Highwheel sbt plugin is an sbt 1.x plugin that allow to run highwheel-modules analysis on scala project as part of the build process run with sbt.
The plugin introduces the following settings:
(highwheelSpecFiles: Seq[File])are the files to the specification files to use for the analysis. Defaults toSeq(baseDirectory / "spec.hwm)"(highwheelAnalysisPaths: Seq[File])specifies the paths to add to the analysis. Defaults toclassDirectory in Compile(highwheelEvidenceLimit: Option[Int])limits the pieces of evidence collected by the plugin to prove that a depdency exists or does not exist. A value ofNoneentails all evidence is collected.
The plugin also introduces the following tasks:
highwheelAnalyse: runs the analysis using the settings specified and fails if any error is thrown. It depends on the compile task.highwheelBaseAnalyseTask: same ashighwheelAnalyse, but with no other task dependency
Add to your project/plugins.sbt file the following line:
//resolvers += Resolver.mavenLocal
addSbtPlugin("com.github.fburato" %% "sbt-highwheel" % "1.8")Uncomment resolvers if you have installed highwheel or the pluing locally.
Add to your build.sbt the plugin as:
lazy val root = project.in(file("."))
.enablePlugins(com.github.fburato.highwheelmodulessbt.AnalyserPlugin)
.settings(...
)When running a multi-modules builds it is possible to run, on the parent module, an analysis that takes into
consideration all the output directories of the children modules. In order to do so, the children output directories
need to be added to the parent highwheelAnalysisPaths setting.
For example, if your multimodule build is defined in your build.sbt as:
lazy val core = project.in(file("core"))
lazy val web = project.in(file("web")).dependsOn(core)
lazy val datagateway = project.in(file("datagateway")).dependsOn(core)
lazy val root = project.in(file("."))
.dependsOn(web, datagateway)
.aggregate(core, web, datagateway)
.enablePlugins(com.github.fburato.highwheelmodulessbt.AnalyserPlugin)
.settings(
name := "my-project",
version := "0.1",
scalaVersion := "2.12.5"
)You can have the parent project analysing all the sub modules by adding the setting:
highwheelAnalysisPaths := Seq((classDirectory in core in Compile).value,(classDirectory in web in Compile).value,(classDirectory in datagateway in Compile).value)
so
lazy val core = project.in(file("core"))
lazy val web = project.in(file("web")).dependsOn(core)
lazy val datagateway = project.in(file("datagateway")).dependsOn(core)
lazy val root = project.in(file("."))
.dependsOn(web, datagateway)
.aggregate(core, web, datagateway)
.enablePlugins(com.github.fburato.highwheelmodulessbt.AnalyserPlugin)
.settings(
name := "my-project",
version := "0.1",
scalaVersion := "2.12.5",
highwheelAnalysisPaths := Seq((classDirectory in core in Compile).value,(classDirectory in web in Compile).value,(classDirectory in datagateway in Compile).value)
)Always in multi-modules build, you can create a task in the parent project that runs the analysis in all submodules and the parent itself as follows:
lazy val core = project.in(file("core"))
lazy val web = project.in(file("web")).dependsOn(core)
lazy val datagateway = project.in(file("datagateway")).dependsOn(core)
lazy val analyseAll = taskKey[Unit]("Run analysis in all submodules and parent")
lazy val root = project.in(file("."))
.dependsOn(web, datagateway)
.aggregate(core, web, datagateway)
.enablePlugins(com.github.fburato.highwheelmodulessbt.AnalyserPlugin)
.settings(
name := "my-project",
version := "0.1",
scalaVersion := "2.12.5",
highwheelAnalysisPaths := Seq((classDirectory in core in Compile).value,(classDirectory in web in Compile).value,(classDirectory in datagateway in Compile).value),
analyseAll := Def.sequential(
highwheelAnalyse in core,
highwheelAnalyse in web,
highwheelAnalyse in datagateway,
highwheelAnalyse
).value
)So, just add the new taskKey and use the sequential combinator on tasks to run the analysis task on every module in sequence.