Scala Kleisli 在 IntelliJ 中引发错误 [英] Scala Kleisli throws an error in IntelliJ
问题描述
尝试在 Scala 中为一个虚构的 Partial 类型实现 Kleisli 类别(阅读 Bartosz Milewski 的程序员的类别理论",这是第 4 章的练习)
trying to implement Kleisli category for a made-up Partial type in Scala (reading Bartosz Milewski's "category theory for programmers", that's exersize for chapter 4)
object Kleisli {
type Partial[A, B] = A => Option[B]
implicit class KleisliOps[A, B](f1: Partial[A, B]) {
def >=>[C](f2: Partial[B, C]): Partial[A, C] =
(x: A) =>
for {
y <- f1(x)
z <- f2(y)
} yield z
def identity(f: Partial[A, B]): Partial[A, B] = x => f(x)
}
val safeRecip: Partial[Double, Double] = {
case 0d => None
case x => Some(1d / x)
}
val safeRoot: Partial[Double, Double] = {
case x if x < 0 => None
case x => Some(Math.sqrt(x))
}
val safeRootRecip: Partial[Double, Double] = safeRoot.>=>(safeRecip)
safeRootRecip(1d)
safeRootRecip(10d)
safeRootRecip(0d)
}
IDE (IntelliJ) 没有显示错误,但是当我运行这个片段时,我得到:
IDE (IntelliJ) shows no errors, but when I run this snippet, I get:
Error:(27, 57) value >=> is not a member of $line5.$read.$iw.$iw.Kleisli.Partial[Double,Double]
val safeRootRecip: Partial[Double, Double] = safeRoot.>=>(safeRecip)
在隐式类之外定义 >=>
工作正常.可能是什么原因?
Defining >=>
outside of implicit class works fine. What could be the reason?
推荐答案
@sinanspd 是对的.在 Dotty 中,代码似乎可以编译:https://scastie.scala-lang.org/n17APWgMQkWqy93ct2cghw
@sinanspd was right. In Dotty the code seems to compile: https://scastie.scala-lang.org/n17APWgMQkWqy93ct2cghw
手动解决
val safeRootRecip: Partial[Double, Double] = KleisliOps(safeRoot).>=>(safeRecip)
编译但编译器没有发现这个转换本身
compiles but compiler doesn't find this conversion itself
Information: KleisliOps{<null>} is not a valid implicit value
for App.safeRoot.type => ?{def >=> : ?} because:
type mismatch;
found : App.safeRoot.type (with underlying type App.Partial[Double,Double])
required: App.Partial[A,Double]
(which expands to) A => Option[Double]
val safeRootRecip: Partial[Double, Double] = safeRoot.>=>(safeRecip)
似乎没有推断类型参数A
.
It seems type parameter A
is not inferred.
(顺便说一下,这里 Martin Odersky 解释了为什么语言中存在隐式转换会使类型推断变得更糟:https://contributors.scala-lang.org/t/can-we-wean-scala-off-implicit-conversions/4388)
(By the way, here Martin Odersky explains why presence of implicit conversions in language makes type inference worse: https://contributors.scala-lang.org/t/can-we-wean-scala-off-implicit-conversions/4388)
尝试使 Partial
对于 B
是协变的,并且(尤其是)对于 A
是逆变的(类似于 A => Option[B]
对于 B
是协变的,对于 A
是逆变的)
Try to make Partial
covariant with respect to B
and (especially) contravariant with respect to A
(similarly to A => Option[B]
being covariant with respect to B
and contravariant with respect to A
)
type Partial[-A, +B] = A => Option[B]
然后代码似乎可以编译了.
Then the code seems to compile.
另一种解决方法是用类型类 (MyTransform
) 和隐式转换替换隐式转换 (X => Y
, KleisliOps
)(myConversion
) 根据此类型类定义(有时 这有助于隐式转换)
Another workaround is to replace implicit conversions (X => Y
, KleisliOps
) with a type class (MyTransform
) and implicit conversion (myConversion
) defined in terms of this type class (sometimes this helps with implicit conversions)
trait MyTransform[X, Y] {
def transform(x: X): Y
}
implicit def myConversion[X, Y](x: X)(implicit mt: MyTransform[X, Y]): Y =
mt.transform(x)
type Partial[A, B] = A => Option[B]
implicit def partialToKleisliOps[A, B]: MyTransform[Partial[A, B], KleisliOps[A, B]] =
f1 => new KleisliOps(f1)
class KleisliOps[A, B](f1: Partial[A, B]) {
def >=>[C](f2: Partial[B, C]): Partial[A, C] =
(x: A) =>
for {
y <- f1(x)
z <- f2(y)
} yield z
def identity(f: Partial[A, B]): Partial[A, B] = x => f(x)
}
这篇关于Scala Kleisli 在 IntelliJ 中引发错误的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!