milessabin/shapeless

shapeless: Generic programming for Scala | This room deprecated in favour of #shapeless on the Typelevel Discord: https://discord.gg/bSQBZA3Ced

Georgi Krastev

@joroKr21

@voonchav_gitlab @mikla I think this is most probably the bug you were experiencing: milessabin/shapeless#1152 - I am very tempted to just cut another release but if you have a way to test if it works for you or for me to reproduce I would hold off.

Found this bug in the update PR for pureconfig: pureconfig/pureconfig#1037

vonchav

@voonchav_gitlab

@joroKr21 Is there a snapshot build that I can download? Alternatively, I can provide the code. Probably need to reproduce the problem in Ammonite. I wonder if I can reproduce it in scastie :)

Georgi Krastev

@joroKr21

Hmm no, looks like we publish snapshots only from main

Let me fix that

vonchav

@voonchav_gitlab

I keep getting this error from Scastie, no matter what version of cats I specify:

sbt.librarymanagement.ResolveException: Error downloading org.typelevel:cats_2.13:2.4.2
  Not found

Georgi Krastev

@joroKr21

I think you need cats-core

vonchav

@voonchav_gitlab

yea, probably. I end up importing kittens instead... That's good. I just need to import 2.2.2 and 2.3.0 to see if I can reproduce the problem with 2.3.0. Almost done.

vonchav

@voonchav_gitlab

Alright, I can reproduce the problem in Scastie: https://scastie.scala-lang.org/vELbT0QtQramsGbHMAsW0w
If you go to Build Settings and change kittens to 2.2.2, it will compile and the values will be asserted.

Georgi Krastev

@joroKr21

Awesome - in the meantime there is a snapshot 2.3.5+6-1b1fb4f0-SNAPSHOT

vonchav

@voonchav_gitlab

Cool. Going to try it. Thanks.

Georgi Krastev

@joroKr21

That looks like a different problem though. cats also gives you map on tuple - mapping on the last element!

I have no clue how it worked before but now they are clashing

vonchav

@voonchav_gitlab

You're right. cats has map on tuple too.... I didn't think about clashing until you pointed it out.

Maybe I can rewrite my stuff. I wrote that a while ago. Perhaps I can find a simpler way and avoid the clashing.... At the same time, I too wonder how it wasn't a problem before :)

Georgi Krastev

@joroKr21

I think somehow the shapeless map was considered more specific before

Because it had P <: Product

Now it's P: IsTuple

vonchav

@voonchav_gitlab

Right. I think I can use FunctionK for the first transformation.

Georgi Krastev

@joroKr21

We could make it P <: Product: IsTuple

I will sleep on it I'm off now

vonchav

@voonchav_gitlab

That looks fine to me but I don't have the knowledge you have. :)

Thanks a lot, Georgi. Have a good night.

Georgi Krastev

@joroKr21

You too :wave:

Georgi Krastev

@joroKr21

We could make it P <: Product: IsTuple

That does't work because it makes it ambiguous with ProductOps - so I'm afraid we can't resolve this 3-way clash (TupleOps <-> ProductOps <-> FunctorOps from cats)

Georgi Krastev

@joroKr21

I think the only other option would be to add an alias for map

Georgi Krastev

@joroKr21

But there are many other methods like flatMap and foldLeft that would need aliases

vonchav

@voonchav_gitlab

@joroKr21 I managed to get my code to work with kittens 2.3.0 by having the "right" implicit in scope only. Not sure if this is just a hack :)
I updated the scastie: https://scastie.scala-lang.org/H8zcJSIkT0mMvK6kSoutkw (sorry, let me know if you need the original version)

Georgi Krastev

@joroKr21

Yeah that's fine it's not a hack

Totally reasonable to do this when you have clashing syntax

vonchav

@voonchav_gitlab

Cool, I think I'm good. I'm in the process of upgrading shapeless and kittens :) It's good to know the root cause. Thank you very much for your help!

Miles Sabin

@milessabin

FYI, there should be a stable release of shapeless 3 (shapeless-deriving module) for Scala 3 to coincide with ScalaCon.

nafg

@nafg

When is that?

Clint Combs

@ClintCombs

What is K0 here? I get "Not found: K0" when attempting something like this in my own code with Scala 3.0.0-RC3: https://github.com/milessabin/shapeless/blob/v3.0.0-M4/modules/deriving/src/test/scala/shapeless3/deriving/type-classes.scala#L48

Miles Sabin

@milessabin

(Discussion is happening in the Typelevel discord)

Miles Sabin

@milessabin

Just published #shapeless 3.0.0 for #Scala 3.0.0: https://github.com/milessabin/shapeless/tree/shapeless-3

(Most discussion will be on the Typelevel discord)

Miles Sabin

@milessabin

shapeless 3 now has it's own repo in the Typelevel GitHub organisation: https://github.com/typelevel/shapeless-3 :tada:

Miles Sabin

@milessabin

Just published #shapeless 3.0.1 for #Scala 3.0.0 and #Scalajs 1.5.

Alex Henning Johannessen

@ahjohannessen

@milessabin Awesome :) How would I go about defining a Typeable for cats.NonEmptyMap. In shapeless 2 I have something like:

  // Adapted from `shapeless.Typeable.genMapTypeable`.
  implicit def typeableForNonEmptyMap[K, V](
    implicit cM: ClassTag[NonEmptyMap[_, _]],
    tK: Typeable[K],
    tV: Typeable[V]
  ): Typeable[NonEmptyMap[K, V]] = new Typeable[NonEmptyMap[K, V]] {

    def cast(t: Any): Option[NonEmptyMap[K, V]] =
      if (t == null) None
      else if (cM.runtimeClass isAssignableFrom t.getClass) {
        val m = t.asInstanceOf[NonEmptyMap[Any, Any]]
        if (m.toSortedMap.forall(_.cast[(K, V)].isDefined)) Some(t.asInstanceOf[NonEmptyMap[K, V]]) else None
      } else None

    def describe = s"${cM.runtimeClass.getSimpleName}[${tK.describe}, ${tV.describe}]"
  }

}

Alex Henning Johannessen

@ahjohannessen

I get this:

[error] -- Error: /Users/ahjohannessen/Development/Work/subsidy-support/marshal/src/test/scala-3/support/marshal/typeables.scala:18:51 
[error] 18 |      val cm     = m.cast[NonEmptyMap[Int, String]]
[error]    |                                                   ^
[error]    |No Typeable for parametrized type cats.data.NonEmptyMapImpl.Type[scala.Int, scala.Predef.String]
[error]    | This location contains code that was inlined from typeables.scala:18
[error]    | This location contains code that was inlined from typeable.scala:244

Victor Viale

@Koroeskohr

hello friends ! I was wondering if there was a way to add a LUBConstraint to a ProductArgs output? I've been playing with it for a bit but I'm hitting a wall with the Dynamic method signature that's not being respected

+- what I'm trying to achieve

object foo extends ProductArgs {
  def applyProduct[L <: HList](l: L)(implicit LUBC: LUBConstraint[L, Bar]): Job[L] = Job(l)
}

Miles Sabin

@milessabin

Hi folks ... your occasional reminder that shapeless discussion has moved to the Typelevel Discord: https://discord.gg/bSQBZA3Ced

Victor Viale

@Koroeskohr

oh... thanks

Where communities thrive

People

Repo info

Activity