Add unorderedReduceOption to UnorderedFoldable

[jorgeadriano]

Add unorderedReduceOption to UnorderedFoldable

Motivation

Foldable has ordered reduction methods such as reduceLeftOption and reduceRightOption,
but UnorderedFoldable does not currently have an unordered equivalent for reducing a
possibly-empty structure.

This PR adds unorderedReduceOption, which reduces an unordered structure to Option[A],
returning None for empty structures and Some(combined) for non-empty ones.

The constraint is CommutativeSemigroup[A], rather than Semigroup[A], because
UnorderedFoldable provides no guarantee on element order.

This is related to the discussion in #2715 and extracts the focused unorderedReduceOption
part of #3309. Unlike #3309, this PR intentionally leaves minimumOption, maximumOption,
minimumByOption, and maximumByOption out of scope, since those raise separate API and
binary-compatibility considerations, as also discussed around #3132.

API

def unorderedReduceOption[A](fa: F[A])(implicit A: CommutativeSemigroup[A]): Option[A]

Also available as syntax via cats.syntax.all._:

import cats.syntax.all._

Set(1, 2, 3).unorderedReduceOption   // Some(6)
Set.empty[Int].unorderedReduceOption // None

Implementation

The implementation is defined in terms of unorderedFoldMap, by lifting each A into
Option[A] and folding with the existing CommutativeMonoid[Option[A]] instance derived
from CommutativeSemigroup[A].

Law

The PR adds a law checking consistency with unorderedFold:

def unorderedReduceOptionConsistentWithUnorderedFold[A: CommutativeMonoid](fa: F[A]): IsEq[Option[A]] =
  F.unorderedReduceOption(fa) <-> (if (F.isEmpty(fa)) None else Some(F.unorderedFold(fa)))

The law is wired into the existing UnorderedFoldableTests.unorderedFoldable rule set.
Eq[Option[A]] is resolved internally from the existing Eq[A] instance via
cats.instances.option, so the public signature of unorderedFoldable is unchanged.

Included changes

• UnorderedFoldable#unorderedReduceOption
• syntax support via cats.syntax.all._
• a law checking consistency with unorderedFold
• discipline test integration
• a focused test in UnorderedFoldableSuite

[Copilot]

Pull request overview

Adds unorderedReduceOption to UnorderedFoldable to support reducing possibly-empty unordered structures, with accompanying syntax and law/test coverage to ensure consistency with unorderedFold.

Changes:

• Add UnorderedFoldable#unorderedReduceOption implemented via unorderedFoldMap over Option[A].
• Add syntax support (fa.unorderedReduceOption) under cats.syntax.all._.
• Add a new law + discipline wiring, plus a focused suite test.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
core/src/main/scala/cats/UnorderedFoldable.scala	Adds the new unorderedReduceOption API on the typeclass.
core/src/main/scala/cats/syntax/unorderedFoldable.scala	Exposes unorderedReduceOption as syntax on F[A].
laws/src/main/scala/cats/laws/UnorderedFoldableLaws.scala	Adds a law relating unorderedReduceOption to unorderedFold/isEmpty.
laws/src/main/scala/cats/laws/discipline/UnorderedFoldableTests.scala	Wires the new law into the existing unorderedFoldable RuleSet.
tests/shared/src/test/scala/cats/tests/UnorderedFoldableSuite.scala	Adds a direct property test for unorderedReduceOption.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[satorg]

Thank you for the PR! The change makes sense to me. A few rather minor comments from my side are inline below.

[satorg]

  def unorderedReduceOption[A: CommutativeSemigroup](fa: F[A]): Option[A] = {
    val reducer = CommutativeMonoid[Option[A]]
    unorderedFoldMap(fa)(a => Some(a): Option[A])(using reducer)
  }

1. implicit A only makes sense if the code needs to refer to A directly. It is not the case here, so I'd suggest using the context bounds syntax, as it seems to be more concise and readable.
2. Referring to catsKernelStdCommutativeMonoidForOption directly is rather brittle and doesn't seem necessary. I'd suggest to summon the required instance and pass it to unorderedFoldMap.

[satorg]

Is this test somehow different from the unorderedReduceOptionConsistentWithUnorderedFold check in UnorderedFoldableLaws? If not, then we can probably drop this one in order to avoid bloating tests with identical checks.