BeProperSubsetOf and BeProperSupersetOf

[Meir017]

Background and motivation

Asserting proper subset and proper superset on collections

proper subset - a collection is included in a superset collection but not equivalent to the superset collection
proper superset - a collection is contains a subset collection but not equivalent to the subset collection

API Proposal

GenericCollectionAssertions<TCollection, T, TAssertions>
{
    public AndConstraint<TAssertions> BeProperSubsetOf(IEnumerable<T> expectedSuperset, string because = "", params object[] becauseArgs);
    public AndConstraint<TAssertions> BeProperSupersetOf(IEnumerable<T> expectedSubset, string because = "", params object[] becauseArgs);
}

API Usage

ISet<int> subset = new HashSet<int>{ 1, 2, 3 };
ISet<int> superset = new HashSet<int>{ 2, 3, 4 };

subset.Should().BeProperSubsetOf(new[] { 1, 2, 3 }); // will throw
subset.Should().BeProperSubsetOf(new[] { 1, 2, 3, 4 }); // will pass

superset.Should().BeProperSupersetOf(new[] { 2, 3, 4 }); // will throw
superset.Should().BeProperSupersetOf(new[] { 2, 3, 4, 5 }); // will pass

Alternative Designs

subset.Should().BeSubsetOf(new[] { 1, 2, 3 }).And.NotBeEquivalentTo(new[] { 1, 2, 3 }); // for ProperSubset
new[] { 2, 3, 4 }.Should().BeSubsetOf(superset).And.NotBeEquivalentTo(superset); // for ProperSuperset

Risks

No response

Are you willing to help with a proof-of-concept (as PR in that or a separate repo) first and as pull-request later on?

Yes

[dennisdoomen]

I like it, although it is a bit niche. What about you @jnyrup ?

[jnyrup]

I like it, although it is a bit niche. What about you @jnyrup ?

I'm also quite niche.
I like set theory, so I've been waiting for years for someone else to also propose this.

Should we also add BeSupersetOf for symmetry even though that would just be forwarding to the existing Contain?

fluentassertions/Src/FluentAssertions/Collections/GenericCollectionAssertions.cs

Line 774 in 6f1035a

public AndConstraint<TAssertions> Contain(IEnumerable<T> expected, string because = "", params object[] becauseArgs)

[Meir017]

@jnyrup maybe adding BeSupersetOf could be added separately (different PR)?

[dennisdoomen]

Assuming you include BeSupersetOf, I'm fine with this proposal. If @jnyrup is too, it's approved.

[jnyrup]

I'm good with it.

[dennisdoomen]

Can we assign this one to you @Meir017 ?

[Meir017]

@dennisdoomen yes please