mirror of https://github.com/sbt/sbt.git
Merge pull request #7444 from adpi2/sbt-2-refactor-execute-2
[2.x] Reduce abstraction in `Execute` and around (bis)
This commit is contained in:
commit
8950b92d33
|
|
@ -25,7 +25,7 @@ import Scope.GlobalScope
|
|||
import sbt.SlashSyntax0.*
|
||||
import sbt.internal.parser.SbtParser
|
||||
import sbt.io.IO
|
||||
import scala.collection.JavaConverters.*
|
||||
import scala.jdk.CollectionConverters.*
|
||||
import xsbti.VirtualFile
|
||||
import xsbti.VirtualFileRef
|
||||
|
||||
|
|
@ -403,14 +403,14 @@ object Index {
|
|||
)
|
||||
}
|
||||
|
||||
private[this] type TriggerMap = collection.mutable.HashMap[Task[Any], Seq[Task[Any]]]
|
||||
private[this] type TriggerMap = collection.mutable.HashMap[TaskId[?], Seq[TaskId[?]]]
|
||||
|
||||
def triggers(ss: Settings[Scope]): Triggers[Task] = {
|
||||
def triggers(ss: Settings[Scope]): Triggers = {
|
||||
val runBefore = new TriggerMap
|
||||
val triggeredBy = new TriggerMap
|
||||
ss.data.values foreach (
|
||||
_.entries foreach {
|
||||
case AttributeEntry(_, value: Task[Any]) =>
|
||||
case AttributeEntry(_, value: Task[?]) =>
|
||||
val as = value.info.attributes
|
||||
update(runBefore, value, as.get(Def.runBefore.asInstanceOf))
|
||||
update(triggeredBy, value, as.get(Def.triggeredBy.asInstanceOf))
|
||||
|
|
@ -418,13 +418,13 @@ object Index {
|
|||
}
|
||||
)
|
||||
val onComplete = (GlobalScope / Def.onComplete) get ss getOrElse (() => ())
|
||||
new Triggers[Task](runBefore, triggeredBy, map => { onComplete(); map })
|
||||
new Triggers(runBefore, triggeredBy, map => { onComplete(); map })
|
||||
}
|
||||
|
||||
private[this] def update(
|
||||
map: TriggerMap,
|
||||
base: Task[Any],
|
||||
tasksOpt: Option[Seq[Task[Any]]]
|
||||
base: Task[?],
|
||||
tasksOpt: Option[Seq[Task[?]]]
|
||||
): Unit =
|
||||
for {
|
||||
tasks <- tasksOpt
|
||||
|
|
|
|||
|
|
@ -382,10 +382,11 @@ object Def extends Init[Scope] with TaskMacroExtra with InitializeImplicits:
|
|||
(TaskKey[A](name, description, DTask), dummyTask(name))
|
||||
|
||||
private[sbt] def dummyTask[T](name: String): Task[T] = {
|
||||
import std.TaskExtra.{ task => newTask, _ }
|
||||
import std.TaskExtra.{ task => newTask, toTaskInfo }
|
||||
val base: Task[T] = newTask(
|
||||
sys.error("Dummy task '" + name + "' did not get converted to a full task.")
|
||||
) named name
|
||||
)
|
||||
.named(name)
|
||||
base.copy(info = base.info.set(isDummyTask, true))
|
||||
}
|
||||
|
||||
|
|
|
|||
|
|
@ -115,7 +115,7 @@ object Previous {
|
|||
/** Persists values of tasks t where there is some task referencing it via t.previous. */
|
||||
private[sbt] def complete(
|
||||
referenced: References,
|
||||
results: RMap[Task, Result],
|
||||
results: RMap[TaskId, Result],
|
||||
streams: Streams
|
||||
): Unit = {
|
||||
val map = referenced.getReferences
|
||||
|
|
@ -124,7 +124,7 @@ object Previous {
|
|||
// We first collect all of the successful tasks and write their scoped key into a map
|
||||
// along with their values.
|
||||
val successfulTaskResults = (for
|
||||
case results.TPair(task, Result.Value(v)) <- results.toTypedSeq
|
||||
results.TPair(task: Task[?], Result.Value(v)) <- results.toTypedSeq
|
||||
key <- task.info.attributes.get(Def.taskDefinitionKey).asInstanceOf[Option[AnyTaskKey]]
|
||||
yield key -> v).toMap
|
||||
// We then traverse the successful results and look up all of the referenced values for
|
||||
|
|
|
|||
|
|
@ -534,7 +534,7 @@ object Scoped:
|
|||
.apply(deps => nop.dependsOn(deps: _*))
|
||||
}
|
||||
|
||||
sealed abstract class RichTaskables[K[L[x]]](final val keys: K[Taskable])(using
|
||||
sealed abstract class RichTaskables[K[+L[x]]](final val keys: K[Taskable])(using
|
||||
a: AList[K]
|
||||
):
|
||||
|
||||
|
|
|
|||
|
|
@ -12,7 +12,6 @@ import java.util.concurrent.atomic.AtomicReference
|
|||
|
||||
import sbt.Def.{ ScopedKey, Setting, dummyState }
|
||||
import sbt.Keys.{ TaskProgress => _, name => _, _ }
|
||||
// import sbt.Project.richInitializeTask
|
||||
import sbt.ProjectExtra.*
|
||||
import sbt.Scope.Global
|
||||
import sbt.SlashSyntax0._
|
||||
|
|
@ -102,7 +101,7 @@ object TaskCancellationStrategy {
|
|||
sealed trait EvaluateTaskConfig {
|
||||
def restrictions: Seq[Tags.Rule]
|
||||
def checkCycles: Boolean
|
||||
def progressReporter: ExecuteProgress[Task]
|
||||
def progressReporter: ExecuteProgress
|
||||
def cancelStrategy: TaskCancellationStrategy
|
||||
|
||||
/** If true, we force a finalizer/gc run (or two) after task execution completes when needed. */
|
||||
|
|
@ -118,7 +117,7 @@ object EvaluateTaskConfig {
|
|||
def apply(
|
||||
restrictions: Seq[Tags.Rule],
|
||||
checkCycles: Boolean,
|
||||
progressReporter: ExecuteProgress[Task],
|
||||
progressReporter: ExecuteProgress,
|
||||
cancelStrategy: TaskCancellationStrategy,
|
||||
forceGarbageCollection: Boolean,
|
||||
minForcegcInterval: Duration
|
||||
|
|
@ -135,7 +134,7 @@ object EvaluateTaskConfig {
|
|||
private[this] case class DefaultEvaluateTaskConfig(
|
||||
restrictions: Seq[Tags.Rule],
|
||||
checkCycles: Boolean,
|
||||
progressReporter: ExecuteProgress[Task],
|
||||
progressReporter: ExecuteProgress,
|
||||
cancelStrategy: TaskCancellationStrategy,
|
||||
forceGarbageCollection: Boolean,
|
||||
minForcegcInterval: Duration
|
||||
|
|
@ -168,7 +167,7 @@ object EvaluateTask {
|
|||
|
||||
@nowarn
|
||||
lazy private val sharedProgress = new TaskTimings(reportOnShutdown = true)
|
||||
def taskTimingProgress: Option[ExecuteProgress[Task]] =
|
||||
def taskTimingProgress: Option[ExecuteProgress] =
|
||||
if (SysProp.taskTimingsOnShutdown) Some(sharedProgress)
|
||||
else None
|
||||
|
||||
|
|
@ -194,14 +193,14 @@ object EvaluateTask {
|
|||
}
|
||||
|
||||
lazy private val sharedTraceEvent = new TaskTraceEvent()
|
||||
def taskTraceEvent: Option[ExecuteProgress[Task]] =
|
||||
def taskTraceEvent: Option[ExecuteProgress] =
|
||||
if (SysProp.traces) {
|
||||
Some(sharedTraceEvent)
|
||||
} else None
|
||||
|
||||
// sbt-pgp calls this
|
||||
@deprecated("No longer used", "1.3.0")
|
||||
private[sbt] def defaultProgress(): ExecuteProgress[Task] = ExecuteProgress.empty[Task]
|
||||
private[sbt] def defaultProgress(): ExecuteProgress = ExecuteProgress.empty
|
||||
|
||||
val SystemProcessors = Runtime.getRuntime.availableProcessors
|
||||
|
||||
|
|
@ -262,26 +261,26 @@ object EvaluateTask {
|
|||
extracted: Extracted,
|
||||
structure: BuildStructure,
|
||||
state: State
|
||||
): ExecuteProgress[Task] = {
|
||||
): ExecuteProgress = {
|
||||
state
|
||||
.get(currentTaskProgress)
|
||||
.map { tp =>
|
||||
new ExecuteProgress[Task] {
|
||||
new ExecuteProgress {
|
||||
val progress = tp.progress
|
||||
override def initial(): Unit = progress.initial()
|
||||
override def afterRegistered(
|
||||
task: Task[Any],
|
||||
allDeps: Iterable[Task[Any]],
|
||||
pendingDeps: Iterable[Task[Any]]
|
||||
task: TaskId[?],
|
||||
allDeps: Iterable[TaskId[?]],
|
||||
pendingDeps: Iterable[TaskId[?]]
|
||||
): Unit =
|
||||
progress.afterRegistered(task, allDeps, pendingDeps)
|
||||
override def afterReady(task: Task[Any]): Unit = progress.afterReady(task)
|
||||
override def beforeWork(task: Task[Any]): Unit = progress.beforeWork(task)
|
||||
override def afterWork[A](task: Task[A], result: Either[Task[A], Result[A]]): Unit =
|
||||
override def afterReady(task: TaskId[?]): Unit = progress.afterReady(task)
|
||||
override def beforeWork(task: TaskId[?]): Unit = progress.beforeWork(task)
|
||||
override def afterWork[A](task: TaskId[A], result: Either[TaskId[A], Result[A]]): Unit =
|
||||
progress.afterWork(task, result)
|
||||
override def afterCompleted[A](task: Task[A], result: Result[A]): Unit =
|
||||
override def afterCompleted[A](task: TaskId[A], result: Result[A]): Unit =
|
||||
progress.afterCompleted(task, result)
|
||||
override def afterAllCompleted(results: RMap[Task, Result]): Unit =
|
||||
override def afterAllCompleted(results: RMap[TaskId, Result]): Unit =
|
||||
progress.afterAllCompleted(results)
|
||||
override def stop(): Unit = {}
|
||||
}
|
||||
|
|
@ -298,9 +297,9 @@ object EvaluateTask {
|
|||
new TaskTimings(reportOnShutdown = false, state.globalLogging.full) :: Nil
|
||||
else Nil)
|
||||
reporters match {
|
||||
case xs if xs.isEmpty => ExecuteProgress.empty[Task]
|
||||
case xs if xs.isEmpty => ExecuteProgress.empty
|
||||
case xs if xs.size == 1 => xs.head
|
||||
case xs => ExecuteProgress.aggregate[Task](xs)
|
||||
case xs => ExecuteProgress.aggregate(xs)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
@ -439,7 +438,7 @@ object EvaluateTask {
|
|||
state: State,
|
||||
streams: Streams,
|
||||
ref: ProjectRef
|
||||
): Option[(Task[T], NodeView[Task])] = {
|
||||
): Option[(Task[T], NodeView)] = {
|
||||
val thisScope = Load.projectScope(ref)
|
||||
val resolvedScope = Scope.replaceThis(thisScope)(taskKey.scope)
|
||||
for (t <- structure.data.get(resolvedScope, taskKey.key))
|
||||
|
|
@ -450,7 +449,7 @@ object EvaluateTask {
|
|||
streams: Streams,
|
||||
roots: Seq[ScopedKey[_]],
|
||||
dummies: DummyTaskMap = DummyTaskMap(Nil)
|
||||
): NodeView[Task] =
|
||||
): NodeView =
|
||||
Transform(
|
||||
(dummyRoots, roots) :: (Def.dummyStreamsManager, streams) :: (dummyState, state) :: dummies
|
||||
)
|
||||
|
|
@ -471,24 +470,21 @@ object EvaluateTask {
|
|||
root: Task[T],
|
||||
state: State,
|
||||
streams: Streams,
|
||||
triggers: Triggers[Task],
|
||||
triggers: Triggers,
|
||||
config: EvaluateTaskConfig
|
||||
)(using taskToNode: NodeView[Task]): (State, Result[T]) = {
|
||||
)(using taskToNode: NodeView): (State, Result[T]) = {
|
||||
import ConcurrentRestrictions.{ cancellableCompletionService, tagged, tagsKey }
|
||||
|
||||
val log = state.log
|
||||
log.debug(
|
||||
s"Running task... Cancel: ${config.cancelStrategy}, check cycles: ${config.checkCycles}, forcegc: ${config.forceGarbageCollection}"
|
||||
)
|
||||
def tagMap(t: Task[_]): Tags.TagMap =
|
||||
t.info.get(tagsKey).getOrElse(Map.empty)
|
||||
val tags =
|
||||
tagged[Task[Any]](tagMap, Tags.predicate(config.restrictions))
|
||||
val tags = tagged(Tags.predicate(config.restrictions))
|
||||
val (service, shutdownThreads) =
|
||||
cancellableCompletionService[Task[Any], Completed](
|
||||
cancellableCompletionService(
|
||||
tags,
|
||||
(s: String) => log.warn(s),
|
||||
(t: Task[_]) => tagMap(t).contains(Tags.Sentinel)
|
||||
(t: TaskId[?]) => t.tags.contains(Tags.Sentinel)
|
||||
)
|
||||
|
||||
def shutdownImpl(force: Boolean): Unit = {
|
||||
|
|
@ -504,18 +500,18 @@ object EvaluateTask {
|
|||
def shutdown(): Unit = shutdownImpl(false)
|
||||
// propagate the defining key for reporting the origin
|
||||
def overwriteNode(i: Incomplete): Boolean = i.node match {
|
||||
case Some(t: Task[_]) => transformNode(t).isEmpty
|
||||
case Some(t: Task[?]) => transformNode(t).isEmpty
|
||||
case _ => true
|
||||
}
|
||||
def run() = {
|
||||
val x = new Execute[Task](
|
||||
val x = new Execute(
|
||||
Execute.config(config.checkCycles, overwriteNode),
|
||||
triggers,
|
||||
config.progressReporter
|
||||
)
|
||||
val (newState, result) =
|
||||
try {
|
||||
given strategy: x.Strategy = service
|
||||
given strategy: CompletionService = service
|
||||
val results = x.runKeep(root)
|
||||
storeValuesForPrevious(results, state, streams)
|
||||
applyResults(results, state, root)
|
||||
|
|
@ -548,7 +544,7 @@ object EvaluateTask {
|
|||
}
|
||||
|
||||
private[this] def storeValuesForPrevious(
|
||||
results: RMap[Task, Result],
|
||||
results: RMap[TaskId, Result],
|
||||
state: State,
|
||||
streams: Streams
|
||||
): Unit =
|
||||
|
|
@ -556,13 +552,13 @@ object EvaluateTask {
|
|||
Previous.complete(referenced, results, streams)
|
||||
|
||||
def applyResults[T](
|
||||
results: RMap[Task, Result],
|
||||
results: RMap[TaskId, Result],
|
||||
state: State,
|
||||
root: Task[T]
|
||||
): (State, Result[T]) = {
|
||||
(stateTransform(results)(state), results(root))
|
||||
}
|
||||
def stateTransform(results: RMap[Task, Result]): State => State =
|
||||
def stateTransform(results: RMap[TaskId, Result]): State => State =
|
||||
Function.chain(
|
||||
results.toTypedSeq flatMap {
|
||||
case results.TPair(_, Result.Value(KeyValue(_, st: StateTransform))) => Some(st.transform)
|
||||
|
|
@ -577,27 +573,25 @@ object EvaluateTask {
|
|||
Incomplete.transformBU(i)(convertCyclicInc andThen taskToKey andThen liftAnonymous)
|
||||
}
|
||||
def taskToKey: Incomplete => Incomplete = {
|
||||
case in @ Incomplete(Some(node: Task[_]), _, _, _, _) => in.copy(node = transformNode(node))
|
||||
case in @ Incomplete(Some(node: Task[?]), _, _, _, _) => in.copy(node = transformNode(node))
|
||||
case i => i
|
||||
}
|
||||
|
||||
type AnyCyclic = Execute[({ type A[_] <: AnyRef })#A]#CyclicException[_]
|
||||
|
||||
def convertCyclicInc: Incomplete => Incomplete = {
|
||||
case in @ Incomplete(
|
||||
_,
|
||||
_,
|
||||
_,
|
||||
_,
|
||||
Some(c: Execute[({ type A[_] <: AnyRef })#A @unchecked]#CyclicException[_])
|
||||
Some(c: Execute#CyclicException)
|
||||
) =>
|
||||
in.copy(directCause = Some(new RuntimeException(convertCyclic(c))))
|
||||
case i => i
|
||||
}
|
||||
|
||||
def convertCyclic(c: AnyCyclic): String =
|
||||
def convertCyclic(c: Execute#CyclicException): String =
|
||||
(c.caller, c.target) match {
|
||||
case (caller: Task[_], target: Task[_]) =>
|
||||
case (caller: Task[?], target: Task[?]) =>
|
||||
c.toString + (if (caller eq target) "(task: " + name(caller) + ")"
|
||||
else "(caller: " + name(caller) + ", target: " + name(target) + ")")
|
||||
case _ => c.toString
|
||||
|
|
|
|||
|
|
@ -590,9 +590,9 @@ object Keys {
|
|||
val state = Def.stateKey
|
||||
val streamsManager = Def.streamsManagerKey
|
||||
// wrapper to work around SI-2915
|
||||
final class TaskProgress(val progress: ExecuteProgress[Task])
|
||||
final class TaskProgress(val progress: ExecuteProgress)
|
||||
object TaskProgress {
|
||||
def apply(progress: ExecuteProgress[Task]): TaskProgress = new TaskProgress(progress)
|
||||
def apply(progress: ExecuteProgress): TaskProgress = new TaskProgress(progress)
|
||||
}
|
||||
private[sbt] val currentTaskProgress = AttributeKey[TaskProgress]("current-task-progress")
|
||||
private[sbt] val taskProgress = AttributeKey[sbt.internal.TaskProgress]("active-task-progress")
|
||||
|
|
|
|||
|
|
@ -10,18 +10,18 @@ package internal
|
|||
|
||||
import java.util.concurrent.ConcurrentHashMap
|
||||
import java.util.concurrent.atomic.AtomicLong
|
||||
import scala.collection.JavaConverters._
|
||||
import scala.jdk.CollectionConverters.*
|
||||
import scala.collection.mutable
|
||||
import scala.collection.immutable.VectorBuilder
|
||||
import scala.concurrent.duration._
|
||||
|
||||
private[sbt] abstract class AbstractTaskExecuteProgress extends ExecuteProgress[Task] {
|
||||
private[sbt] abstract class AbstractTaskExecuteProgress extends ExecuteProgress {
|
||||
import AbstractTaskExecuteProgress.Timer
|
||||
|
||||
private[this] val showScopedKey = Def.showShortKey(None)
|
||||
private[this] val anonOwners = new ConcurrentHashMap[Task[_], Task[_]]
|
||||
private[this] val calledBy = new ConcurrentHashMap[Task[_], Task[_]]
|
||||
private[this] val timings = new ConcurrentHashMap[Task[_], Timer]
|
||||
private[this] val anonOwners = new ConcurrentHashMap[TaskId[_], TaskId[_]]
|
||||
private[this] val calledBy = new ConcurrentHashMap[TaskId[_], TaskId[_]]
|
||||
private[this] val timings = new ConcurrentHashMap[TaskId[_], Timer]
|
||||
private[sbt] def timingsByName: mutable.Map[String, AtomicLong] = {
|
||||
val result = new ConcurrentHashMap[String, AtomicLong]
|
||||
timings.forEach { (task, timing) =>
|
||||
|
|
@ -34,18 +34,18 @@ private[sbt] abstract class AbstractTaskExecuteProgress extends ExecuteProgress[
|
|||
result.asScala
|
||||
}
|
||||
private[sbt] def anyTimings = !timings.isEmpty
|
||||
def currentTimings: Iterator[(Task[_], Timer)] = timings.asScala.iterator
|
||||
def currentTimings: Iterator[(TaskId[_], Timer)] = timings.asScala.iterator
|
||||
|
||||
private[internal] def exceededThreshold(task: Task[_], threshold: FiniteDuration): Boolean =
|
||||
private[internal] def exceededThreshold(task: TaskId[_], threshold: FiniteDuration): Boolean =
|
||||
timings.get(task) match {
|
||||
case null => false
|
||||
case t => t.durationMicros > threshold.toMicros
|
||||
}
|
||||
private[internal] def timings(
|
||||
tasks: java.util.Set[Task[_]],
|
||||
tasks: java.util.Set[TaskId[_]],
|
||||
thresholdMicros: Long
|
||||
): Vector[(Task[_], Long)] = {
|
||||
val result = new VectorBuilder[(Task[_], Long)]
|
||||
): Vector[(TaskId[_], Long)] = {
|
||||
val result = new VectorBuilder[(TaskId[_], Long)]
|
||||
val now = System.nanoTime
|
||||
tasks.forEach { t =>
|
||||
timings.get(t) match {
|
||||
|
|
@ -60,7 +60,7 @@ private[sbt] abstract class AbstractTaskExecuteProgress extends ExecuteProgress[
|
|||
result.result()
|
||||
}
|
||||
def activeTasks(now: Long) = {
|
||||
val result = new VectorBuilder[(Task[_], FiniteDuration)]
|
||||
val result = new VectorBuilder[(TaskId[_], FiniteDuration)]
|
||||
timings.forEach { (task, timing) =>
|
||||
if (timing.isActive) result += task -> (now - timing.startNanos).nanos
|
||||
}
|
||||
|
|
@ -68,25 +68,26 @@ private[sbt] abstract class AbstractTaskExecuteProgress extends ExecuteProgress[
|
|||
}
|
||||
|
||||
override def afterRegistered(
|
||||
task: Task[Any],
|
||||
allDeps: Iterable[Task[Any]],
|
||||
pendingDeps: Iterable[Task[Any]]
|
||||
task: TaskId[?],
|
||||
allDeps: Iterable[TaskId[?]],
|
||||
pendingDeps: Iterable[TaskId[?]]
|
||||
): Unit = {
|
||||
// we need this to infer anonymous task names
|
||||
pendingDeps foreach { t =>
|
||||
if (TaskName.transformNode(t).isEmpty) {
|
||||
anonOwners.put(t, task)
|
||||
pendingDeps
|
||||
.filter {
|
||||
case t: Task[?] => TaskName.transformNode(t).isEmpty
|
||||
case _ => true
|
||||
}
|
||||
}
|
||||
.foreach(anonOwners.put(_, task))
|
||||
}
|
||||
|
||||
override def beforeWork(task: Task[Any]): Unit = {
|
||||
override def beforeWork(task: TaskId[?]): Unit = {
|
||||
timings.put(task, new Timer)
|
||||
()
|
||||
}
|
||||
|
||||
protected def clearTimings: Boolean = false
|
||||
override def afterWork[A](task: Task[A], result: Either[Task[A], Result[A]]): Unit = {
|
||||
override def afterWork[A](task: TaskId[A], result: Either[TaskId[A], Result[A]]): Unit = {
|
||||
if (clearTimings) timings.remove(task)
|
||||
else
|
||||
timings.get(task) match {
|
||||
|
|
@ -100,21 +101,23 @@ private[sbt] abstract class AbstractTaskExecuteProgress extends ExecuteProgress[
|
|||
}
|
||||
}
|
||||
|
||||
private[this] val taskNameCache = new ConcurrentHashMap[Task[_], String]
|
||||
protected def taskName(t: Task[_]): String = taskNameCache.get(t) match {
|
||||
private[this] val taskNameCache = new ConcurrentHashMap[TaskId[_], String]
|
||||
protected def taskName(t: TaskId[_]): String = taskNameCache.get(t) match {
|
||||
case null =>
|
||||
val name = taskName0(t)
|
||||
taskNameCache.putIfAbsent(t, name)
|
||||
name
|
||||
case name => name
|
||||
}
|
||||
private[this] def taskName0(t: Task[_]): String = {
|
||||
private[this] def taskName0(t: TaskId[_]): String = {
|
||||
def definedName(node: Task[_]): Option[String] =
|
||||
node.info.name orElse TaskName.transformNode(node).map(showScopedKey.show)
|
||||
node.info.name.orElse(TaskName.transformNode(node).map(showScopedKey.show))
|
||||
def inferredName(t: Task[_]): Option[String] = nameDelegate(t) map taskName
|
||||
def nameDelegate(t: Task[_]): Option[Task[_]] =
|
||||
Option(anonOwners.get(t)) orElse Option(calledBy.get(t))
|
||||
definedName(t) orElse inferredName(t) getOrElse TaskName.anonymousName(t)
|
||||
def nameDelegate(t: Task[_]): Option[TaskId[_]] =
|
||||
Option(anonOwners.get(t)).orElse(Option(calledBy.get(t)))
|
||||
t match
|
||||
case t: Task[?] => definedName(t).orElse(inferredName(t)).getOrElse(TaskName.anonymousName(t))
|
||||
case _ => TaskName.anonymousName(t)
|
||||
}
|
||||
}
|
||||
|
||||
|
|
|
|||
|
|
@ -71,7 +71,7 @@ final class BuildStructure(
|
|||
final class StructureIndex(
|
||||
val keyMap: Map[String, AttributeKey[_]],
|
||||
val taskToKey: Map[Task[_], ScopedKey[Task[_]]],
|
||||
val triggers: Triggers[Task],
|
||||
val triggers: Triggers,
|
||||
val keyIndex: KeyIndex,
|
||||
val aggregateKeyIndex: KeyIndex,
|
||||
)
|
||||
|
|
|
|||
|
|
@ -12,11 +12,11 @@ import Def.{ displayFull, ScopedKey }
|
|||
import Keys.taskDefinitionKey
|
||||
|
||||
private[sbt] object TaskName {
|
||||
def name(node: Task[_]): String = definedName(node) getOrElse anonymousName(node)
|
||||
def name(node: Task[_]): String = definedName(node).getOrElse(anonymousName(node))
|
||||
def definedName(node: Task[_]): Option[String] =
|
||||
node.info.name orElse transformNode(node).map(displayFull)
|
||||
def anonymousName(node: Task[_]): String =
|
||||
node.info.name.orElse(transformNode(node).map(displayFull))
|
||||
def anonymousName(node: TaskId[_]): String =
|
||||
"<anon-" + System.identityHashCode(node).toHexString + ">"
|
||||
def transformNode(node: Task[_]): Option[ScopedKey[_]] =
|
||||
node.info.attributes get taskDefinitionKey
|
||||
node.info.attributes.get(taskDefinitionKey)
|
||||
}
|
||||
|
|
|
|||
|
|
@ -26,11 +26,11 @@ private[sbt] class TaskProgress(
|
|||
threshold: FiniteDuration,
|
||||
logger: Logger
|
||||
) extends AbstractTaskExecuteProgress
|
||||
with ExecuteProgress[Task]
|
||||
with ExecuteProgress
|
||||
with AutoCloseable {
|
||||
private[this] val lastTaskCount = new AtomicInteger(0)
|
||||
private[this] val reportLoop = new AtomicReference[AutoCloseable]
|
||||
private[this] val active = new ConcurrentHashMap[Task[_], AutoCloseable]
|
||||
private[this] val active = new ConcurrentHashMap[TaskId[_], AutoCloseable]
|
||||
private[this] val nextReport = new AtomicReference(Deadline.now)
|
||||
private[this] val scheduler =
|
||||
Executors.newSingleThreadScheduledExecutor(r => new Thread(r, "sbt-progress-report-scheduler"))
|
||||
|
|
@ -90,7 +90,7 @@ private[sbt] class TaskProgress(
|
|||
}
|
||||
Util.ignoreResult(pending.add(executor.submit(runnable)))
|
||||
}
|
||||
override def beforeWork(task: Task[Any]): Unit =
|
||||
override def beforeWork(task: TaskId[?]): Unit =
|
||||
if (!closed.get) {
|
||||
super.beforeWork(task)
|
||||
reportLoop.get match {
|
||||
|
|
@ -108,7 +108,7 @@ private[sbt] class TaskProgress(
|
|||
logger.debug(s"called beforeWork for ${taskName(task)} after task progress was closed")
|
||||
}
|
||||
|
||||
override def afterReady(task: Task[Any]): Unit =
|
||||
override def afterReady(task: TaskId[?]): Unit =
|
||||
if (!closed.get) {
|
||||
try {
|
||||
Util.ignoreResult(executor.submit((() => {
|
||||
|
|
@ -124,7 +124,7 @@ private[sbt] class TaskProgress(
|
|||
}
|
||||
override def stop(): Unit = {}
|
||||
|
||||
override def afterCompleted[A](task: Task[A], result: Result[A]): Unit =
|
||||
override def afterCompleted[A](task: TaskId[A], result: Result[A]): Unit =
|
||||
active.remove(task) match {
|
||||
case null =>
|
||||
case a =>
|
||||
|
|
@ -132,7 +132,7 @@ private[sbt] class TaskProgress(
|
|||
if (exceededThreshold(task, threshold)) report()
|
||||
}
|
||||
|
||||
override def afterAllCompleted(results: RMap[Task, Result]): Unit = {
|
||||
override def afterAllCompleted(results: RMap[TaskId, Result]): Unit = {
|
||||
reportLoop.getAndSet(null) match {
|
||||
case null =>
|
||||
case l => l.close()
|
||||
|
|
@ -164,7 +164,7 @@ private[sbt] class TaskProgress(
|
|||
val ltc = lastTaskCount.get
|
||||
if (currentTasks.nonEmpty || ltc != 0) {
|
||||
val currentTasksCount = currentTasks.size
|
||||
def event(tasks: Vector[(Task[_], Long)]): ProgressEvent = {
|
||||
def event(tasks: Vector[(TaskId[_], Long)]): ProgressEvent = {
|
||||
if (tasks.nonEmpty) nextReport.set(Deadline.now + sleepDuration)
|
||||
val toWrite = tasks.sortBy(_._2)
|
||||
val distinct = new java.util.LinkedHashMap[String, ProgressItem]
|
||||
|
|
@ -187,7 +187,7 @@ private[sbt] class TaskProgress(
|
|||
}
|
||||
}
|
||||
|
||||
private[this] def getShortName(task: Task[_]): String = {
|
||||
private[this] def getShortName(task: TaskId[_]): String = {
|
||||
val name = taskName(task)
|
||||
name.lastIndexOf('/') match {
|
||||
case -1 => name
|
||||
|
|
@ -199,12 +199,13 @@ private[sbt] class TaskProgress(
|
|||
|
||||
}
|
||||
private[this] def filter(
|
||||
tasks: Vector[(Task[_], Long)]
|
||||
): (Vector[(Task[_], Long)], Boolean) = {
|
||||
tasks.foldLeft((Vector.empty[(Task[_], Long)], false)) { case ((tasks, skip), pair @ (t, _)) =>
|
||||
val shortName = getShortName(t)
|
||||
val newSkip = skip || skipReportTasks.contains(shortName)
|
||||
if (hiddenTasks.contains(shortName)) (tasks, newSkip) else (tasks :+ pair, newSkip)
|
||||
tasks: Vector[(TaskId[_], Long)]
|
||||
): (Vector[(TaskId[_], Long)], Boolean) = {
|
||||
tasks.foldLeft((Vector.empty[(TaskId[_], Long)], false)) {
|
||||
case ((tasks, skip), pair @ (t, _)) =>
|
||||
val shortName = getShortName(t)
|
||||
val newSkip = skip || skipReportTasks.contains(shortName)
|
||||
if (hiddenTasks.contains(shortName)) (tasks, newSkip) else (tasks :+ pair, newSkip)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -23,7 +23,7 @@ import sbt.util.{ Level, Logger }
|
|||
*/
|
||||
private[sbt] final class TaskTimings(reportOnShutdown: Boolean, logger: Logger)
|
||||
extends AbstractTaskExecuteProgress
|
||||
with ExecuteProgress[Task] {
|
||||
with ExecuteProgress {
|
||||
@deprecated("Use the constructor that takes an sbt.util.Logger parameter.", "1.3.3")
|
||||
def this(reportOnShutdown: Boolean) =
|
||||
this(
|
||||
|
|
@ -50,9 +50,9 @@ private[sbt] final class TaskTimings(reportOnShutdown: Boolean, logger: Logger)
|
|||
start = System.nanoTime
|
||||
}
|
||||
|
||||
override def afterReady(task: Task[Any]): Unit = ()
|
||||
override def afterCompleted[T](task: Task[T], result: Result[T]): Unit = ()
|
||||
override def afterAllCompleted(results: RMap[Task, Result]): Unit =
|
||||
override def afterReady(task: TaskId[?]): Unit = ()
|
||||
override def afterCompleted[T](task: TaskId[T], result: Result[T]): Unit = ()
|
||||
override def afterAllCompleted(results: RMap[TaskId, Result]): Unit =
|
||||
if (!reportOnShutdown) {
|
||||
report()
|
||||
}
|
||||
|
|
|
|||
|
|
@ -21,17 +21,15 @@ import sjsonnew.support.scalajson.unsafe.CompactPrinter
|
|||
* as Chrome Trace Event Format.
|
||||
* This class is activated by adding -Dsbt.traces=true to the JVM options.
|
||||
*/
|
||||
private[sbt] final class TaskTraceEvent
|
||||
extends AbstractTaskExecuteProgress
|
||||
with ExecuteProgress[Task] {
|
||||
private[sbt] final class TaskTraceEvent extends AbstractTaskExecuteProgress with ExecuteProgress {
|
||||
import AbstractTaskExecuteProgress.Timer
|
||||
private[this] var start = 0L
|
||||
private[this] val console = ConsoleOut.systemOut
|
||||
|
||||
override def initial(): Unit = ()
|
||||
override def afterReady(task: Task[Any]): Unit = ()
|
||||
override def afterCompleted[T](task: Task[T], result: Result[T]): Unit = ()
|
||||
override def afterAllCompleted(results: RMap[Task, Result]): Unit = ()
|
||||
override def afterReady(task: TaskId[?]): Unit = ()
|
||||
override def afterCompleted[T](task: TaskId[T], result: Result[T]): Unit = ()
|
||||
override def afterAllCompleted(results: RMap[TaskId, Result]): Unit = ()
|
||||
override def stop(): Unit = ()
|
||||
|
||||
start = System.nanoTime
|
||||
|
|
|
|||
|
|
@ -66,7 +66,7 @@ object BspCompileTask {
|
|||
case class BspCompileTask private (
|
||||
targetId: BuildTargetIdentifier,
|
||||
targetName: String,
|
||||
id: TaskId,
|
||||
id: sbt.internal.bsp.TaskId,
|
||||
startTimeMillis: Long
|
||||
) {
|
||||
import sbt.internal.bsp.codec.JsonProtocol._
|
||||
|
|
|
|||
|
|
@ -16,7 +16,7 @@ import sbt.util.Monad
|
|||
/**
|
||||
* Combines metadata `info` and a computation `work` to define a task.
|
||||
*/
|
||||
final case class Task[A](info: Info[A], work: Action[A]):
|
||||
final case class Task[A](info: Info[A], work: Action[A]) extends TaskId[A]:
|
||||
override def toString = info.name getOrElse ("Task(" + info + ")")
|
||||
override def hashCode = info.hashCode
|
||||
|
||||
|
|
@ -28,7 +28,10 @@ final case class Task[A](info: Info[A], work: Action[A]):
|
|||
withInfo(info = nextInfo)
|
||||
}
|
||||
|
||||
def tags: TagMap = info get tagsKey getOrElse TagMap.empty
|
||||
def tags: TagMap = info.get(tagsKey).getOrElse(TagMap.empty)
|
||||
def name: Option[String] = info.name
|
||||
|
||||
def attributes: AttributeMap = info.attributes
|
||||
|
||||
private[sbt] def withInfo(info: Info[A]): Task[A] =
|
||||
Task(info = info, work = this.work)
|
||||
|
|
|
|||
|
|
@ -27,11 +27,11 @@ enum Action[A]:
|
|||
// private[sbt] def mapTask(f: [A1] => Task[A1] => Task[A1]) = this
|
||||
|
||||
/** Applies a function to the result of evaluating a heterogeneous list of other tasks. */
|
||||
case Mapped[A, K[F[_]]](in: K[Task], f: K[Result] => A, alist: AList[K]) extends Action[A]
|
||||
case Mapped[A, K[+F[_]]](in: K[Task], f: K[Result] => A, alist: AList[K]) extends Action[A]
|
||||
// private[sbt] def mapTask(g: Task ~> Task) = Mapped[A, K](alist.transform(in, g), f, alist)
|
||||
|
||||
/** Computes another task to evaluate based on results from evaluating other tasks. */
|
||||
case FlatMapped[A, K[F[_]]](
|
||||
case FlatMapped[A, K[+F[_]]](
|
||||
in: K[Task],
|
||||
f: K[Result] => Task[A],
|
||||
alist: AList[K],
|
||||
|
|
@ -63,6 +63,7 @@ enum Action[A]:
|
|||
end Action
|
||||
|
||||
object Action:
|
||||
import sbt.std.TaskExtra.*
|
||||
|
||||
/**
|
||||
* Encode this computation as a flatMap.
|
||||
|
|
@ -72,12 +73,12 @@ object Action:
|
|||
): Action.FlatMapped[A2, [F[_]] =>> Tuple1[F[Either[A1, A2]]]] =
|
||||
val alist = AList.tuple[Tuple1[Either[A1, A2]]]
|
||||
val f: Either[A1, A2] => Task[A2] = {
|
||||
case Right(b) => std.TaskExtra.task(b)
|
||||
case Left(a) => std.TaskExtra.singleInputTask(s.fin).map(_(a))
|
||||
case Right(b) => task(b)
|
||||
case Left(a) => singleInputTask(s.fin).map(_(a))
|
||||
}
|
||||
Action.FlatMapped[A2, [F[_]] =>> Tuple1[F[Either[A1, A2]]]](
|
||||
Tuple1(s.fab),
|
||||
{ case Tuple1(r) => (f compose std.TaskExtra.successM)(r) },
|
||||
{ case Tuple1(r) => f.compose(successM)(r) },
|
||||
alist,
|
||||
)
|
||||
end Action
|
||||
|
|
|
|||
|
|
@ -144,7 +144,7 @@ trait TaskExtra extends TaskExtra0 {
|
|||
: Conversion[(Task[A1], Task[A2]), MultiInTask[[F[_]] =>> Tuple.Map[(A1, A2), F]]] =
|
||||
multT2Task(_)
|
||||
|
||||
final implicit def multInputTask[K[F[_]]: AList](tasks: K[Task]): MultiInTask[K] =
|
||||
final implicit def multInputTask[K[+F[_]]: AList](tasks: K[Task]): MultiInTask[K] =
|
||||
new MultiInTask[K]:
|
||||
override def flatMapN[A](f: K[Id] => Task[A]): Task[A] =
|
||||
Task(Info(), Action.FlatMapped[A, K](tasks, f compose allM, AList[K]))
|
||||
|
|
|
|||
|
|
@ -30,8 +30,8 @@ object Transform:
|
|||
|
||||
final class TaskAndValue[T](val task: Task[T], val value: T)
|
||||
|
||||
def dummyMap(dummyMap: DummyTaskMap): Task ~>| Task = {
|
||||
val pmap = new DelegatingPMap[Task, Task](new collection.mutable.ListMap)
|
||||
def dummyMap(dummyMap: DummyTaskMap): TaskId ~>| Task = {
|
||||
val pmap = new DelegatingPMap[TaskId, Task](new collection.mutable.ListMap)
|
||||
def add[T](dummy: TaskAndValue[T]): Unit = {
|
||||
pmap(dummy.task) = fromDummyStrict(dummy.task, dummy.value)
|
||||
}
|
||||
|
|
@ -40,15 +40,15 @@ object Transform:
|
|||
}
|
||||
|
||||
/** Applies `map`, returning the result if defined or returning the input unchanged otherwise. */
|
||||
implicit def getOrId(map: Task ~>| Task): [A] => Task[A] => Task[A] =
|
||||
[A] => (in: Task[A]) => map(in).getOrElse(in)
|
||||
implicit def getOrId(map: TaskId ~>| Task): [A] => TaskId[A] => Task[A] =
|
||||
[A] => (in: TaskId[A]) => map(in).getOrElse(in.asInstanceOf)
|
||||
|
||||
def apply(dummies: DummyTaskMap) = taskToNode(getOrId(dummyMap(dummies)))
|
||||
|
||||
def taskToNode(pre: [A] => Task[A] => Task[A]): NodeView[Task] =
|
||||
new NodeView[Task]:
|
||||
def taskToNode(pre: [A] => TaskId[A] => Task[A]): NodeView =
|
||||
new NodeView:
|
||||
import Action.*
|
||||
def apply[T](t: Task[T]): Node[Task, T] = pre(t).work match
|
||||
def apply[T](t: TaskId[T]): Node[T] = pre(t).work match
|
||||
case Pure(eval, _) => uniform(Nil)(_ => Right(eval()))
|
||||
case m: Mapped[a, k] => toNode[a, k](m.in)(right[a] compose m.f)(m.alist)
|
||||
case m: FlatMapped[a, k] =>
|
||||
|
|
@ -61,19 +61,19 @@ object Transform:
|
|||
case DependsOn(in, deps) => uniform(existToAny(deps))(const(Left(in)) compose all)
|
||||
case Join(in, f) => uniform(in)(f)
|
||||
|
||||
def inline1[T](t: Task[T]): Option[() => T] = t.work match
|
||||
case Action.Pure(eval, true) => Some(eval)
|
||||
case _ => None
|
||||
def inline1[T](t: TaskId[T]): Option[() => T] = t match
|
||||
case Task(_, Action.Pure(eval, true)) => Some(eval)
|
||||
case _ => None
|
||||
|
||||
def uniform[A1, D](tasks: Seq[Task[D]])(
|
||||
f: Seq[Result[D]] => Either[Task[A1], A1]
|
||||
): Node[Task, A1] =
|
||||
): Node[A1] =
|
||||
toNode[A1, [F[_]] =>> List[F[D]]](tasks.toList)(f)(AList.list[D])
|
||||
|
||||
def toNode[A1, K1[F[_]]: AList](
|
||||
inputs: K1[Task]
|
||||
)(f: K1[Result] => Either[Task[A1], A1]): Node[Task, A1] =
|
||||
new Node[Task, A1]:
|
||||
inputs: K1[TaskId]
|
||||
)(f: K1[Result] => Either[Task[A1], A1]): Node[A1] =
|
||||
new Node[A1]:
|
||||
type K[F[_]] = K1[F]
|
||||
val in = inputs
|
||||
lazy val alist: AList[K] = AList[K]
|
||||
|
|
|
|||
|
|
@ -22,15 +22,15 @@ object TaskGen extends std.TaskExtra {
|
|||
val TaskListGen = MaxTasksGen.flatMap(size => Gen.listOfN(size, Arbitrary.arbInt.arbitrary))
|
||||
|
||||
def run[T](root: Task[T], checkCycles: Boolean, maxWorkers: Int): Result[T] = {
|
||||
val (service, shutdown) = CompletionService[Task[_], Completed](maxWorkers)
|
||||
val (service, shutdown) = CompletionService(maxWorkers)
|
||||
val dummies = std.Transform.DummyTaskMap(Nil)
|
||||
val x = new Execute[Task](
|
||||
val x = new Execute(
|
||||
Execute.config(checkCycles),
|
||||
Execute.noTriggers,
|
||||
ExecuteProgress.empty[Task]
|
||||
ExecuteProgress.empty
|
||||
)(using std.Transform(dummies))
|
||||
try {
|
||||
x.run(root)(using service.asInstanceOf)
|
||||
x.run(root)(using service)
|
||||
} finally {
|
||||
shutdown()
|
||||
}
|
||||
|
|
|
|||
|
|
@ -50,5 +50,5 @@ object TaskRunnerCircularTest extends Properties("TaskRunner Circular") {
|
|||
def cyclic(i: Incomplete) =
|
||||
Incomplete
|
||||
.allExceptions(i)
|
||||
.exists(_.isInstanceOf[Execute[({ type A[_] <: AnyRef })#A @unchecked]#CyclicException[_]])
|
||||
.exists(_.isInstanceOf[Execute#CyclicException])
|
||||
}
|
||||
|
|
|
|||
|
|
@ -16,7 +16,7 @@ import TaskGen.MaxWorkers
|
|||
import org.scalacheck._
|
||||
import Prop.forAll
|
||||
import Transform.taskToNode
|
||||
import ConcurrentRestrictions.{ completionService, limitTotal, tagged => tagged0, TagMap }
|
||||
import ConcurrentRestrictions.{ completionService, limitTotal, tagged }
|
||||
|
||||
import java.util.concurrent.{ CountDownLatch, TimeUnit }
|
||||
|
||||
|
|
@ -52,7 +52,7 @@ object TaskSerial extends Properties("task serial") {
|
|||
|
||||
def checkArbitrary(
|
||||
size: Int,
|
||||
restrictions: ConcurrentRestrictions[Task[_]],
|
||||
restrictions: ConcurrentRestrictions,
|
||||
shouldSucceed: Boolean
|
||||
) = {
|
||||
val latch = task { new CountDownLatch(size) }
|
||||
|
|
@ -70,8 +70,7 @@ object TaskSerial extends Properties("task serial") {
|
|||
"Some tasks were unschedulable: verify this is an actual failure by extending the timeout to several seconds."
|
||||
def scheduledMsg = "All tasks were unexpectedly scheduled."
|
||||
|
||||
def tagged(f: TagMap => Boolean) = tagged0[Task[_]](_.tags, f)
|
||||
def evalRestricted[T](t: Task[T])(restrictions: ConcurrentRestrictions[Task[_]]): T =
|
||||
def evalRestricted[T](t: Task[T])(restrictions: ConcurrentRestrictions): T =
|
||||
tryRun[T](t, checkCycles, restrictions)
|
||||
}
|
||||
|
||||
|
|
@ -79,18 +78,18 @@ object TaskTest {
|
|||
def run[T](
|
||||
root: Task[T],
|
||||
checkCycles: Boolean,
|
||||
restrictions: ConcurrentRestrictions[Task[_]]
|
||||
restrictions: ConcurrentRestrictions
|
||||
): Result[T] = {
|
||||
val (service, shutdown) =
|
||||
completionService[Task[_], Completed](restrictions, (x: String) => System.err.println(x))
|
||||
completionService(restrictions, (x: String) => System.err.println(x))
|
||||
|
||||
val x = new Execute[Task](
|
||||
val x = new Execute(
|
||||
Execute.config(checkCycles),
|
||||
Execute.noTriggers,
|
||||
ExecuteProgress.empty[Task]
|
||||
)(using taskToNode(idK[Task]))
|
||||
ExecuteProgress.empty
|
||||
)(using taskToNode([A] => (id: TaskId[A]) => id.asInstanceOf))
|
||||
try {
|
||||
x.run(root)(using service.asInstanceOf)
|
||||
x.run(root)(using service)
|
||||
} finally {
|
||||
shutdown()
|
||||
}
|
||||
|
|
@ -98,7 +97,7 @@ object TaskTest {
|
|||
def tryRun[T](
|
||||
root: Task[T],
|
||||
checkCycles: Boolean,
|
||||
restrictions: ConcurrentRestrictions[Task[_]]
|
||||
restrictions: ConcurrentRestrictions
|
||||
): T =
|
||||
run(root, checkCycles, restrictions) match {
|
||||
case Result.Value(v) => v
|
||||
|
|
|
|||
|
|
@ -7,19 +7,19 @@
|
|||
|
||||
package sbt
|
||||
|
||||
trait CompletionService[A, R]:
|
||||
trait CompletionService:
|
||||
|
||||
/**
|
||||
* Submits a work node A with work that returns R. In Execute this is used for tasks returning
|
||||
* sbt.Completed.
|
||||
*/
|
||||
def submit(node: A, work: () => R): Unit
|
||||
def submit(node: TaskId[?], work: () => Completed): Unit
|
||||
|
||||
/**
|
||||
* Retrieves and removes the result from the next completed task, waiting if none are yet present.
|
||||
* In Execute this is used for tasks returning sbt.Completed.
|
||||
*/
|
||||
def take(): R
|
||||
def take(): Completed
|
||||
end CompletionService
|
||||
|
||||
import java.util.concurrent.atomic.AtomicInteger
|
||||
|
|
@ -35,35 +35,40 @@ import java.util.concurrent.{
|
|||
|
||||
object CompletionService {
|
||||
val poolID = new AtomicInteger(1)
|
||||
def apply[A, T](poolSize: Int): (CompletionService[A, T], () => Unit) = {
|
||||
def apply(poolSize: Int): (CompletionService, () => Unit) = {
|
||||
val i = new AtomicInteger(1)
|
||||
val id = poolID.getAndIncrement()
|
||||
val pool = Executors.newFixedThreadPool(
|
||||
poolSize,
|
||||
(r: Runnable) => new Thread(r, s"sbt-completion-thread-$id-${i.getAndIncrement}")
|
||||
)
|
||||
(apply[A, T](pool), () => { pool.shutdownNow(); () })
|
||||
(apply(pool), () => { pool.shutdownNow(); () })
|
||||
}
|
||||
|
||||
def apply[A, T](x: Executor): CompletionService[A, T] =
|
||||
apply(new ExecutorCompletionService[T](x))
|
||||
def apply(x: Executor): CompletionService =
|
||||
apply(new ExecutorCompletionService[Completed](x))
|
||||
|
||||
def apply[A, T](completion: JCompletionService[T]): CompletionService[A, T] =
|
||||
new CompletionService[A, T] {
|
||||
def submit(node: A, work: () => T) = { CompletionService.submit(work, completion); () }
|
||||
def apply(completion: JCompletionService[Completed]): CompletionService =
|
||||
new CompletionService {
|
||||
def submit(node: TaskId[?], work: () => Completed) = {
|
||||
CompletionService.submit(work, completion); ()
|
||||
}
|
||||
def take() = completion.take().get()
|
||||
}
|
||||
|
||||
def submit[T](work: () => T, completion: JCompletionService[T]): () => T = {
|
||||
val future = submitFuture[T](work, completion)
|
||||
def submit(work: () => Completed, completion: JCompletionService[Completed]): () => Completed = {
|
||||
val future = submitFuture(work, completion)
|
||||
() => future.get
|
||||
}
|
||||
|
||||
private[sbt] def submitFuture[A](work: () => A, completion: JCompletionService[A]): JFuture[A] = {
|
||||
private[sbt] def submitFuture(
|
||||
work: () => Completed,
|
||||
completion: JCompletionService[Completed]
|
||||
): JFuture[Completed] = {
|
||||
val future =
|
||||
try
|
||||
completion.submit {
|
||||
new Callable[A] {
|
||||
new Callable[Completed] {
|
||||
def call =
|
||||
try {
|
||||
work()
|
||||
|
|
@ -79,9 +84,9 @@ object CompletionService {
|
|||
}
|
||||
future
|
||||
}
|
||||
def manage[A, T](
|
||||
service: CompletionService[A, T]
|
||||
)(setup: A => Unit, cleanup: A => Unit): CompletionService[A, T] =
|
||||
def manage(
|
||||
service: CompletionService
|
||||
)(setup: TaskId[?] => Unit, cleanup: TaskId[?] => Unit): CompletionService =
|
||||
wrap(service) { (node, work) => () =>
|
||||
setup(node)
|
||||
try {
|
||||
|
|
@ -90,11 +95,11 @@ object CompletionService {
|
|||
cleanup(node)
|
||||
}
|
||||
}
|
||||
def wrap[A, T](
|
||||
service: CompletionService[A, T]
|
||||
)(w: (A, () => T) => (() => T)): CompletionService[A, T] =
|
||||
new CompletionService[A, T] {
|
||||
def submit(node: A, work: () => T) = service.submit(node, w(node, work))
|
||||
def wrap(
|
||||
service: CompletionService
|
||||
)(w: (TaskId[?], () => Completed) => (() => Completed)): CompletionService =
|
||||
new CompletionService {
|
||||
def submit(node: TaskId[?], work: () => Completed) = service.submit(node, w(node, work))
|
||||
def take() = service.take()
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -13,14 +13,12 @@ import sbt.internal.util.AttributeKey
|
|||
import java.util.concurrent.atomic.AtomicBoolean
|
||||
import java.util.concurrent.{ Future => JFuture, RejectedExecutionException }
|
||||
import scala.collection.mutable
|
||||
import scala.jdk.CollectionConverters.*
|
||||
|
||||
/**
|
||||
* Describes restrictions on concurrent execution for a set of tasks.
|
||||
*
|
||||
* @tparam A
|
||||
* the type of a task
|
||||
*/
|
||||
trait ConcurrentRestrictions[A] {
|
||||
trait ConcurrentRestrictions {
|
||||
|
||||
/** Internal state type used to describe a set of tasks. */
|
||||
type G
|
||||
|
|
@ -29,10 +27,10 @@ trait ConcurrentRestrictions[A] {
|
|||
def empty: G
|
||||
|
||||
/** Updates the description `g` to include a new task `a`. */
|
||||
def add(g: G, a: A): G
|
||||
def add(g: G, a: TaskId[?]): G
|
||||
|
||||
/** Updates the description `g` to remove a previously added task `a`. */
|
||||
def remove(g: G, a: A): G
|
||||
def remove(g: G, a: TaskId[?]): G
|
||||
|
||||
/**
|
||||
* Returns true if the tasks described by `g` are allowed to execute concurrently. The methods in
|
||||
|
|
@ -54,8 +52,7 @@ import java.util.concurrent.{ Executor, Executors, ExecutorCompletionService }
|
|||
import annotation.tailrec
|
||||
|
||||
object ConcurrentRestrictions {
|
||||
private[this] val completionServices = new java.util.WeakHashMap[CompletionService[_, _], Boolean]
|
||||
import scala.collection.JavaConverters._
|
||||
private[this] val completionServices = new java.util.WeakHashMap[CompletionService, Boolean]
|
||||
def cancelAll() = completionServices.keySet.asScala.toVector.foreach {
|
||||
case a: AutoCloseable => a.close()
|
||||
case _ =>
|
||||
|
|
@ -71,22 +68,22 @@ object ConcurrentRestrictions {
|
|||
* @param zero
|
||||
* the constant placeholder used for t
|
||||
*/
|
||||
def unrestricted[A]: ConcurrentRestrictions[A] =
|
||||
new ConcurrentRestrictions[A] {
|
||||
def unrestricted: ConcurrentRestrictions =
|
||||
new ConcurrentRestrictions {
|
||||
type G = Unit
|
||||
def empty = ()
|
||||
def add(g: G, a: A) = ()
|
||||
def remove(g: G, a: A) = ()
|
||||
def add(g: G, a: TaskId[?]) = ()
|
||||
def remove(g: G, a: TaskId[?]) = ()
|
||||
def valid(g: G) = true
|
||||
}
|
||||
|
||||
def limitTotal[A](i: Int): ConcurrentRestrictions[A] = {
|
||||
def limitTotal(i: Int): ConcurrentRestrictions = {
|
||||
assert(i >= 1, "Maximum must be at least 1 (was " + i + ")")
|
||||
new ConcurrentRestrictions[A] {
|
||||
new ConcurrentRestrictions {
|
||||
type G = Int
|
||||
def empty = 0
|
||||
def add(g: Int, a: A) = g + 1
|
||||
def remove(g: Int, a: A) = g - 1
|
||||
def add(g: Int, a: TaskId[?]) = g + 1
|
||||
def remove(g: Int, a: TaskId[?]) = g - 1
|
||||
def valid(g: Int) = g <= i
|
||||
}
|
||||
}
|
||||
|
|
@ -108,26 +105,25 @@ object ConcurrentRestrictions {
|
|||
|
||||
/**
|
||||
* Implements concurrency restrictions on tasks based on Tags.
|
||||
* @tparam A
|
||||
* type of a task
|
||||
* @param get
|
||||
* extracts tags from a task
|
||||
* @param validF
|
||||
* defines whether a set of tasks are allowed to execute concurrently based on their merged tags
|
||||
*/
|
||||
def tagged[A](get: A => TagMap, validF: TagMap => Boolean): ConcurrentRestrictions[A] =
|
||||
new ConcurrentRestrictions[A] {
|
||||
def tagged(validF: TagMap => Boolean): ConcurrentRestrictions =
|
||||
new ConcurrentRestrictions {
|
||||
type G = TagMap
|
||||
def empty = Map.empty
|
||||
def add(g: TagMap, a: A) = merge(g, a, get)(_ + _)
|
||||
def remove(g: TagMap, a: A) = merge(g, a, get)(_ - _)
|
||||
def add(g: TagMap, a: TaskId[?]) = merge(g, a)(_ + _)
|
||||
def remove(g: TagMap, a: TaskId[?]) = merge(g, a)(_ - _)
|
||||
def valid(g: TagMap) = validF(g)
|
||||
}
|
||||
|
||||
private[this] def merge[A](m: TagMap, a: A, get: A => TagMap)(f: (Int, Int) => Int): TagMap = {
|
||||
val aTags = get(a)
|
||||
val base = merge(m, aTags)(f)
|
||||
val un = if (aTags.isEmpty) update(base, Untagged, 1)(f) else base
|
||||
private[this] def merge(m: TagMap, a: TaskId[?])(
|
||||
f: (Int, Int) => Int
|
||||
): TagMap = {
|
||||
val base = merge(m, a.tags)(f)
|
||||
val un = if (a.tags.isEmpty) update(base, Untagged, 1)(f) else base
|
||||
update(un, All, 1)(f)
|
||||
}
|
||||
|
||||
|
|
@ -154,26 +150,26 @@ object ConcurrentRestrictions {
|
|||
* @tparam R
|
||||
* the type of data that will be computed by the CompletionService.
|
||||
*/
|
||||
def completionService[A, R](
|
||||
tags: ConcurrentRestrictions[A],
|
||||
def completionService(
|
||||
tags: ConcurrentRestrictions,
|
||||
warn: String => Unit
|
||||
): (CompletionService[A, R], () => Unit) = {
|
||||
): (CompletionService, () => Unit) = {
|
||||
val id = poolID.getAndIncrement
|
||||
val i = new AtomicInteger(1)
|
||||
val pool = Executors.newCachedThreadPool { r =>
|
||||
new Thread(r, s"sbt-completion-service-pool-$id-${i.getAndIncrement()}")
|
||||
}
|
||||
val service = completionService[A, R](pool, tags, warn)
|
||||
val service = completionService(pool, tags, warn)
|
||||
(service, () => { pool.shutdownNow(); () })
|
||||
}
|
||||
|
||||
def completionService[A, R](
|
||||
tags: ConcurrentRestrictions[A],
|
||||
def completionService(
|
||||
tags: ConcurrentRestrictions,
|
||||
warn: String => Unit,
|
||||
isSentinel: A => Boolean
|
||||
): (CompletionService[A, R], () => Unit) = {
|
||||
isSentinel: TaskId[?] => Boolean
|
||||
): (CompletionService, () => Unit) = {
|
||||
val pool = Executors.newCachedThreadPool()
|
||||
val service = completionService[A, R](pool, tags, warn, isSentinel)
|
||||
val service = completionService(pool, tags, warn, isSentinel)
|
||||
(
|
||||
service,
|
||||
() => {
|
||||
|
|
@ -183,13 +179,13 @@ object ConcurrentRestrictions {
|
|||
)
|
||||
}
|
||||
|
||||
def cancellableCompletionService[A, R](
|
||||
tags: ConcurrentRestrictions[A],
|
||||
def cancellableCompletionService(
|
||||
tags: ConcurrentRestrictions,
|
||||
warn: String => Unit,
|
||||
isSentinel: A => Boolean
|
||||
): (CompletionService[A, R], Boolean => Unit) = {
|
||||
isSentinel: TaskId[?] => Boolean
|
||||
): (CompletionService, Boolean => Unit) = {
|
||||
val pool = Executors.newCachedThreadPool()
|
||||
val service = completionService[A, R](pool, tags, warn, isSentinel)
|
||||
val service = completionService(pool, tags, warn, isSentinel)
|
||||
(
|
||||
service,
|
||||
force => {
|
||||
|
|
@ -200,12 +196,12 @@ object ConcurrentRestrictions {
|
|||
)
|
||||
}
|
||||
|
||||
def completionService[A, R](
|
||||
def completionService(
|
||||
backing: Executor,
|
||||
tags: ConcurrentRestrictions[A],
|
||||
tags: ConcurrentRestrictions,
|
||||
warn: String => Unit
|
||||
): CompletionService[A, R] with AutoCloseable = {
|
||||
completionService[A, R](backing, tags, warn, (_: A) => false)
|
||||
): CompletionService with AutoCloseable = {
|
||||
completionService(backing, tags, warn, _ => false)
|
||||
}
|
||||
|
||||
/**
|
||||
|
|
@ -213,17 +209,17 @@ object ConcurrentRestrictions {
|
|||
* restrictions on concurrent task execution and using the provided Executor to manage execution
|
||||
* on threads.
|
||||
*/
|
||||
def completionService[A, R](
|
||||
def completionService(
|
||||
backing: Executor,
|
||||
tags: ConcurrentRestrictions[A],
|
||||
tags: ConcurrentRestrictions,
|
||||
warn: String => Unit,
|
||||
isSentinel: A => Boolean,
|
||||
): CompletionService[A, R] with CancelSentiels with AutoCloseable = {
|
||||
isSentinel: TaskId[?] => Boolean,
|
||||
): CompletionService with CancelSentiels with AutoCloseable = {
|
||||
|
||||
// Represents submitted work for a task.
|
||||
final class Enqueue(val node: A, val work: () => R)
|
||||
final class Enqueue(val node: TaskId[?], val work: () => Completed)
|
||||
|
||||
new CompletionService[A, R] with CancelSentiels with AutoCloseable {
|
||||
new CompletionService with CancelSentiels with AutoCloseable {
|
||||
completionServices.put(this, true)
|
||||
private[this] val closed = new AtomicBoolean(false)
|
||||
override def close(): Unit = if (closed.compareAndSet(false, true)) {
|
||||
|
|
@ -232,7 +228,7 @@ object ConcurrentRestrictions {
|
|||
}
|
||||
|
||||
/** Backing service used to manage execution on threads once all constraints are satisfied. */
|
||||
private[this] val jservice = new ExecutorCompletionService[R](backing)
|
||||
private[this] val jservice = new ExecutorCompletionService[Completed](backing)
|
||||
|
||||
/** The description of the currently running tasks, used by `tags` to manage restrictions. */
|
||||
private[this] var tagState = tags.empty
|
||||
|
|
@ -255,7 +251,7 @@ object ConcurrentRestrictions {
|
|||
sentinels.clear()
|
||||
}
|
||||
|
||||
def submit(node: A, work: () => R): Unit = synchronized {
|
||||
def submit(node: TaskId[?], work: () => Completed): Unit = synchronized {
|
||||
if (closed.get) throw new RejectedExecutionException
|
||||
else if (isSentinel(node)) {
|
||||
// skip all checks for sentinels
|
||||
|
|
@ -276,7 +272,7 @@ object ConcurrentRestrictions {
|
|||
}
|
||||
()
|
||||
}
|
||||
private[this] def submitValid(node: A, work: () => R): Unit = {
|
||||
private[this] def submitValid(node: TaskId[?], work: () => Completed): Unit = {
|
||||
running += 1
|
||||
val wrappedWork = () =>
|
||||
try work()
|
||||
|
|
@ -284,7 +280,7 @@ object ConcurrentRestrictions {
|
|||
CompletionService.submitFuture(wrappedWork, jservice)
|
||||
()
|
||||
}
|
||||
private[this] def cleanup(node: A): Unit = synchronized {
|
||||
private[this] def cleanup(node: TaskId[?]): Unit = synchronized {
|
||||
running -= 1
|
||||
tagState = tags.remove(tagState, node)
|
||||
if (!tags.valid(tagState)) {
|
||||
|
|
@ -320,7 +316,7 @@ object ConcurrentRestrictions {
|
|||
submitValid(tried)
|
||||
}
|
||||
|
||||
def take(): R = {
|
||||
def take(): Completed = {
|
||||
if (closed.get)
|
||||
throw new RejectedExecutionException(
|
||||
"Tried to get values for a closed completion service"
|
||||
|
|
|
|||
|
|
@ -11,23 +11,24 @@ import java.util.concurrent.ExecutionException
|
|||
|
||||
import sbt.internal.util.ErrorHandling.wideConvert
|
||||
import sbt.internal.util.{ DelegatingPMap, IDSet, PMap, RMap, ~> }
|
||||
import sbt.internal.util.Types._
|
||||
import sbt.internal.util.Types.*
|
||||
import sbt.internal.util.Util.nilSeq
|
||||
import Execute._
|
||||
|
||||
import scala.annotation.tailrec
|
||||
import scala.collection.mutable
|
||||
import scala.collection.JavaConverters._
|
||||
import scala.jdk.CollectionConverters.*
|
||||
import mutable.Map
|
||||
import sbt.internal.util.AList
|
||||
|
||||
private[sbt] object Execute {
|
||||
def idMap[A1, A2]: Map[A1, A2] = (new java.util.IdentityHashMap[A1, A2]).asScala
|
||||
def pMap[F1[_], F2[_]]: PMap[F1, F2] = new DelegatingPMap[F1, F2](idMap)
|
||||
def taskMap[A]: Map[TaskId[?], A] = (new java.util.IdentityHashMap[TaskId[?], A]).asScala
|
||||
def taskPMap[F[_]]: PMap[TaskId, F] = new DelegatingPMap(
|
||||
(new java.util.IdentityHashMap[TaskId[Any], F[Any]]).asScala
|
||||
)
|
||||
private[sbt] def completed(p: => Unit): Completed = new Completed {
|
||||
def process(): Unit = p
|
||||
}
|
||||
def noTriggers[F[_]] = new Triggers[F](Map.empty, Map.empty, idFun)
|
||||
def noTriggers[TaskId[_]] = new Triggers(Map.empty, Map.empty, idFun)
|
||||
|
||||
def config(checkCycles: Boolean, overwriteNode: Incomplete => Boolean = const(false)): Config =
|
||||
new Config(checkCycles, overwriteNode)
|
||||
|
|
@ -44,33 +45,33 @@ sealed trait Completed {
|
|||
def process(): Unit
|
||||
}
|
||||
|
||||
private[sbt] trait NodeView[F[_]] {
|
||||
def apply[A](a: F[A]): Node[F, A]
|
||||
def inline1[A](a: F[A]): Option[() => A]
|
||||
private[sbt] trait NodeView {
|
||||
def apply[A](a: TaskId[A]): Node[A]
|
||||
def inline1[A](a: TaskId[A]): Option[() => A]
|
||||
}
|
||||
|
||||
final class Triggers[F[_]](
|
||||
val runBefore: collection.Map[F[Any], Seq[F[Any]]],
|
||||
val injectFor: collection.Map[F[Any], Seq[F[Any]]],
|
||||
val onComplete: RMap[F, Result] => RMap[F, Result],
|
||||
final class Triggers(
|
||||
val runBefore: collection.Map[TaskId[?], Seq[TaskId[?]]],
|
||||
val injectFor: collection.Map[TaskId[?], Seq[TaskId[?]]],
|
||||
val onComplete: RMap[TaskId, Result] => RMap[TaskId, Result],
|
||||
)
|
||||
|
||||
private[sbt] final class Execute[F[_] <: AnyRef](
|
||||
config: Config,
|
||||
triggers: Triggers[F],
|
||||
progress: ExecuteProgress[F]
|
||||
)(using view: NodeView[F]) {
|
||||
type Strategy = CompletionService[F[Any], Completed]
|
||||
private[sbt] final class Execute(
|
||||
config: Execute.Config,
|
||||
triggers: Triggers,
|
||||
progress: ExecuteProgress
|
||||
)(using view: NodeView) {
|
||||
import Execute.*
|
||||
|
||||
private[this] val forward = idMap[F[Any], IDSet[F[Any]]]
|
||||
private[this] val reverse = idMap[F[Any], Iterable[F[Any]]]
|
||||
private[this] val callers = pMap[F, Compose[IDSet, F]]
|
||||
private[this] val state = idMap[F[Any], State]
|
||||
private[this] val viewCache = pMap[F, Node[F, *]]
|
||||
private[this] val results = pMap[F, Result]
|
||||
private[this] val forward = taskMap[IDSet[TaskId[?]]]
|
||||
private[this] val reverse = taskMap[Iterable[TaskId[?]]]
|
||||
private[this] val callers = taskPMap[[X] =>> IDSet[TaskId[X]]]
|
||||
private[this] val state = taskMap[State]
|
||||
private[this] val viewCache = taskPMap[Node]
|
||||
private[this] val results = taskPMap[Result]
|
||||
|
||||
private[this] val getResult: [A] => F[A] => Result[A] = [A] =>
|
||||
(a: F[A]) =>
|
||||
private[this] val getResult: [A] => TaskId[A] => Result[A] = [A] =>
|
||||
(a: TaskId[A]) =>
|
||||
view.inline1(a) match
|
||||
case Some(v) => Result.Value(v())
|
||||
case None => results(a)
|
||||
|
|
@ -85,12 +86,12 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
def dump: String =
|
||||
"State: " + state.toString + "\n\nResults: " + results + "\n\nCalls: " + callers + "\n\n"
|
||||
|
||||
def run[A](root: F[A])(using strategy: Strategy): Result[A] =
|
||||
def run[A](root: TaskId[A])(using strategy: CompletionService): Result[A] =
|
||||
try {
|
||||
runKeep(root)(root)
|
||||
} catch { case i: Incomplete => Result.Inc(i) }
|
||||
|
||||
def runKeep[A](root: F[A])(using strategy: Strategy): RMap[F, Result] = {
|
||||
def runKeep[A](root: TaskId[A])(using strategy: CompletionService): RMap[TaskId, Result] = {
|
||||
assert(state.isEmpty, "Execute already running/ran.")
|
||||
|
||||
addNew(root)
|
||||
|
|
@ -102,7 +103,7 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
finalResults
|
||||
}
|
||||
|
||||
def processAll()(using strategy: Strategy): Unit = {
|
||||
def processAll()(using strategy: CompletionService): Unit = {
|
||||
@tailrec def next(): Unit = {
|
||||
pre {
|
||||
assert(reverse.nonEmpty, "Nothing to process.")
|
||||
|
|
@ -135,7 +136,7 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
}
|
||||
def dumpCalling: String = state.filter(_._2 == Calling).mkString("\n\t")
|
||||
|
||||
def call[A](node: F[A], target: F[A])(using strategy: Strategy): Unit = {
|
||||
def call[A](node: TaskId[A], target: TaskId[A])(using strategy: CompletionService): Unit = {
|
||||
if (config.checkCycles) cycleCheck(node, target)
|
||||
pre {
|
||||
assert(running(node))
|
||||
|
|
@ -145,7 +146,7 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
results.get(target) match {
|
||||
case Some(result) => retire(node, result)
|
||||
case None =>
|
||||
state(node.asInstanceOf) = Calling
|
||||
state(node) = Calling
|
||||
addChecked(target)
|
||||
addCaller(node, target)
|
||||
}
|
||||
|
|
@ -160,18 +161,16 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
}
|
||||
}
|
||||
|
||||
def retire[A](node: F[A], result: Result[A])(using strategy: Strategy): Unit = {
|
||||
def retire[A](node: TaskId[A], result: Result[A])(using strategy: CompletionService): Unit = {
|
||||
pre {
|
||||
assert(running(node) | calling(node))
|
||||
readyInv(node)
|
||||
}
|
||||
|
||||
results(node) = result
|
||||
state(node.asInstanceOf) = Done
|
||||
state(node) = Done
|
||||
progress.afterCompleted(node, result)
|
||||
remove(reverse.asInstanceOf[Map[F[A], Iterable[F[Any]]]], node) foreach { dep =>
|
||||
notifyDone(node, dep.asInstanceOf)
|
||||
}
|
||||
remove(reverse, node).foreach(dep => notifyDone(node, dep))
|
||||
callers.remove(node).toList.flatten.foreach { c =>
|
||||
retire(c, callerResult(c, result))
|
||||
}
|
||||
|
|
@ -183,23 +182,23 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
assert(done(node))
|
||||
assert(results(node) == result)
|
||||
readyInv(node)
|
||||
assert(!(reverse.contains(node.asInstanceOf)))
|
||||
assert(!(reverse.contains(node)))
|
||||
assert(!(callers.contains(node)))
|
||||
assert(triggeredBy(node) forall added)
|
||||
}
|
||||
}
|
||||
def callerResult[A](node: F[A], result: Result[A]): Result[A] =
|
||||
def callerResult[A](node: TaskId[A], result: Result[A]): Result[A] =
|
||||
result match {
|
||||
case _: Result.Value[A] => result
|
||||
case Result.Inc(i) => Result.Inc(Incomplete(Some(node), tpe = i.tpe, causes = i :: Nil))
|
||||
}
|
||||
|
||||
def notifyDone[A](node: F[A], dependent: F[Any])(using strategy: Strategy): Unit = {
|
||||
def notifyDone(node: TaskId[?], dependent: TaskId[?])(using strategy: CompletionService): Unit = {
|
||||
val f = forward(dependent)
|
||||
f -= node.asInstanceOf
|
||||
f -= node
|
||||
if (f.isEmpty) {
|
||||
remove[F[Any], IDSet[F[Any]]](forward.asInstanceOf, dependent)
|
||||
ready[Any](dependent)
|
||||
remove(forward, dependent)
|
||||
ready(dependent)
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -208,7 +207,7 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
* inputs and dependencies have completed. Its computation is then evaluated and made available
|
||||
* for nodes that have it as an input.
|
||||
*/
|
||||
def addChecked[A](node: F[A])(using strategy: Strategy): Unit = {
|
||||
def addChecked[A](node: TaskId[A])(using strategy: CompletionService): Unit = {
|
||||
if (!added(node)) addNew(node)
|
||||
|
||||
post { addedInv(node) }
|
||||
|
|
@ -219,14 +218,14 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
* have finished, the node's computation is scheduled to run. The node's dependencies will be
|
||||
* added (transitively) if they are not already registered.
|
||||
*/
|
||||
def addNew[A](node: F[A])(using strategy: Strategy): Unit = {
|
||||
def addNew(node: TaskId[?])(using strategy: CompletionService): Unit = {
|
||||
pre { newPre(node) }
|
||||
|
||||
val v = register(node)
|
||||
val deps: Iterable[F[Any]] = dependencies(v) ++ runBefore(node.asInstanceOf)
|
||||
val active = IDSet[F[Any]](deps filter notDone.asInstanceOf)
|
||||
val deps = dependencies(v) ++ runBefore(node)
|
||||
val active = IDSet[TaskId[?]](deps filter notDone)
|
||||
progress.afterRegistered(
|
||||
node.asInstanceOf,
|
||||
node,
|
||||
deps,
|
||||
active.toList
|
||||
/* active is mutable, so take a snapshot */
|
||||
|
|
@ -234,10 +233,10 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
|
||||
if (active.isEmpty) ready(node)
|
||||
else {
|
||||
forward(node.asInstanceOf) = active.asInstanceOf
|
||||
forward(node) = active
|
||||
for (a <- active) {
|
||||
addChecked[Any](a.asInstanceOf)
|
||||
addReverse[Any](a.asInstanceOf, node.asInstanceOf)
|
||||
addChecked(a)
|
||||
addReverse(a, node)
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -253,45 +252,47 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
* Called when a pending 'node' becomes runnable. All of its dependencies must be done. This
|
||||
* schedules the node's computation with 'strategy'.
|
||||
*/
|
||||
def ready[A](node: F[A])(using strategy: Strategy): Unit = {
|
||||
def ready(node: TaskId[?])(using strategy: CompletionService): Unit = {
|
||||
pre {
|
||||
assert(pending(node))
|
||||
readyInv(node)
|
||||
assert(reverse.contains(node.asInstanceOf))
|
||||
assert(reverse.contains(node))
|
||||
}
|
||||
|
||||
state(node.asInstanceOf) = Running
|
||||
progress.afterReady(node.asInstanceOf)
|
||||
state(node) = Running
|
||||
progress.afterReady(node)
|
||||
submit(node)
|
||||
|
||||
post {
|
||||
readyInv(node)
|
||||
assert(reverse.contains(node.asInstanceOf))
|
||||
assert(reverse.contains(node))
|
||||
assert(running(node))
|
||||
}
|
||||
}
|
||||
|
||||
/** Enters the given node into the system. */
|
||||
def register[A](node: F[A]): Node[F, A] = {
|
||||
state(node.asInstanceOf) = Pending
|
||||
reverse(node.asInstanceOf) = Seq()
|
||||
def register[A](node: TaskId[A]): Node[A] = {
|
||||
state(node) = Pending
|
||||
reverse(node) = Seq()
|
||||
viewCache.getOrUpdate(node, view(node))
|
||||
}
|
||||
|
||||
/** Send the work for this node to the provided Strategy. */
|
||||
def submit[A](node: F[A])(using strategy: Strategy): Unit = {
|
||||
def submit(node: TaskId[?])(using strategy: CompletionService): Unit = {
|
||||
val v = viewCache(node)
|
||||
val rs = v.alist.transform[F, Result](v.in)(getResult)
|
||||
val rs = v.alist.transform(v.in)(getResult)
|
||||
// v.alist.transform(v.in)(getResult)
|
||||
strategy.submit(node.asInstanceOf, () => work(node, v.work(rs)))
|
||||
strategy.submit(node, () => work(node, v.work(rs)))
|
||||
}
|
||||
|
||||
/**
|
||||
* Evaluates the computation 'f' for 'node'. This returns a Completed instance, which contains the
|
||||
* post-processing to perform after the result is retrieved from the Strategy.
|
||||
*/
|
||||
def work[A](node: F[A], f: => Either[F[A], A])(using strategy: Strategy): Completed = {
|
||||
progress.beforeWork(node.asInstanceOf)
|
||||
def work[A](node: TaskId[A], f: => Either[TaskId[A], A])(using
|
||||
strategy: CompletionService
|
||||
): Completed = {
|
||||
progress.beforeWork(node)
|
||||
val rawResult = wideConvert(f).left.map {
|
||||
case i: Incomplete => if (config.overwriteNode(i)) i.copy(node = Some(node)) else i
|
||||
case e => Incomplete(Some(node), Incomplete.Error, directCause = Some(e))
|
||||
|
|
@ -306,8 +307,8 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
}
|
||||
}
|
||||
private[this] def rewrap[A](
|
||||
rawResult: Either[Incomplete, Either[F[A], A]]
|
||||
): Either[F[A], Result[A]] =
|
||||
rawResult: Either[Incomplete, Either[TaskId[A], A]]
|
||||
): Either[TaskId[A], Result[A]] =
|
||||
rawResult match {
|
||||
case Left(i) => Right(Result.Inc(i))
|
||||
case Right(Right(v)) => Right(Result.Value(v))
|
||||
|
|
@ -317,108 +318,105 @@ private[sbt] final class Execute[F[_] <: AnyRef](
|
|||
def remove[K, V](map: Map[K, V], k: K): V =
|
||||
map.remove(k).getOrElse(sys.error("Key '" + k + "' not in map :\n" + map))
|
||||
|
||||
def addReverse[A](node: F[A], dependent: F[Any]): Unit =
|
||||
reverse(node.asInstanceOf) ++= Seq(dependent)
|
||||
def addCaller[A](caller: F[A], target: F[A]): Unit =
|
||||
callers.getOrUpdate(target, IDSet.create[F[A]]) += caller
|
||||
def addReverse(node: TaskId[?], dependent: TaskId[?]): Unit =
|
||||
reverse(node) ++= Seq(dependent)
|
||||
def addCaller[A](caller: TaskId[A], target: TaskId[A]): Unit =
|
||||
callers.getOrUpdate(target, IDSet.create) += caller
|
||||
|
||||
def dependencies[A](node: F[A]): Iterable[F[Any]] = dependencies(viewCache(node.asInstanceOf))
|
||||
def dependencies[A](v: Node[F, A]): Iterable[F[Any]] =
|
||||
v.alist.toList[F](v.in).filter(dep => view.inline1(dep).isEmpty)
|
||||
def dependencies(node: TaskId[?]): Iterable[TaskId[?]] = dependencies(viewCache(node))
|
||||
def dependencies(v: Node[?]): Iterable[TaskId[?]] =
|
||||
v.alist.toList(v.in).filter(dep => view.inline1(dep).isEmpty)
|
||||
|
||||
def runBefore[A](node: F[A]): Seq[F[A]] =
|
||||
getSeq[A](triggers.runBefore, node)
|
||||
def triggeredBy[A](node: F[A]): Seq[F[A]] = getSeq(triggers.injectFor, node)
|
||||
def getSeq[A](map: collection.Map[F[Any], Seq[F[Any]]], node: F[A]): Seq[F[A]] =
|
||||
map.getOrElse(node.asInstanceOf, nilSeq[F[Any]]).asInstanceOf
|
||||
def runBefore(node: TaskId[?]): Seq[TaskId[?]] = triggers.runBefore.getOrElse(node, nilSeq)
|
||||
def triggeredBy(node: TaskId[?]): Seq[TaskId[?]] = triggers.injectFor.getOrElse(node, nilSeq)
|
||||
|
||||
// Contracts
|
||||
|
||||
def addedInv[A](node: F[A]): Unit = topologicalSort(node) foreach addedCheck
|
||||
def addedCheck[A](node: F[A]): Unit = {
|
||||
def addedInv(node: TaskId[?]): Unit = topologicalSort(node).foreach(addedCheck)
|
||||
def addedCheck(node: TaskId[?]): Unit = {
|
||||
assert(added(node), "Not added: " + node)
|
||||
assert(viewCache.contains[Any](node.asInstanceOf), "Not in view cache: " + node)
|
||||
dependencyCheck(node.asInstanceOf)
|
||||
assert(viewCache.contains(node), "Not in view cache: " + node)
|
||||
dependencyCheck(node)
|
||||
}
|
||||
def dependencyCheck(node: F[Any]): Unit = {
|
||||
def dependencyCheck(node: TaskId[?]): Unit = {
|
||||
dependencies(node) foreach { dep =>
|
||||
def onOpt[A](o: Option[A])(f: A => Boolean) = o match {
|
||||
case None => false; case Some(x) => f(x)
|
||||
}
|
||||
def checkForward = onOpt(forward.get(node.asInstanceOf)) { _ contains dep.asInstanceOf }
|
||||
def checkReverse = onOpt(reverse.get(dep.asInstanceOf)) { _.exists(_ == node) }
|
||||
assert(done(dep.asInstanceOf) ^ (checkForward && checkReverse))
|
||||
def checkForward = onOpt(forward.get(node))(_.contains(dep))
|
||||
def checkReverse = onOpt(reverse.get(dep))(_.exists(_ == node))
|
||||
assert(done(dep) ^ (checkForward && checkReverse))
|
||||
}
|
||||
}
|
||||
def pendingInv[A](node: F[A]): Unit = {
|
||||
def pendingInv(node: TaskId[?]): Unit = {
|
||||
assert(atState(node, Pending))
|
||||
assert((dependencies(node) ++ runBefore(node)) exists notDone.asInstanceOf)
|
||||
assert((dependencies(node) ++ runBefore(node)).exists(notDone))
|
||||
}
|
||||
def runningInv[A](node: F[A]): Unit = {
|
||||
assert(dependencies(node) forall done.asInstanceOf)
|
||||
assert(!(forward.contains(node.asInstanceOf)))
|
||||
def runningInv(node: TaskId[?]): Unit = {
|
||||
assert(dependencies(node).forall(done))
|
||||
assert(!(forward.contains(node)))
|
||||
}
|
||||
def newPre[A](node: F[A]): Unit = {
|
||||
def newPre(node: TaskId[?]): Unit = {
|
||||
isNew(node)
|
||||
assert(!(reverse.contains(node.asInstanceOf)))
|
||||
assert(!(forward.contains(node.asInstanceOf)))
|
||||
assert(!(callers.contains[Any](node.asInstanceOf)))
|
||||
assert(!(viewCache.contains[Any](node.asInstanceOf)))
|
||||
assert(!(results.contains[Any](node.asInstanceOf)))
|
||||
assert(!(reverse.contains(node)))
|
||||
assert(!(forward.contains(node)))
|
||||
assert(!(callers.contains(node)))
|
||||
assert(!(viewCache.contains(node)))
|
||||
assert(!(results.contains(node)))
|
||||
}
|
||||
|
||||
def topologicalSort[A](node: F[A]): Seq[F[Any]] = {
|
||||
val seen = IDSet.create[F[Any]]
|
||||
def visit(n: F[Any]): List[F[Any]] =
|
||||
(seen process n)(List[F[Any]]()) {
|
||||
node.asInstanceOf :: dependencies(n).foldLeft(List[F[Any]]()) { (ss, dep) =>
|
||||
visit(dep.asInstanceOf) ::: ss
|
||||
}
|
||||
def topologicalSort(node: TaskId[?]): Seq[TaskId[?]] = {
|
||||
val seen = IDSet.create[TaskId[?]]
|
||||
def visit(n: TaskId[?]): List[TaskId[?]] =
|
||||
seen.process(n)(List.empty) {
|
||||
val deps: List[TaskId[?]] =
|
||||
dependencies(n).foldLeft(List.empty)((ss, dep) => visit(dep) ::: ss)
|
||||
node :: deps
|
||||
}
|
||||
|
||||
visit(node.asInstanceOf).reverse
|
||||
visit(node).reverse
|
||||
}
|
||||
|
||||
def readyInv[A](node: F[A]): Unit = {
|
||||
assert(dependencies(node) forall done.asInstanceOf)
|
||||
assert(!(forward.contains(node.asInstanceOf)))
|
||||
def readyInv(node: TaskId[?]): Unit = {
|
||||
assert(dependencies(node).forall(done))
|
||||
assert(!(forward.contains(node)))
|
||||
}
|
||||
|
||||
// cyclic reference checking
|
||||
|
||||
def snapshotCycleCheck(): Unit =
|
||||
callers.toSeq foreach { case (called: F[c], callers) =>
|
||||
for (caller <- callers) cycleCheck(caller.asInstanceOf[F[c]], called)
|
||||
callers.toSeq foreach { case (called, callers) =>
|
||||
for (caller <- callers) cycleCheck(caller, called)
|
||||
}
|
||||
|
||||
def cycleCheck[A](node: F[A], target: F[A]): Unit = {
|
||||
def cycleCheck(node: TaskId[?], target: TaskId[?]): Unit = {
|
||||
if (node eq target) cyclic(node, target, "Cannot call self")
|
||||
val all = IDSet.create[F[A]]
|
||||
def allCallers(n: F[A]): Unit = (all process n)(()) {
|
||||
val all = IDSet.create[TaskId[?]]
|
||||
def allCallers(n: TaskId[?]): Unit = (all process n)(()) {
|
||||
callers.get(n).toList.flatten.foreach(allCallers)
|
||||
}
|
||||
allCallers(node)
|
||||
if (all contains target) cyclic(node, target, "Cyclic reference")
|
||||
}
|
||||
def cyclic[A](caller: F[A], target: F[A], msg: String) =
|
||||
def cyclic(caller: TaskId[?], target: TaskId[?], msg: String) =
|
||||
throw new Incomplete(
|
||||
Some(caller),
|
||||
message = Some(msg),
|
||||
directCause = Some(new CyclicException(caller, target, msg))
|
||||
)
|
||||
final class CyclicException[A](val caller: F[A], val target: F[A], msg: String)
|
||||
final class CyclicException(val caller: TaskId[?], val target: TaskId[?], msg: String)
|
||||
extends Exception(msg)
|
||||
|
||||
// state testing
|
||||
|
||||
def pending[A](d: F[A]) = atState(d, Pending)
|
||||
def running[A](d: F[A]) = atState(d, Running)
|
||||
def calling[A](d: F[A]) = atState(d, Calling)
|
||||
def done[A](d: F[A]) = atState(d, Done)
|
||||
def notDone[A](d: F[A]) = !done(d)
|
||||
private def atState[A](d: F[A], s: State) = state.get(d.asInstanceOf) == Some(s)
|
||||
def isNew[A](d: F[A]) = !added(d)
|
||||
def added[A](d: F[A]) = state.contains(d.asInstanceOf)
|
||||
def pending(d: TaskId[?]) = atState(d, Pending)
|
||||
def running(d: TaskId[?]) = atState(d, Running)
|
||||
def calling(d: TaskId[?]) = atState(d, Calling)
|
||||
def done(d: TaskId[?]) = atState(d, Done)
|
||||
def notDone(d: TaskId[?]) = !done(d)
|
||||
private def atState(d: TaskId[?], s: State) = state.get(d) == Some(s)
|
||||
def isNew(d: TaskId[?]) = !added(d)
|
||||
def added(d: TaskId[?]) = state.contains(d)
|
||||
def complete = state.values.forall(_ == Done)
|
||||
|
||||
def pre(f: => Unit) = if (checkPreAndPostConditions) f
|
||||
|
|
|
|||
|
|
@ -14,7 +14,7 @@ import sbt.internal.util.RMap
|
|||
* except `started` and `finished`, which is called from the executing task's thread. All methods
|
||||
* should return quickly to avoid task execution overhead.
|
||||
*/
|
||||
trait ExecuteProgress[F[_]] {
|
||||
trait ExecuteProgress {
|
||||
def initial(): Unit
|
||||
|
||||
/**
|
||||
|
|
@ -22,20 +22,24 @@ trait ExecuteProgress[F[_]] {
|
|||
* `task` are `allDeps` and the subset of those dependencies that have not completed are
|
||||
* `pendingDeps`.
|
||||
*/
|
||||
def afterRegistered(task: F[Any], allDeps: Iterable[F[Any]], pendingDeps: Iterable[F[Any]]): Unit
|
||||
def afterRegistered(
|
||||
task: TaskId[?],
|
||||
allDeps: Iterable[TaskId[?]],
|
||||
pendingDeps: Iterable[TaskId[?]]
|
||||
): Unit
|
||||
|
||||
/**
|
||||
* Notifies that all of the dependencies of `task` have completed and `task` is therefore ready to
|
||||
* run. The task has not been scheduled on a thread yet.
|
||||
*/
|
||||
def afterReady(task: F[Any]): Unit
|
||||
def afterReady(task: TaskId[?]): Unit
|
||||
|
||||
/**
|
||||
* Notifies that the work for `task` is starting after this call returns. This is called from the
|
||||
* thread the task executes on, unlike most other methods in this callback. It is called
|
||||
* immediately before the task's work starts with minimal intervening executor overhead.
|
||||
*/
|
||||
def beforeWork(task: F[Any]): Unit
|
||||
def beforeWork(task: TaskId[?]): Unit
|
||||
|
||||
/**
|
||||
* Notifies that the work for `task` work has finished. The task may have computed the next task
|
||||
|
|
@ -45,16 +49,16 @@ trait ExecuteProgress[F[_]] {
|
|||
* executes on, unlike most other methods in this callback. It is immediately called after the
|
||||
* task's work is complete with minimal intervening executor overhead.
|
||||
*/
|
||||
def afterWork[A](task: F[A], result: Either[F[A], Result[A]]): Unit
|
||||
def afterWork[A](task: TaskId[A], result: Either[TaskId[A], Result[A]]): Unit
|
||||
|
||||
/**
|
||||
* Notifies that `task` has completed. The task's work is done with a final `result`. Any tasks
|
||||
* called by `task` have completed.
|
||||
*/
|
||||
def afterCompleted[A](task: F[A], result: Result[A]): Unit
|
||||
def afterCompleted[A](task: TaskId[A], result: Result[A]): Unit
|
||||
|
||||
/** All tasks have completed with the final `results` provided. */
|
||||
def afterAllCompleted(results: RMap[F, Result]): Unit
|
||||
def afterAllCompleted(results: RMap[TaskId, Result]): Unit
|
||||
|
||||
/** Notifies that either all tasks have finished or cancelled. */
|
||||
def stop(): Unit
|
||||
|
|
@ -64,46 +68,46 @@ trait ExecuteProgress[F[_]] {
|
|||
* This module is experimental and subject to binary and source incompatible changes at any time.
|
||||
*/
|
||||
object ExecuteProgress {
|
||||
def empty[F[_]]: ExecuteProgress[F] = new ExecuteProgress[F] {
|
||||
def empty: ExecuteProgress = new ExecuteProgress {
|
||||
override def initial(): Unit = ()
|
||||
override def afterRegistered(
|
||||
task: F[Any],
|
||||
allDeps: Iterable[F[Any]],
|
||||
pendingDeps: Iterable[F[Any]]
|
||||
task: TaskId[?],
|
||||
allDeps: Iterable[TaskId[?]],
|
||||
pendingDeps: Iterable[TaskId[?]]
|
||||
): Unit =
|
||||
()
|
||||
override def afterReady(task: F[Any]): Unit = ()
|
||||
override def beforeWork(task: F[Any]): Unit = ()
|
||||
override def afterWork[A](task: F[A], result: Either[F[A], Result[A]]): Unit = ()
|
||||
override def afterCompleted[A](task: F[A], result: Result[A]): Unit = ()
|
||||
override def afterAllCompleted(results: RMap[F, Result]): Unit = ()
|
||||
override def afterReady(task: TaskId[?]): Unit = ()
|
||||
override def beforeWork(task: TaskId[?]): Unit = ()
|
||||
override def afterWork[A](task: TaskId[A], result: Either[TaskId[A], Result[A]]): Unit = ()
|
||||
override def afterCompleted[A](task: TaskId[A], result: Result[A]): Unit = ()
|
||||
override def afterAllCompleted(results: RMap[TaskId, Result]): Unit = ()
|
||||
override def stop(): Unit = ()
|
||||
}
|
||||
|
||||
def aggregate[F[_]](reporters: Seq[ExecuteProgress[F]]) = new ExecuteProgress[F] {
|
||||
def aggregate(reporters: Seq[ExecuteProgress]) = new ExecuteProgress {
|
||||
override def initial(): Unit = {
|
||||
reporters foreach { _.initial() }
|
||||
}
|
||||
override def afterRegistered(
|
||||
task: F[Any],
|
||||
allDeps: Iterable[F[Any]],
|
||||
pendingDeps: Iterable[F[Any]]
|
||||
task: TaskId[?],
|
||||
allDeps: Iterable[TaskId[?]],
|
||||
pendingDeps: Iterable[TaskId[?]]
|
||||
): Unit = {
|
||||
reporters foreach { _.afterRegistered(task, allDeps, pendingDeps) }
|
||||
}
|
||||
override def afterReady(task: F[Any]): Unit = {
|
||||
override def afterReady(task: TaskId[?]): Unit = {
|
||||
reporters foreach { _.afterReady(task) }
|
||||
}
|
||||
override def beforeWork(task: F[Any]): Unit = {
|
||||
override def beforeWork(task: TaskId[?]): Unit = {
|
||||
reporters foreach { _.beforeWork(task) }
|
||||
}
|
||||
override def afterWork[A](task: F[A], result: Either[F[A], Result[A]]): Unit = {
|
||||
override def afterWork[A](task: TaskId[A], result: Either[TaskId[A], Result[A]]): Unit = {
|
||||
reporters foreach { _.afterWork(task, result) }
|
||||
}
|
||||
override def afterCompleted[A](task: F[A], result: Result[A]): Unit = {
|
||||
override def afterCompleted[A](task: TaskId[A], result: Result[A]): Unit = {
|
||||
reporters foreach { _.afterCompleted(task, result) }
|
||||
}
|
||||
override def afterAllCompleted(results: RMap[F, Result]): Unit = {
|
||||
override def afterAllCompleted(results: RMap[TaskId, Result]): Unit = {
|
||||
reporters foreach { _.afterAllCompleted(results) }
|
||||
}
|
||||
override def stop(): Unit = {
|
||||
|
|
|
|||
|
|
@ -11,6 +11,7 @@ import scala.collection.mutable.ListBuffer
|
|||
|
||||
import sbt.internal.util.IDSet
|
||||
import Incomplete.{ Error, Value => IValue }
|
||||
import scala.jdk.CollectionConverters.*
|
||||
|
||||
/**
|
||||
* Describes why a task did not complete.
|
||||
|
|
@ -45,7 +46,6 @@ object Incomplete extends Enumeration {
|
|||
def transformTD(i: Incomplete)(f: Incomplete => Incomplete): Incomplete = transform(i, true)(f)
|
||||
def transformBU(i: Incomplete)(f: Incomplete => Incomplete): Incomplete = transform(i, false)(f)
|
||||
def transform(i: Incomplete, topDown: Boolean)(f: Incomplete => Incomplete): Incomplete = {
|
||||
import collection.JavaConverters._
|
||||
val visited: collection.mutable.Map[Incomplete, Incomplete] =
|
||||
(new java.util.IdentityHashMap[Incomplete, Incomplete]).asScala
|
||||
def visit(inc: Incomplete): Incomplete =
|
||||
|
|
|
|||
|
|
@ -17,11 +17,11 @@ import sbt.internal.util.AList
|
|||
* @tparam A
|
||||
* the type computed by this node
|
||||
*/
|
||||
private[sbt] trait Node[Effect[_], A]:
|
||||
private[sbt] trait Node[A]:
|
||||
type K[L[x]]
|
||||
def in: K[Effect]
|
||||
def in: K[TaskId]
|
||||
def alist: AList[K]
|
||||
|
||||
/** Computes the result of this task given the results from the inputs. */
|
||||
def work(inputs: K[Result]): Either[Effect[A], A]
|
||||
def work(inputs: K[Result]): Either[TaskId[A], A]
|
||||
end Node
|
||||
|
|
|
|||
|
|
@ -0,0 +1,6 @@
|
|||
package sbt
|
||||
|
||||
import sbt.internal.util.AttributeMap
|
||||
|
||||
trait TaskId[A]:
|
||||
def tags: ConcurrentRestrictions.TagMap
|
||||
Loading…
Reference in New Issue