Merge pull request #14 from jmalkin/ci_workflow

Ci workflow

.github/workflows/ci.yaml

@@ -0,0 +1,70 @@
+name: Continuous Integration
+
+on:
+  push:
+    paths-ignore: [ '**/*.html', '**/*.md', '**/*.txt', '**/*.xml', '**/*.yaml', '**/*.yml', '**/LICENSE', '**/NOTICE' ]
+    branches: [ 'main', '[0-9]+.[0-9]+.[Xx]' ]
+  pull_request:
+    paths-ignore: [ '**/*.html', '**/*.md', '**/*.txt', '**/*.xml', '**/*.yaml', '**/*.yml', '**/LICENSE', '**/NOTICE' ]
+    # The branches below must be a subset of the branches above
+    branches: [ 'main', '[0-9]+.[0-9]+.[Xx]' ]
+  workflow_dispatch:
+
+#env:
+#  MAVEN_OPTS: -Xmx4g -Xms1g
+
+jobs:
+  build:
+    name: Build and Test
+    runs-on: ubuntu-latest
+
+    strategy:
+      fail-fast: false
+      matrix:
+        jdk: [ 8, 11, 17 ]
+        spark: [ 3.4.4, 3.5.4 ]
+
+    env:
+      JDK_VERSION: ${{ matrix.jdk }}
+      SPARK_VERSION: ${{ matrix.spark }}
+
+    steps:
+      - name: "Checkout ${{ github.ref }} ( ${{ github.sha }} )"
+        uses: actions/checkout@v4
+        with:
+          persist-credentials: false
+
+      - name: Cache local Maven repository
+        uses: actions/cache@v4
+        with:
+          path: ~/.m2/repository
+          key: build-${{ runner.os }}-jdk-${{ matrix.jdk }}-spark-${{ matrix.spark }}-${{ hashFiles('**/pom.xml') }}
+          restore-keys: build-${{ runner.os }}-jdk-${{matrix.jdk}}-spark-${{ matrix.spark }}-maven-
+
+      - name: Setup JDK
+        uses: actions/setup-java@v4
+        with:
+          java-version: ${{ matrix.jdk }}
+          distribution: 'temurin'
+          java-package: jdk
+          architecture: x64
+          cache: sbt
+
+      - name: Setup SBT
+        uses: sbt/setup-sbt@v1
+
+      - name: Echo Java Version
+        run: >
+          java -version
+
+      - name: Build and test
+        run: >
+          sbt --batch clean test
+
+# Architecture options: x86, x64, armv7, aarch64, ppc64le
+# setup-java@v4 has a "with cache" option
+# Lifecycle: validate, compile, test, package, verify, install, deploy
+# -B batch mode, never stops for user input
+# -V show Version without stopping
+# -X debug mode
+# -q quiet, only show errors

build.sbt

@@ -24,28 +24,34 @@ description := "The Apache DataSketches package for Spark"
 
 licenses += ("Apache-2.0", url("http://www.apache.org/licenses/LICENSE-2.0"))
 
+val sparkVersion = settingKey[String]("The version of Spark")
+sparkVersion := sys.env.getOrElse("SPARK_VERSION", "3.5.4")
+
 // determine our java version
 val jvmVersionString = settingKey[String]("The JVM version")
 jvmVersionString := sys.props("java.version")
 
+// Map of JVM version prefix to:
+// (JVM major version, datasketches-java version)
+val jvmVersionMap = Map(
+  "21" -> ("21", "8.0.0"),
+  "17" -> ("17", "7.0.1"),
+  "11" -> ("11", "6.2.0"),
+  "8"  -> ("8",  "6.2.0"),
+  "1.8" -> ("8", "6.2.0")
+)
+
+// determine the JVM major verison (default: 11)
 val jvmVersion = settingKey[String]("The JVM major version")
-jvmVersion := {
-  val version = jvmVersionString.value
-  if (version.startsWith("21")) "21"
-  else if (version.startsWith("17")) "17"
-  else if (version.startsWith("11")) "11"
-  else "8"
-}
+jvmVersion := jvmVersionMap.collectFirst {
+  case (prefix, (major, _)) if jvmVersionString.value.startsWith(prefix) => major
+}.getOrElse("11")
 
+// look up the associated datasketches-java version
 val dsJavaVersion = settingKey[String]("The DataSketches Java version")
-dsJavaVersion := {
-  if (jvmVersion.value == "11") "6.2.0"
-  else if (jvmVersion.value == "17") "7.0.1"
-  else if (jvmVersion.value == "21") "8.0.0"
-  else "6.2.0"
-}
+dsJavaVersion := jvmVersionMap.get(jvmVersion.value).map(_._2).get
+
 
-// these do not impact code generation in spark
 javacOptions ++= Seq("-source", jvmVersion.value, "-target", jvmVersion.value)
 scalacOptions ++= Seq("-encoding", "UTF-8", "-release", jvmVersion.value)
 Test / javacOptions ++= Seq("-source", jvmVersion.value, "-target", jvmVersion.value)
@@ -54,7 +60,7 @@ Test / scalacOptions ++= Seq("-encoding", "UTF-8", "-release", jvmVersion.value)
 libraryDependencies ++= Seq(
   "org.apache.datasketches" % "datasketches-java" % dsJavaVersion.value % "compile",
   "org.scala-lang" % "scala-library" % "2.12.6",
-  "org.apache.spark" %% "spark-sql" % "3.5.4" % "provided",
+  "org.apache.spark" %% "spark-sql" % sparkVersion.value % "provided",
   "org.scalatest" %% "scalatest" % "3.2.19" % "test",
   "org.scalatestplus" %% "junit-4-13" % "3.2.19.0" % "test"
 )