Improvements to Tuples #19172
Improvements to Tuples #19172
Conversation
odersky
force-pushed
the
augment-tuples
branch
from
bc8fa14 to
eff2d54
Compare
|
|
||
| object Tuple { | ||
| /** A tuple with the elements of this tuple in reversed order added in front of `acc` */ | ||
| inline def reverseOnto[This >: this.type <: Tuple, Acc <: Tuple](acc: Acc): ReverseOnto[This, Acc] = |
There was a problem hiding this comment.
| inline def reverseOnto[This >: this.type <: Tuple, Acc <: Tuple](acc: Acc): ReverseOnto[This, Acc] = | |
| inline def reverseOnto[This >: this.type <: Tuple, That <: Tuple](taht: That): ReverseOnto[This, That] = |
There was a problem hiding this comment.
I think I prefer acc/Acc here.
| @experimental | ||
| object Helpers: | ||
| /** A tuple with the elements of tuple `X` in reversed order added in front of `Acc` */ | ||
| type ReverseOnto[X <: Tuple, Acc <: Tuple] <: Tuple = X match |
There was a problem hiding this comment.
| type ReverseOnto[X <: Tuple, Acc <: Tuple] <: Tuple = X match | |
| type ReverseOnto[X <: Tuple, That <: Tuple] <: Tuple = X match |
| fromArray(xs, xs.length) | ||
|
|
||
| /** Convert the first `n` elements of an array into a tuple of unknown arity and types */ | ||
| def fromArray[T](xs: Array[T], n: Int): Tuple = { |
There was a problem hiding this comment.
We could compute the type (Any, Any, ..., Any) if we have a known n.
There was a problem hiding this comment.
Yes, we could. Needs to be discussed, and can be in a separate PR. I don't think the functionality is so important in fromArray, but having a constructor for "type level list of known size" might be useful elsewhere.
| @@ -82,82 +81,126 @@ sealed trait Tuple extends Product { | |||
| /** Given a tuple `(a1, ..., am)`, returns the reversed tuple `(am, ..., a1)` | |||
| * consisting all its elements. | |||
| */ | |||
| @experimental | |||
There was a problem hiding this comment.
We usually do not mix the stabilization of API with the addition of new ones.
| * for which the given type level predicate `P` reduces to the literal | ||
| * constant `true`. | ||
| */ | ||
| inline def filter[This >: this.type <: Tuple, P[_] <: Boolean]: Filter[This, P] = |
There was a problem hiding this comment.
| inline def filter[This >: this.type <: Tuple, P[_] <: Boolean]: Filter[This, P] = | |
| @exerimental | |
| inline def filter[This >: this.type <: Tuple, P[_] <: Boolean]: Filter[This, P] = |
|
I don't really have the bandwidth to do further changes here. Should someone else take this over? |
New methods: filter, indicesWhere, reverseOnto
odersky
force-pushed
the
augment-tuples
branch
from
eff2d54 to
254b72a
Compare
|
I broke it down to several commits. |
|
I will take this over. I will split it onto different PRs for each addition and add tests for each one. The first part is in #19183. |
|
Just be aware that the named tuples PR has the same series of commits. I would not like to have to get back to this and sort out again a huge amount of merge conflicts. I already spent an afternoon doing that yesterday. So let's get these things in as they are. |
|
This is superseded by the work we did on NamedTuples. @EugeneFlesselle but it might be useful as a starting point for your improvements. |
Indeed, there seems to be some overlap. I'll rebase everything we've got and hopefully we'll be able to improve theses tuples soo enough 🤞 |
Improve organization and doc comments