TUnit support?

[siewers]

I started experimenting with TUnit and realized there isn't great combinatorial support (yet).
Would it make sense to expand Xunit.Combinatorial to support TUnit as well?
I haven't made much discovery in this yet, but Xunit.Combinatorial could become a base library for generating combinatorial data for any testing framework.

TUnit itself works in many ways like xUnit, so I don't believe supporting it would be a massive undertaking.

I suggest splitting Xunit.Combinatorial into the following packages:

• Combinatorial.Core
• Xunit.Combinatorial
• TUnit.Combinatorial

Maybe the framework-specific packages should be suffixed with the framework name instead...

Am I completely off the rails here, or is there some sense to my proposal?

[AArnott]

It's an interesting idea.
I haven't succeeded in adopting TUnit in any of my repos yet. The combinatorial algorithm in this repo was actually taken from NUnit. Whether there's something shareable there or if it should just be forked again, I couldn't say until I study how TUnit discovery works.