Introduce parReplicateA #4007
Introduce parReplicateA #4007
Conversation
| /** | ||
| * Like [[Applicative.replicateA]], but uses the apply instance | ||
| * corresponding to the Parallel instance instead. | ||
| */ | ||
| def parReplicateA[A](n: Int, ma: M[A]): M[List[A]] = | ||
| if (n <= 0) monad.pure(List.empty[A]) | ||
| else Parallel.parSequence(List.fill(n)(ma))(UnorderedFoldable.catsTraverseForList, this) |
There was a problem hiding this comment.
To me it seems that trait Parallel is designed to just represent a relationship between a Monad and Applicative with parallel composition. But most of complimentary methods are placed into object Parallel (including parAp per se). I'm not absolutely confident, but shouldn't parReplicateA be also placed in the object rather than the trait?
| if (n <= 0) P.monad.pure(List.empty[A]) | ||
| else Parallel.parSequence(List.fill(n)(ma)) |
There was a problem hiding this comment.
I guess these two lines can be replaced with just P.parReplicateA(...) if you decide to leave Parallel#parReplicateA in the trait Parallel or with Parallel.parReplicateA(...) if you move it into the object. Just to avoid code duplication.
| @@ -184,6 +184,7 @@ object SyntaxSuite { | |||
| val mb4: M[B] = ma.parProductR(mb) | |||
| val mab2: M[(A, B)] = ma.parProduct(mb) | |||
| val mb5: M[B] = mab.parAp(ma) | |||
| val mla: M[List[A]] = ma.parReplicateA(3) | |||
There was a problem hiding this comment.
mock[Int] instead of just 3?
|
Hello @satorg I agree with you comments and so I've pushed corresponding changes. |
|
Hello @satorg I added some tests as you've requested. |
| * Like `Applicative[F].replicateA`, but uses the apply instance | ||
| * corresponding to the Parallel instance instead. | ||
| */ | ||
| def parReplicateA[M[_], A](n: Int, ma: M[A])(implicit P: Parallel[M]): M[List[A]] = |
There was a problem hiding this comment.
why not P.applicative.replicateA(n, ma)?
Since in principle, that implementation could contain some optimization/specialization of some kind.
I think that is a better than having a second implementation here.
There was a problem hiding this comment.
It has to be a bit more complicated than that.
P.sequential(P.applicative.replicateA(n, P.parallel(ma)))
But yes, I agree, it's better to reuse it.
This PR introduces
parReplicateAas a parallel version ofApplicative.replicateA.