SBT: How to make one task depend on another in multi-project builds, and not run in the root project? - scala

For my multi-project build, I'm trying to create a verify task that just results in scct:test and then scalastyle being executed in order. I would like scct:test to execute for all the subprojects, but not the top-level project. (If it executes for the top-level project, I get "timed out waiting for coverage report" from scct, since there's no source and no tests in that project.) What I had thought to do was to create verify as a task with dependencies on scct:test and scalastyle. This has turned out to be fairly baroque. Here is my Build.scala from my top-level project/ directory:
object MyBuild extends Build {
val verifyTask = TaskKey[Unit]("verify", "Compiles, runs tests via scct:test and then runs scalastyle")
val scctTestTask = (test in ScctPlugin.Scct).scopedKey
val scalastyleTask = PluginKeys.scalastyleTarget.scopedKey
lazy val root = Project("rootProject",
file("."),
settings = Defaults.defaultSettings ++
ScalastylePlugin.Settings ++
ScctPlugin.instrumentSettings ++
ScctPlugin.mergeReportSettings ++
Seq(
verifyTask in Global := {},
verifyTask <<= verifyTask.dependsOn(scctTestTask, scalastyleTask)
)
) aggregate(lift_webapp, selenium_tests)
lazy val subproject_1 = Project(id = "subproject_1", base = file("subproject_1"))
lazy val subproject_2 = Project(id = "subproject_2", base = file("subproject_2"))
}
However, the verify task only seems to exist for the root project; when I run it I don't see the same task being run in the subprojects. This is exactly the opposite of what I want; I'd like to issue sbt verify and have scct:test and scalastyle run in each of the subprojects but not in the top-level project. How might I go about doing that?

solution 1: define verifyTask in subprojects
First thing to note is that if you want some task (verify, test, etc) to run in some projects, you need to define them scoped to the subprojects. So in your case, the most straightforward thing to do this is to define verifyTask in subproject_1 and subproject_2.
lazy val scalaTest = "org.scalatest" %% "scalatest" % "3.0.4"
lazy val verify = taskKey[Unit]("verify")
def verifySettings = Seq(
skip in verify := false,
verify := (Def.taskDyn {
val sk = (skip in verify).value
if (sk) Def.task { println("skipping verify...") }
else (test in Test)
}).value
)
lazy val root = (project in file("."))
.aggregate(sub1, sub2)
.settings(
verifySettings,
scalaVersion in ThisBuild := "2.12.4",
skip in verify := true
)
lazy val sub1 = (project in file("sub1"))
.settings(
verifySettings,
libraryDependencies += scalaTest % Test
)
lazy val sub2 = (project in file("sub2"))
.settings(
verifySettings,
libraryDependencies += scalaTest % Test
)
solution 2: ScopeFilter
There was a recent Reddit thread that mentioned this question, so I'll post what I've done there.
If you want to manually aggregate on some subprojects, there's also a technique called ScopeFilter.
Note that I am using sbt 1.x here, but it should work with sbt 0.13 some minor change.
lazy val packageAll = taskKey[Unit]("package all the projects")
lazy val myTask = inputKey[Unit]("foo")
lazy val root = (project in file("."))
.aggregate(sub1, sub2)
.settings(
scalaVersion in ThisBuild := "2.12.4",
packageAll := {
(packageBin in Compile).all(nonRootsFilter).value
()
},
myTask := {
packageAll.value
}
)
lazy val sub1 = (project in file("sub1"))
lazy val sub2 = (project in file("sub2"))
def nonRootsFilter = {
import sbt.internal.inc.ReflectUtilities
def nonRoots: List[ProjectReference] =
allProjects filter {
case LocalProject(p) => p != "root"
case _ => false
}
def allProjects: List[ProjectReference] =
ReflectUtilities.allVals[Project](this).values.toList map { p =>
p: ProjectReference
}
ScopeFilter(inProjects(nonRoots: _*), inAnyConfiguration)
}
In the above, myTask depends on packageAll, which aggregates (packageBin in Compile) for all non-root subprojects.
sbt:root> myTask
[info] Packaging /Users/xxx/packageall/sub1/target/scala-2.12/sub1_2.12-0.1.0-SNAPSHOT.jar ...
[info] Done packaging.
[info] Packaging /Users/xxx/packageall/sub2/target/scala-2.12/sub2_2.12-0.1.0-SNAPSHOT.jar ...
[info] Done packaging.
[success] Total time: 0 s, completed Feb 2, 2018 7:23:23 PM

I may be wrong, but you are defining the verify task dependency only for the current project.
Maybe you can try:
Seq(
verifyTask in Global := {},
verifyTask <<= (verifyTask in Global).dependsOn(scctTestTask, scalastyleTask)
)
Or you can add the verifyTask settings to all your modules.

Related

build.sbt - iteration over sub projects for common settings in monorepo

I'm implementing a monorepo using SBT. I would like to iterate over my subprojects in order to initialize them (as the have the same configuration) and prevent code duplication.
In my build.sbt:
lazy val root = (project in file("."))
.aggregate(projects: _*)
.settings(
crossScalaVersions := Nil,
publish / skip := true
)
lazy val projects = Seq("projectA", "projectB", "projectC")
.map((projectName: String) => (project in file(projectName))
.settings(
name := projectName,
commonSettings,
libraryDependencies ++= ModulesDependencies.get(projectName))
.project
)
I'm getting the error:
error: project must be directly assigned to a val, such as `val x = project`. Alternatively, you can use `sbt.Project.apply`
Based on the error message, I also tried to use sbt.Project.apply(projectName, file(projectName)).settings(...) instead, but I'm also facing some funny errors.
From what I understand, it seems that SBT expects me to declare as lazy val projectA = (project in file("projectA")).settings(...), which works fine but I would have to duplicate this code for all my sub projects.
Is this iteration that I try to implement even possible?
Utility method might help with some of the duplication, for example
def createProject(projectName: String) = {
Project(projectName, file(projectName))
.settings(
name := projectName,
commonSettings,
libraryDependencies ++= ModulesDependencies.get(projectName)
)
}
lazy val projectA = createProject("projectA")
lazy val projectB = createProject("projectB")
lazy val projectC = createProject("projectC")
lazy val root = (project in file("."))
.aggregate(projectA, projectB, projectB)
.settings(
crossScalaVersions := Nil,
publish / skip := true
)

How to create a generic SBT root project with varying sub projects

I'm working on the Scala track in Exercism which means I have a lot of SBT projects in a root folder. I'd like to create a root SBT project which will automatically add new sub-projects as I download new exercises. Currently I have to add them manually, so my root build.sbt looks like this:
lazy val root = (project in file("."))
.aggregate(
hello_world,
sum_of_multiples,
robot_name)
lazy val hello_world = project in file("hello-world")
lazy val sum_of_multiples = project in file("sum-of-multiples")
lazy val robot_name = project in file("robot-name")
...but I'd like to avoid having to add every project manually. Is there a way to add new projects automatically?
I'd like to avoid having to add every project manually. Is there a way to add new projects automatically?
Sure. It's a bit advanced use of sbt, but you can create an ad-hoc plugin that generates subprojects programmatically.
build.sbt
ThisBuild / scalaVersion := "2.12.8"
ThisBuild / version := "0.1.0-SNAPSHOT"
ThisBuild / organization := "com.example"
ThisBuild / organizationName := "example"
project/build.properties
sbt.version=1.2.8
project/SubprojectPlugin.scala
import sbt._
object SubprojectPlugin extends AutoPlugin {
override val requires = sbt.plugins.JvmPlugin
override val trigger = allRequirements
override lazy val extraProjects: Seq[Project] = {
val dirs = (file(".") * ("*" -- "project" -- "target")) filter { _.isDirectory }
dirs.get.toList map { dir =>
Project(dir.getName.replaceAll("""\W""", "_"), dir)
}
}
}
Now if you start up sbt, any directories that are not named target or project will result to a subproject.
sbt:generic-root> projects
[info] In file:/private/tmp/generic-root/
[info] * generic-root
[info] hello_world
[info] robot_name
[info] sum_of_multiple
hello-world/build.sbt
To add further settings you can create build.sbt file under the directory as follows:
libraryDependencies += "commons-io" % "commons-io" % "2.6"

Sbt: How to define task for all projects?

I would like to be able to define a task for all projects in my sbt.build:
lazy val project1 = project.in(`.` / "project1)
...
lazy val project2 =
...
lazy val upload = taskKey[Unit]("upload a config file from project to server)
upload := {
val file = baseDirectory.value / "config.json"
...
}
The problem is this definition works only when I call sbt upload, but I would like to be able to call it for each subproject: sbt project1/upload and sbt project2/upload.
Is there a way to do it, without using inputKey?
See Organizing the build:
For more advanced users, another way of organizing your build is to define one-off auto plugins in project/*.scala. By defining triggered plugins, auto plugins can be used as a convenient way to inject custom tasks and commands across all subprojects.
project/UploadPlugin.scala
package something
import sbt._
import Keys._
object UploadPlugin extends AutoPlugin {
override def requires = sbt.plugins.JvmPlugin
override def trigger = allRequirements
object autoImport {
val upload = taskKey[Unit]("upload a config file from project to server")
}
import autoImport._
override lazy val projectSettings = Seq(
upload := {
val n = name.value
println(s"uploading $n..")
}
)
}
build.sbt
Here's how you can use it:
ThisBuild / organization := "com.example"
ThisBuild / scalaVersion := "2.12.5"
ThisBuild / version := "0.1.0-SNAPSHOT"
lazy val root = (project in file("."))
.aggregate(project1, project2)
.settings(
name := "Hello"
)
lazy val project1 = (project in file("project1"))
lazy val project2 = (project in file("project2"))
build.sbt does not have to mention anything about UploadPlugin, since it's a triggered plugin. From the shell you can call:
sbt:Hello> project1/upload
uploading project1..
[success] Total time: 0 s, completed Jul 20, 2018
sbt:Hello> project2/upload
uploading project2..
[success] Total time: 0 s, completed Jul 20, 2018
You can add the task as a setting of the project you want :
lazy val uploadTask = {
lazy val upload = taskKey[Unit]("upload a config file from project to server)
upload := {
val file = baseDirectory.value / "config.json"
...
}
}
project.in(`.` / "project1).settings(uploadTask)

sbt-assembly does not pick up configuration specific settings

I am updating an old 0.7.x build file from the tool sbt that thankfully removed the reference to "simple" from its name in the meantime.
Something that once worked, does not do so any longer. I had different config entries for platform specific assembly tasks. These include specific filters that for some reason are now called assemblyExcludedJars instead of excludedJars, and specific jar names that for some reason are now called assemblyJarName instead of jarName.
Basically:
val Foo = config("foo") extend Compile
lazy val assemblyFoo = TaskKey[File]("assembly-foo")
lazy val root = Project(id = "root", base = file("."))
// .configs(Foo) // needed? doesn't change anything
.settings(
inConfig(Foo)(inTask(assembly) {
assemblyJarName := "wtf.jar"
}),
scalaVersion := "2.11.7",
assemblyFoo <<= assembly in Foo
)
Now I would expect that if I run sbt assembly-foo or sbt foo:assembly, it would produce a file wtf.jar. But I am getting the default root-assembly-0.1-snapshot.jar. The same problem happens when I try to specify assemblyExcludedJars, they are simply ignored and still included.
If I remove the inConfig it works:
lazy val root = Project(id = "root", base = file("."))
.settings(
inTask(assembly) {
assemblyJarName := "wtf.jar"
},
scalaVersion := "2.11.7",
assemblyFoo <<= assembly in Foo
)
But now I cannot use different jar names for different configurations (which is the whole point).
As described in a blog post by one of sbt's authors and the author of sbt-assembly, this should work. It was also written in this Stackoverflow question. But the example requires an antique version of sbt-assembly (0.9.0 from 2013, before auto plugins etc.) and doesn't seem to apply to the current versions.
If one defines a new configuration, one has to redefine (?) all the tasks one is about to use. Apparently for sbt-assembly, this means running baseAssemblySettings:
val Foo = config("foo") extend Compile
lazy val assemblyFoo = TaskKey[File]("assembly-foo")
lazy val root = Project(id = "root", base = file("."))
.settings(
inConfig(Foo)(baseAssemblySettings /* !!! */ ++ inTask(assembly) {
jarName := "wtf.jar"
}),
scalaVersion := "2.11.7",
assemblyFoo := (assembly in Foo).value
)
Tested with sbt 0.13.9 and sbt-assembly 0.14.1.

How to use custom plugin in multi-module project?

In a multi-module project, with one module implementing a custom SBT plugin with custom TaskKey, how this plugin can be imported for the project settings of another submodule.
If the submodule using the plugin is define in submodule1/build.sbt, then submodule1/project/plugins.sbt is not loaded.
If plugin is registered in project/plugins.sbt it will fails when loading top/aggregate project as plugin is not necessarily already built.
Is there any other way to define a custom task requiring custom dependency so that it can be used by a submodule?
Here is how I finally make it works:
import sbt._
import Keys._
object MyBuild extends Build {
private lazy val myGenerator =
// Private project with generator code and its specific dependencies
// (e.g. Javassist)
Project(id = "my-generator",
base = file("project") / "my-generator").settings(
name := "my-generator",
javacOptions in Test ++= Seq("-Xlint:unchecked", "-Xlint:deprecation"),
autoScalaLibrary := false,
scalacOptions += "-feature",
resolvers +=
"Typesafe Snapshots" at "http://repo.typesafe.com/typesafe/releases/",
libraryDependencies ++= Seq( // Dependencies required to generate classes
"org.javassist" % "javassist" % "3.18.2-GA")
)
// Some custom setting & task
lazy val generatedClassDirectory = settingKey[File](
"Directory where classes get generated")
lazy val generatedClasses = taskKey[Seq[(File, String)]]("Generated classes")
lazy val myProject =
Project(id = "my-project", base = file("my-project")).settings(
name := "my-project",
javacOptions in Test ++= Seq("-Xlint:unchecked", "-Xlint:deprecation"),
autoScalaLibrary := false,
scalacOptions += "-feature",
libraryDependencies ++= Seq(/* ... */),
generatedClassDirectory := {
// Defines setting for path to generated classes
val dir = target.value / "generated_classes"
if (!dir.exists) dir.mkdirs()
dir
},
generatedClasses <<= Def.task { // Define task generating .class files
// first get classloader including generator and its dependencies
val cp = (fullClasspath in (myGenerator, Compile)).value
val cl = classpath.ClasspathUtilities.toLoader(cp.files)
// then loaded generator class, and instantiate with structural type
val genClass = cl loadClass "my.custom.GeneratorClass"
val generator = genClass.newInstance.
asInstanceOf[{def writeTo(out: File): File}]
// finally we can call the
val outdir = generatedClassDirectory.value
val generated = generator writeTo outdir
val path = generated.getAbsolutePath
// Mappings describing generated classes
Seq[(File, String)](generated -> path.
drop(outdir.getAbsolutePath.length+1))
} dependsOn(compile in (myGenerator, Compile))/* awkward? */,
managedClasspath in Compile := {
// Add generated classes to compilation classpath,
// so it can be used in my-project sources
val cp = (managedClasspath in Compile).value
cp :+ Attributed.blank(generatedClassDirectory.value)
},
// Make sure custom class generation is done before compile
compile in Compile <<= (compile in Compile) dependsOn generatedClasses,
mappings in (Compile, packageBin) := {
val ms = mappings.in(Compile, packageBin).value
ms ++ generatedClasses.value // add generated classes to package
}
).dependsOn(myGenerator/* required even if there dependsOn(compile in (myGenerator, Compile)) */)
}
Not sure there is better solution, especially about the redondant dependsOn(compile in (myGenerator, Compile)) and .dependsOn(myGenerator).