build.sbt

name := """sf-converter"""
version := "1.0.0"
scalaVersion := "2.11.12"
organization := "fi.seco"

libraryDependencies ++= Seq(
  "ch.qos.logback" % "logback-classic" % "1.2.3",
  "com.typesafe" % "config" % "1.3.1",
  "com.typesafe.play" %% "play-json" % "2.6.8",
  "com.typesafe.scala-logging" %% "scala-logging" % "3.7.2",
  "joda-time" % "joda-time" % "2.9.7",
  "net.sourceforge.htmlcleaner" % "htmlcleaner" % "2.6.1",
  "org.apache.jena" % "jena-core" % "3.4.0" exclude("org.slf4j","slf4j-log4j12") exclude("log4j","log4j") exclude("commons-logging", "commons-logging"),
  "org.apache.jena" % "jena-arq" % "3.4.0" exclude("org.slf4j","slf4j-log4j12") exclude("log4j","log4j") exclude("commons-logging", "commons-logging"),
  "org.clapper" %% "argot" % "1.0.3",
  "org.scala-lang.modules" %% "scala-xml" % "1.0.2",
  "org.scalaj" %% "scalaj-http" % "2.3.0",
  "org.scalatest" % "scalatest_2.11" % "2.2.6" % "test"
)

// Change confDir to test to run tasks with test configuration
val confDir = "main"
lazy val configuration = com.typesafe.config.ConfigFactory.parseFile(new File(s"src/$confDir/resources/application.conf")).resolve

lazy val sourceDataPath = configuration.getString("sourcedata.path")
lazy val outputDataPath = configuration.getString("outputdata.path")
lazy val wwwDataPath = configuration.getString("wwwdata.path")
lazy val outputRdfPath = s"$outputDataPath/rdf"
lazy val tdbDataPath = s"$outputDataPath/jena/tdb"
lazy val luceneDataPath = s"$outputDataPath/jena/lucene"
lazy val backupsPath = s"$outputDataPath/jena/backups"
lazy val jenaPath = configuration.getString("jena.path")
lazy val fusekiPath = configuration.getString("fuseki.path")
lazy val sourceDataUrl = configuration.getString("sourcedata.url")
lazy val assemblerPath = s"${configuration.getString("app.path")}/src/main/resources/jena/assembler.ttl"

val makeDataDir = taskKey[Unit]("Build data directory tree")
val downloadSourceData = taskKey[Unit]("Download source data")
val loadDataToTripleStore = taskKey[Unit]("Create TDB and Lucene indices")
val publishData = taskKey[Unit]("Publish RDF and source XML")
val publishSourceData = taskKey[Unit]("Publish only source XML")
val runSparqlServer = taskKey[Unit]("Run SPARQL server locally")

makeDataDir := {
  import sys.process._
  val sourceDatasets = List("sd", "asd", "kho", "kko", "ho", "hao", "misc", "release")
  sourceDatasets.foreach(
    ds => {
      println(s"Create directory: $sourceDataPath/$ds")
      (s"mkdir -p $sourceDataPath/$ds").!
    }
  )
  val rdfDatasets = List("sd", "asd", "kho", "kko", "ho", "hao", "misc", "voc", "schema")
  rdfDatasets.foreach{
    ds => {
      println(s"Create directory: $outputRdfPath/$ds")
      (s"mkdir -p $outputRdfPath/$ds").!
    }
  }
  rdfDatasets.foreach{
    ds => {
      println(s"Create directory: $outputRdfPath/bulk/$ds")
      (s"mkdir -p $outputRdfPath/bulk/$ds").!
    }
  }
  println(s"Create directories:", wwwDataPath, tdbDataPath, luceneDataPath, backupsPath)
  (s"mkdir -p $wwwDataPath $tdbDataPath $luceneDataPath $backupsPath").!
  println(s"Succesfully created all the directories")
}
downloadSourceData := {
  import sys.process._
  (s"./src/main/resources/shell/download_source_data.sh $sourceDataPath $sourceDataPath/zip $sourceDataPath/tmp $sourceDataUrl").!
}
loadDataToTripleStore := {
  import sys.process._
  (s"./src/main/resources/shell/load_data_to_triple_store.sh $outputRdfPath $tdbDataPath $luceneDataPath $jenaPath $fusekiPath $assemblerPath $backupsPath").!
}
publishData := {
  import sys.process._
  (s"./src/main/resources/shell/publish.sh $wwwDataPath $sourceDataPath $outputRdfPath $tdbDataPath $luceneDataPath").!
}
publishSourceData := {
  import sys.process._
  (s"./src/main/resources/shell/publish_source_data.sh $wwwDataPath $sourceDataPath").!
}
runSparqlServer := {
   import sys.process._
   println(s"SPARQL SERVER: $fusekiPath/fuseki-server")
   println(s"ASSEMBLER PATH: $assemblerPath")
   (s"$fusekiPath/fuseki-server --config=$assemblerPath").!
}

parallelExecution in Test := false
fork in run := true
fork in Test := true