Added Databricks cluster deployment to the user guide

.gitignore

@@ -7,6 +7,7 @@ project/.gnupg/local*
 project/target
 target/
 .history
+.bsp
 
 # Java
 *.class

.travis.yml

@@ -9,29 +9,22 @@ addons:
     packages:
       - shellcheck
 
-before_install:
-  - git fetch --tags
-  - docker pull exasol/docker-db:7.0.2
-  - docker pull exasol/docker-db:6.2.9-d1
-  - docker network create -d bridge --subnet 192.168.0.0/24 --gateway 192.168.0.1 dockernet
-
-matrix:
-  include:
-    - jdk: openjdk8
-      scala: 2.11.12
-      env: SPARK_VERSION="2.4.5" EXASOL_DOCKER_VERSION="6.2.9-d1"
+jdk:
+  - openjdk11
 
-    - jdk: openjdk8
-      scala: 2.11.12
-      env: SPARK_VERSION="2.4.5" EXASOL_DOCKER_VERSION="7.0.2"
+scala:
+  - 2.12.12
 
-    - jdk: openjdk8
-      scala: 2.12.12
-      env: SPARK_VERSION="2.4.5" EXASOL_DOCKER_VERSION="6.2.9-d1"
+env:
+  - SPARK_VERSION="2.4.5" EXASOL_DOCKER_VERSION="6.2.12-d1"
+  - SPARK_VERSION="2.4.5" EXASOL_DOCKER_VERSION="7.0.4"
+  - SPARK_VERSION="3.0.1" EXASOL_DOCKER_VERSION="6.2.12-d1"
+  - SPARK_VERSION="3.0.1" EXASOL_DOCKER_VERSION="7.0.4"
 
-    - jdk: openjdk8
-      scala: 2.12.12
-      env: SPARK_VERSION="2.4.5" EXASOL_DOCKER_VERSION="7.0.2"
+before_install:
+  - git fetch --tags
+  - docker pull "exasol/docker-db:$EXASOL_DOCKER_VERSION"
+  - docker network create -d bridge --subnet 192.168.0.0/24 --gateway 192.168.0.1 dockernet
 
 script:
   - travis_wait 30 ./scripts/ci.sh

doc/changes/changelog.md

@@ -1,5 +1,6 @@
 # Releases
 
+* [1.0.0](changes_1.0.0.md)
 * [0.3.2](changes_0.3.2.md)
 * [0.3.1](changes_0.3.1.md)
 * [0.3.0](changes_0.3.0.md)

doc/changes/changes_1.0.0.md

@@ -0,0 +1,33 @@
+# Spark Exasol Connector 1.0.0, released 2020-12-DD
+
+## Features / Improvements
+
+
+## Documentation
+
+* #85: Updated documentation with configuration for the Databricks cluster (PR #86)
+
+## Dependency Updates
+
+### Runtime Dependency Updates
+
+* Updated to `com.exasol:exasol-jdbc:7.0.4` (was `7.0.0`)
+* Updated to `org.apache.spark:spark-core:3.0.1` (was `2.4.5`)
+* Updated to `org.apache.spark:spark-sql:3.0.1` (was `2.4.5`)
+
+### Test Dependency Updates
+
+* Updated to `org.scalatest:scalatest:3.2.2` (was `3.2.2`)
+* Updated to `org.testcontainers:jdbc:1.15.0` (was `1.14.3`)
+* Updated to `com.holdenkarau:spark-testing-base:3.0.1_1.0.0` (was `2.4.5_0.14.0)
+* Updated to `org.mockito:mockito-core:3.6.28` (was `3.5.13`)
+* Updated to `com.dimafeng:testcontainers-scala:0.38.7` (was `0.38.4`)
+
+### Plugin Updates
+
+* Updated to `sbt.version:1.4.4` (was `1.3.13`)
+* Updated to `org.wartremover:sbt-wartremover:2.4.13` (was `2.4.10`)
+* Updated to `org.wartremover:sbt-wartremover-contrib:1.3.11` (was `1.3.8`)
+* Updated to `com.jsuereth:sbt-pgp:2.0.2` (was `2.0.1`)
+* Updated to `org.xerial.sbt:sbt-sonatype:3.9.5` (was `3.9.4`)
+* Removed `io.get-coursier:sbt-coursier`

doc/user_guide/user_guide.md

@@ -86,6 +86,29 @@ libraryDependencies += "com.exasol" % "spark-connector" %% "<VERSION>"
 Similar to the Java dependency, we add the resolver to the Exasol Artifactory so
 that the Exasol JDBC driver can be found. 
 
+### Spark Exasol Connector as Databricks Cluster Dependency
+
+Similar to using maven, you should provide maven artifact coordinates to the
+[Databricks Workspace Library][databricks-install].
+
+[databricks-install]: https://docs.databricks.com/libraries/workspace-libraries.html#maven-libraries
+
+<img alt="databricks-deployment" src="../images/databricks-deployment.png"/>
+
+Go to your cluster, then to `Libraries`, and click `Install New`:
+
+- Select Maven as a Library Source.
+- In the Coordinate field, enter artifact coordinates
+  `com.exasol:spark-connector_2.12:<VERSION>`. Please notice that we use the
+  Scala version 2.12, change it to 2.11 if your Databricks Runtime version
+  requires it.
+- In the Repository field, enter the Exasol Artifactory
+  `https://maven.exasol.com/artifactory/exasol-releases`.
+- Click Install.
+
+Please change the `<VERSION>` to one of the latest Spark Exasol Connector
+releases.
+
 ### Spark Exasol Connector With Spark Shell
 
 You can also integrate the Spark Exasol Connector to the Spark Shell. Provide

project/Dependencies.scala

@@ -6,19 +6,19 @@ import sbt._
 object Dependencies {
 
   // Versions
-  private val DefaultSparkVersion = "2.4.5"
-  private val ExasolJdbcVersion = "7.0.0"
+  private val DefaultSparkVersion = "3.0.1"
+  private val ExasolJdbcVersion = "7.0.4"
 
-  private val ScalaTestVersion = "3.2.2"
+  private val ScalaTestVersion = "3.2.3"
   private val ScalaTestMockitoVersion = "1.0.0-M2"
-  private val MockitoVersion = "3.5.13"
-  private val ContainersJdbcVersion = "1.14.3"
-  private val ContainersScalaVersion = "0.38.4"
+  private val MockitoVersion = "3.6.28"
+  private val ContainersJdbcVersion = "1.15.0"
+  private val ContainersScalaVersion = "0.38.7"
 
   private val sparkCurrentVersion =
     sys.env.getOrElse("SPARK_VERSION", DefaultSparkVersion)
 
-  private val SparkTestingBaseVersion = s"${sparkCurrentVersion}_0.14.0"
+  private val SparkTestingBaseVersion = s"${sparkCurrentVersion}_1.0.0"
 
   val Resolvers: Seq[Resolver] = Seq(
     "Exasol Releases" at "https://maven.exasol.com/artifactory/exasol-releases"

project/Settings.scala

@@ -19,6 +19,7 @@ object Settings {
     buildSettings(scalaVersion) ++
       miscSettings ++
       assemblySettings ++
+      apiDocSettings ++
       scalaStyleSettings ++
       Publishing.publishSettings()
 
@@ -70,6 +71,31 @@ object Settings {
     }
   )
 
+  def apiDocSettings(): Seq[Setting[_]] = Seq(
+    autoAPIMappings := true,
+    apiMappings ++= scalaInstance.value.libraryJars.collect {
+      case file if file.getName.startsWith("scala-library") && file.getName.endsWith(".jar") =>
+        file -> url(s"http://www.scala-lang.org/api/${scalaVersion.value}/")
+    }.toMap ++
+      // Since Java 9+ introduced modules, API links changed, update these
+      // links based on used Java modules.
+      Map(
+        file("/modules/java.sql") -> url(
+          "https://docs.oracle.com/en/java/javase/11/docs/api/java.sql"
+        )
+      ),
+    // Override doc task in 2.11.x versions since linking external Java
+    // 11+ classes does not work.
+    (Compile / doc) := Def.taskDyn {
+      val docTaskValue = (Compile / doc).taskValue
+      if (scalaBinaryVersion.value == "2.11") {
+        (Compile / doc / target).toTask
+      } else {
+        Def.task(docTaskValue.value)
+      }
+    }.value
+  )
+
   def scalaStyleSettings(): Seq[Setting[_]] = {
     // Creates a Scalastyle task that runs with tests
     lazy val mainScalastyle = taskKey[Unit]("mainScalastyle")

project/build.properties

@@ -1 +1 @@
-sbt.version=1.3.13
+sbt.version=1.4.4

project/plugins.sbt

@@ -1,14 +1,10 @@
-// Adds a pure Scala artifact fetcher `coursier`
-// https://github.com/coursier/coursier
-addSbtPlugin("io.get-coursier" % "sbt-coursier" % "1.0.3")
-
 // Adds a `wartremover` a flexible Scala code linting tool
 // http://github.com/puffnfresh/wartremover
-addSbtPlugin("org.wartremover" % "sbt-wartremover" % "2.4.10")
+addSbtPlugin("org.wartremover" % "sbt-wartremover" % "2.4.13")
 
 // Adds Contrib Warts
 // http://github.com/wartremover/wartremover-contrib/
-addSbtPlugin("org.wartremover" % "sbt-wartremover-contrib" % "1.3.8")
+addSbtPlugin("org.wartremover" % "sbt-wartremover-contrib" % "1.3.11")
 
 // Adds a `assembly` task to create a fat JAR with all of its dependencies
 // https://github.com/sbt/sbt-assembly
@@ -45,11 +41,11 @@ addSbtPlugin("net.virtual-void" % "sbt-dependency-graph" % "0.9.2")
 
 // Adds a `sonatype` release tasks
 // https://github.com/xerial/sbt-sonatype
-addSbtPlugin("org.xerial.sbt" % "sbt-sonatype" % "3.9.4")
+addSbtPlugin("org.xerial.sbt" % "sbt-sonatype" % "3.9.5")
 
 // Adds a `gnu-pgp` plugin
 // https://github.com/sbt/sbt-pgp
-addSbtPlugin("com.jsuereth" % "sbt-pgp" % "2.0.1")
+addSbtPlugin("com.jsuereth" % "sbt-pgp" % "2.0.2")
 
 // Adds a `git` plugin
 // https://github.com/sbt/sbt-git

sbtx

@@ -34,10 +34,10 @@
 
 set -o pipefail
 
-declare -r sbt_release_version="1.3.13"
-declare -r sbt_unreleased_version="1.4.0-M1"
+declare -r sbt_release_version="1.4.4"
+declare -r sbt_unreleased_version="1.4.4"
 
-declare -r latest_213="2.13.3"
+declare -r latest_213="2.13.4"
 declare -r latest_212="2.12.12"
 declare -r latest_211="2.11.12"
 declare -r latest_210="2.10.7"
@@ -471,7 +471,7 @@ process_args() {
       -trace)       require_arg integer "$1" "$2" && trace_level="$2" && shift 2 ;;
       -debug-inc)   addJava "-Dxsbt.inc.debug=true" && shift ;;
 
-      -no-colors)   addJava "-Dsbt.log.noformat=true" && shift ;;
+      -no-colors)   addJava "-Dsbt.log.noformat=true" && addJava "-Dsbt.color=false" && shift ;;
       -sbt-create)  sbt_create=true && shift ;;
       -sbt-dir)     require_arg path "$1" "$2" && sbt_dir="$2" && shift 2 ;;
       -sbt-boot)    require_arg path "$1" "$2" && addJava "-Dsbt.boot.directory=$2" && shift 2 ;;

scripts/ci.sh

@@ -10,7 +10,7 @@ BASE_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )"/.. && pwd )"
 cd "$BASE_DIR"
 
 MAIN_SCALA_VERSION=2.12.12
-MAIN_SPARK_VERSION=2.4.5
+MAIN_SPARK_VERSION=3.0.1
 
 if [[ -z "${TRAVIS_SCALA_VERSION:-}" ]]; then
   echo "Environment variable TRAVIS_SCALA_VERSION is not set"
@@ -104,7 +104,7 @@ run_dependency_info () {
   ./sbtx \
     -Dspark.currentVersion=$SPARK_VERSION \
     ++$TRAVIS_SCALA_VERSION \
-    dependencyUpdates pluginUpdates coursierDependencyTree
+    dependencyUpdates pluginUpdates dependencyTree
 }
 
 run_shell_check () {

src/it/java/org/testcontainers/containers/ExasolDockerContainer.java

@@ -10,7 +10,7 @@
 
 public class ExasolDockerContainer<SELF extends ExasolDockerContainer<SELF>>
   extends JdbcDatabaseContainer<SELF> {
-  private static final String DEFAULT_EXASOL_VERSION = "7.0.2";
+  private static final String DEFAULT_EXASOL_VERSION = "7.0.4";
   // wait for 5 minutes to startup
   private static final Integer EXASOL_STARTUP_TIME = 15 * 60;
   private String username = "sys";

src/main/scala/com/exasol/spark/rdd/ExasolRDD.scala

@@ -13,11 +13,11 @@ import org.apache.spark.rdd.RDD
 import org.apache.spark.scheduler.SparkListener
 import org.apache.spark.scheduler.SparkListenerApplicationEnd
 import org.apache.spark.sql.Row
+import org.apache.spark.sql.execution.datasources.jdbc.JdbcUtils
 import org.apache.spark.sql.types.StructType
 
 import com.exasol.jdbc.EXAConnection
 import com.exasol.jdbc.EXAResultSet
-import com.exasol.spark.util.Converter
 import com.exasol.spark.util.ExasolConnectionManager
 
 /**
@@ -155,7 +155,7 @@ class ExasolRDD(
     conn = manager.subConnection(partition.connectionUrl)
     resultSet = conn.DescribeResult(subHandle)
 
-    Converter.resultSetToRows(resultSet, querySchema)
+    JdbcUtils.resultSetToRows(resultSet, querySchema)
   }
   // scalastyle:on null return
 

src/main/scala/com/exasol/spark/util/Converter.scala

@@ -6,7 +6,6 @@ import java.sql.ResultSet
 import org.apache.spark.internal.Logging
 import org.apache.spark.sql.Row
 import org.apache.spark.sql.catalyst.InternalRow
-import org.apache.spark.sql.catalyst.encoders.RowEncoder
 import org.apache.spark.sql.catalyst.expressions.SpecificInternalRow
 import org.apache.spark.sql.catalyst.util.DateTimeUtils
 import org.apache.spark.sql.types._
@@ -21,16 +20,6 @@ import org.apache.spark.unsafe.types.UTF8String
  */
 object Converter extends Logging {
 
-  /**
-   * Converts a [[java.sql.ResultSet]] into an iterator of
-   * [[org.apache.spark.sql.Row]]-s.
-   */
-  def resultSetToRows(resultSet: ResultSet, schema: StructType): Iterator[Row] = {
-    val encoder = RowEncoder(schema).resolveAndBind()
-    val internalRows = resultSetToSparkInternalRows(resultSet, schema)
-    internalRows.map(encoder.fromRow)
-  }
-
   @SuppressWarnings(Array("org.wartremover.warts.AsInstanceOf"))
   def resultSetToSparkInternalRows(
     resultSet: ResultSet,

src/test/scala/com/exasol/spark/util/FiltersSuite.scala

@@ -41,7 +41,7 @@ class FiltersSuite extends AnyFunSuite with Matchers {
         |    str_col <= '123'
         |AND int_col > 42
         |AND (in_col NOT IN (1,2,3))
-    """.stripMargin.lines.mkString(" ").trim
+    """.stripMargin.replaceAll("\\s+", " ").trim()
 
     assert(createWhereClause(testSchema, filters) === expected)
   }
@@ -78,7 +78,7 @@ class FiltersSuite extends AnyFunSuite with Matchers {
         |AND (str_col LIKE '%inside%')
         |AND (str_col LIKE 'prefix%')
         |AND (in_col IN (1,2,3))
-      """.stripMargin.lines.mkString(" ").trim
+      """.stripMargin.replaceAll("\\s+", " ").trim()
 
     assert(createWhereClause(testSchema, filters) === expected)
   }
@@ -108,7 +108,7 @@ class FiltersSuite extends AnyFunSuite with Matchers {
         |AND double_col = 100.0
         |AND date_col = date '2018-01-01'
         |AND datetime_col = timestamp '2018-01-01 00:00:59.123'
-      """.stripMargin.lines.mkString(" ").trim
+      """.stripMargin.replaceAll("\\s+", " ").trim()
 
     assert(createWhereClause(testSchema, filters) === expected)
   }
@@ -126,7 +126,7 @@ class FiltersSuite extends AnyFunSuite with Matchers {
         |((str_col = 'abc') OR (int_col = 123))
         |AND (((NOT ((int_col IS NULL)))) OR ((str_col IS NOT NULL)))
         |AND ((str_col = 'xyz') OR (((float_col = 3.14) AND (int_col != 3))))
-    """.stripMargin.lines.mkString(" ").trim
+      """.stripMargin.replaceAll("\\s+", " ").trim()
 
     assert(createWhereClause(testSchema, filters) === expected)
   }