Preparation for Scala 3 migration #201
Conversation
| // for { p <- this.copy(boundT = boundT + NatKind.IDWrapper(x)).`type`(e)} | ||
| // yield (p._1, NatToDataLambda(x, e)) | ||
| val pp: Pure[((OrderedSet[Identifier], OrderedSet[Kind.Identifier]), DataType)] = | ||
| this.copy(boundT = boundT + NatKind.IDWrapper(x)).`type`(e) | ||
| pp.map(p => (p._1, NatToDataLambda(x, e))) |
There was a problem hiding this comment.
is the for syntax not supported anymore?
There was a problem hiding this comment.
No, of course it is, but here the system got confused somehow (and I with it ...).
I think that maybe the type inference has changed, and therefore the type annotation helps here, but I am not sure.
The explicit version with the map should be the correct one - from my point of view at least.
| @@ -414,7 +414,7 @@ object NamedRewriteDSL { | |||
|
|
|||
| import scala.language.implicitConversions | |||
|
|
|||
| val `_`: rct.TypePlaceholder.type = rct.TypePlaceholder | |||
| val `__`: rct.TypePlaceholder.type = rct.TypePlaceholder | |||
There was a problem hiding this comment.
maybe we can also just remove that syntax at that point and always use names ..
There was a problem hiding this comment.
That is completely up to you, I just renamed it in a way that Scala 3 does not compain about ...
| object DepLambda { | ||
| def apply[T, I](kind: Kind[T, I], x: I): Object { | ||
| def apply[U <: PhraseType](body: Phrase[U]): DepLambda[T, I, U] | ||
| } = new { | ||
| def apply[T, I](kind: Kind[T, I], x: I): SyntaxHelper[T, I] = SyntaxHelper(kind, x) | ||
|
|
||
| case class SyntaxHelper[T, I](kind: Kind[T, I], x: I) { | ||
| def apply[U <: PhraseType](body: Phrase[U]): DepLambda[T, I, U] = DepLambda(kind, x, body) | ||
| } | ||
| } |
There was a problem hiding this comment.
Can we use a case class DepLambda to avoid the helper?
There was a problem hiding this comment.
Of course, there is a class DepLambda that represents a dependent lambda phrase.
We usually keep those classes free of any syntactic sugar (which I think is a good idea) and even for this here it is not clear if this syntax should be defined here or elsewhere in a DSL package.
Anyway, I think the purpose of this is only to infer the type U and not have to annotate this, while still allowing the annotation of T and I.
We might want to look into this after having migrated to Scala 3 again, I supose ...
This PR introduces changes that will help to ease the migration to Scala 3, but the code in this PR remains valid Scala 2 code.
In detail the changes are:
()is replaced withUnit(as Scala 3 seems to no longer accept()as a type)T#U) in theFunctionHelperclass is avoided as these have been dropped in Scala 3._1and._2notations on Rise and DPIA expressions do not work in Scala 3 (I don't know why) I have replaced them with`1`and`2``_`notation is also not accepted by Scala 3, I have replaced it with`__`unapplymethod generated for case classes.