kotlinx-html-jvm:${Versions.kotlinxHtml}"
     inline val sparkSql get() = "org.apache.spark:spark-sql_${Versions.scalaCompat}:${Versions.spark}"
+    inline val sparkSqlApi get() = "org.apache.spark:spark-sql-api_${Versions.scalaCompat}:${Versions.spark}"
+    inline val sparkConnectClient get() = "org.apache.spark:spark-connect-client-jvm_${Versions.scalaCompat}:${Versions.spark}"
     inline val sparkMl get() = "org.apache.spark:spark-mllib_${Versions.scalaCompat}:${Versions.spark}"
     inline val sparkStreaming get() = "org.apache.spark:spark-streaming_${Versions.scalaCompat}:${Versions.spark}"
     inline val hadoopClient get() = "org.apache.hadoop:hadoop-client:${Versions.hadoop}"

jupyter/src/main/kotlin/org/jetbrains/kotlinx/spark/api/jupyter/Integration.kt

@@ -162,10 +162,11 @@ abstract class Integration(private val notebook: Notebook, private val options:
         }
 
         beforeCellExecution {
-            if (scalaCompatVersion.toDouble() >= 2.13)
+            if (scalaCompatVersion.toDouble() >= 2.13) {
                 execute("scala.`Console\$`.`MODULE\$`.setOutDirect(System.out)")
-            else
+            } else {
                 execute("""scala.Console.setOut(System.out)""")
+            }
 
             beforeCellExecution()
         }

kotlin-spark-api/build.gradle.kts

@@ -42,7 +42,7 @@ dependencies {
         // https://github.com/FasterXML/jackson-bom/issues/52
         if (Versions.spark == "3.3.1") implementation(jacksonDatabind)
 
-        if (Versions.sparkConnect) TODO("unsupported for now")
+        // if (Versions.sparkConnect) TODO("unsupported for now")
 
         implementation(
             kotlinStdLib,

kotlin-spark-api/src/test/kotlin/org/jetbrains/kotlinx/spark/api/RddTest.kt

@@ -1,14 +1,12 @@
 package org.jetbrains.kotlinx.spark.api
 
 import io.kotest.core.spec.style.ShouldSpec
-import io.kotest.core.spec.style.Test
-import io.kotest.core.test.TestScope
 import io.kotest.matchers.collections.shouldContainAll
 import io.kotest.matchers.shouldBe
 import org.apache.spark.api.java.JavaRDD
-import org.jetbrains.kotlinx.spark.api.tuples.*
+import org.jetbrains.kotlinx.spark.api.tuples.X
+import org.jetbrains.kotlinx.spark.api.tuples.t
 import scala.Tuple2
-import java.io.Serializable
 
 class RddTest : ShouldSpec({
     context("RDD extension functions") {

settings.gradle.kts

@@ -21,9 +21,11 @@ gradleEnterprise {
 val spark: String by settings
 val scala: String by settings
 val skipScalaOnlyDependent: String by settings
+val sparkConnect: String by settings
 System.setProperty("spark", spark)
 System.setProperty("scala", scala)
 System.setProperty("skipScalaOnlyDependent", skipScalaOnlyDependent)
+System.setProperty("sparkConnect", sparkConnect)
 
 val scalaCompat
     get() = scala.substringBeforeLast('.')
@@ -37,6 +39,7 @@ include("scala-tuples-in-kotlin")
 include("kotlin-spark-api")
 include("jupyter")
 include("examples")
+include("spark-connect-examples")
 include("compiler-plugin")
 include("gradle-plugin")
 

spark-connect-examples/build.gradle.kts

@@ -0,0 +1,60 @@
+import org.jetbrains.kotlin.gradle.dsl.JvmTarget
+
+plugins {
+    // Needs to be installed in the local maven repository or have the bootstrap jar on the classpath
+    id("org.jetbrains.kotlinx.spark.api")
+    kotlin("jvm")
+    application
+}
+
+// run with `./gradlew run`
+application {
+    mainClass = "org.jetbrains.kotlinx.spark.examples.MainKt"
+
+    // workaround for java 17
+    applicationDefaultJvmArgs = listOf("--add-opens", "java.base/java.nio=ALL-UNNAMED")
+}
+
+kotlinSparkApi {
+    enabled = true
+    sparkifyAnnotationFqNames = listOf("org.jetbrains.kotlinx.spark.api.plugin.annotations.Sparkify")
+}
+
+group = Versions.groupID
+version = Versions.project
+
+repositories {
+    mavenLocal()
+    mavenCentral()
+}
+
+dependencies {
+    Projects {
+        implementation(
+            // TODO kotlinSparkApi,
+        )
+    }
+
+    Dependencies {
+
+        // IMPORTANT!
+        compileOnly(sparkSqlApi)
+        implementation(sparkConnectClient)
+    }
+}
+
+// spark-connect seems to work well with java 17 as client and java 1.8 as server
+// also set gradle and your project sdk to java 17
+kotlin {
+    jvmToolchain {
+        languageVersion = JavaLanguageVersion.of(17)
+    }
+    compilerOptions {
+        jvmTarget = JvmTarget.JVM_17
+    }
+}
+
+tasks.withType<JavaCompile> {
+    sourceCompatibility = JavaVersion.VERSION_17.toString()
+    targetCompatibility = JavaVersion.VERSION_17.toString()
+}

spark-connect-examples/src/main/kotlin/org/jetbrains/kotlinx/spark/examples/Main.kt

@@ -0,0 +1,27 @@
+package org.jetbrains.kotlinx.spark.examples
+
+import org.apache.spark.sql.SparkSession
+import org.apache.spark.sql.connect.client.REPLClassDirMonitor
+
+// run with `./gradlew run` or set VM options: "--add-opens=java.base/java.nio=ALL-UNNAMED" in the IDE
+fun main() {
+    val spark =
+        SparkSession
+            .builder()
+            .remote("sc://localhost")
+            .create()
+
+    val classFinder = REPLClassDirMonitor("/mnt/data/Projects/kotlin-spark-api/spark-connect-examples/build/classes")
+    spark.registerClassFinder(classFinder)
+    spark.addArtifact("/mnt/data/Projects/kotlin-spark-api/spark-connect-examples/build/libs/spark-connect-examples-2.0.0-SNAPSHOT.jar")
+
+    spark.sql("select 1").show()
+
+    spark.stop()
+}
+
+//@Sparkify
+//data class Person(
+//    val name: String,
+//    val age: Int,
+//)