Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Optimize foreachExec and underlying private foreach* methods #9804

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Apr 21, 2025
Merged

Optimize foreachExec and underlying private foreach* methods #9804

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7a7fdd4
Optimize `foreachExec`
kyri-petrou Apr 17, 2025
6b2aadb
Remove redundant `min` op
kyri-petrou Apr 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
84 changes: 46 additions & 38 deletions core/shared/src/main/scala/zio/ZIO.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3547,23 +3547,25 @@ object ZIO extends ZIOCompanionPlatformSpecific with ZIOCompanionVersionSpecific
* Applies the function `f` to each element of the `Collection[A]` and returns
* the result in a new `Collection[B]` using the specified execution strategy.
*/
final def foreachExec[R, E, A, B, Collection[+Element] <: Iterable[Element]](as: Collection[A])(
def foreachExec[R, E, A, B, Collection[+Element] <: Iterable[Element]](
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I assume final was unnecessary in the object?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yeap 👍

as: Collection[A]
)(
exec: => ExecutionStrategy
)(
f: A => ZIO[R, E, B]
)(implicit bf: BuildFrom[Collection[A], B, Collection[B]], trace: Trace): ZIO[R, E, Collection[B]] =
if (as.isEmpty) ZIO.succeed(bf.fromSpecific(as)(Nil))
else
ZIO.suspendSucceed {
exec match {
case ExecutionStrategy.Parallel =>
ZIO.withParallelismUnboundedMask(restore => ZIO.foreachPar(as)(a => restore(f(a))))
case ExecutionStrategy.ParallelN(n) =>
ZIO.withParallelismMask(n)(restore => ZIO.foreachPar(as)(a => restore(f(a))))
case ExecutionStrategy.Sequential =>
ZIO.foreach(as)(f)
as.size match {
case 0 => ZIO.succeed(bf.fromSpecific(as)(Nil))
case 1 => ZIO.suspendSucceed(f(as.head)).map(b => bf.fromSpecific(as)(as.map(_ => b)))
case size =>
ZIO.suspendSucceed {
exec match {
case ExecutionStrategy.Parallel => ZIO.foreachParUnbounded(as, size)(f)
case ExecutionStrategy.ParallelN(n) => ZIO.foreachPar(n)(as, size)(f)
case ExecutionStrategy.Sequential => ZIO.foreach(as)(f)
}
}
}
}

/**
* Applies the function `f` to each element of the `Collection[A]` in
Expand Down Expand Up @@ -6285,44 +6287,50 @@ object ZIO extends ZIOCompanionPlatformSpecific with ZIOCompanionVersionSpecific
): ZIO[R, E, Unit] =
foreachParUnboundedDiscard(as, size)(ZIO.identityFn)

private def foreachPar[R, E, A, B, Collection[+Element] <: Iterable[Element]](n: => Int)(
private def foreachPar[R, E, A, B, Collection[+Element] <: Iterable[Element]](
parallelism: Int
)(
as: Collection[A],
size: Int
)(
fn: A => ZIO[R, E, B]
)(implicit bf: BuildFrom[Collection[A], B, Collection[B]], trace: Trace): ZIO[R, E, Collection[B]] =
ZIO.suspendSucceed {
val array = Array.ofDim[AnyRef](size)
val zioFunction: ((A, Int)) => ZIO[R, E, Any] = { case (a, i) =>
fn(a).flatMap { b => array(i) = b.asInstanceOf[AnyRef]; Exit.unit }
if (parallelism <= 1)
foreach(as)(fn)
else if (parallelism >= size)
foreachParUnbounded(as, size)(fn)
else
ZIO.suspendSucceed {
val array = Array.ofDim[AnyRef](size)
val zioFunction: ((A, Int)) => ZIO[R, E, Any] = { case (a, i) =>
fn(a).flatMap { b => array(i) = b.asInstanceOf[AnyRef]; Exit.unit }
}
foreachParDiscard(parallelism)(as.zipWithIndex, size)(zioFunction)
.as(bf.fromSpecific(as)(array.asInstanceOf[Array[B]]))
}
foreachParDiscard(n)(as.zipWithIndex, size)(zioFunction)
.as(bf.fromSpecific(as)(array.asInstanceOf[Array[B]]))
}

private def foreachParDiscard[R, E, A](
n: Int
parallelism: Int
)(as: Iterable[A], size: Int)(f: A => ZIO[R, E, Any])(implicit trace: Trace): ZIO[R, E, Unit] =
size match {
case 0 => Exit.unit
case 1 => f(as.head).unit
case size =>
ZIO.suspendSucceed {
import scala.jdk.CollectionConverters._
if (parallelism <= 1)
foreachDiscard(as)(f)
else if (parallelism >= size)
foreachParUnboundedDiscard(as, size)(f)
Comment on lines +6317 to +6318
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

does this need to be a distinct branch? Wouldn't val nWorkers = parallelism.min(size) take care of this?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is more optimal because in this branch we don't have to go through polling via the queue. But you're right this makes val nWorkers = parallelism.min(size) obsolete, we can simply use parallelism with this new implementation

else
ZIO.suspendSucceed {
import scala.jdk.CollectionConverters._

val queue = new ConcurrentLinkedQueue[A](as.asJavaCollection)
val queue = new ConcurrentLinkedQueue[A](as.asJavaCollection)

lazy val worker: ZIO[R, E, Unit] =
ZIO.suspendSucceed(queue.poll() match {
case null => Exit.unit
case a => f(a) *> worker
})
lazy val worker: ZIO[R, E, Unit] =
ZIO.suspendSucceed(queue.poll() match {
case null => Exit.unit
case a => f(a) *> worker
})

val nWorkers = n.min(size)
val workers = ZIO.replicate(nWorkers)(worker)
ZIO.collectAllParUnboundedDiscard(workers, nWorkers)
}
}
val workers = ZIO.replicate(parallelism)(worker)
ZIO.collectAllParUnboundedDiscard(workers, parallelism)
}

private def foreachParUnbounded[R, E, A, B, Collection[+Element] <: Iterable[Element]](
as: Collection[A],
Expand Down
Loading