Add Diamond Test Generator

[jpreese]

This issue is part of an overall initiative to complete #195.

The documentation on how to add generators can be found here. If you get stuck or have any questions about adding this generator, please do not hesitate to reach out!

[avjgit]

Current tests are property-based.
Current canonical testcases are value-based.
Should canonical data be updated by adding property-based testcases?
Should overwrite them?
What would be an impact on generators in other languages?

[ErikSchierboom]

Great observation. I think in this case we might not actually want to use the canonical data, as I think there is value in the property-based testing approach. @jpreese @robkeim what do you think?

[felix91gr]

Property Testing for the win!

It's so much stronger and descriptive than Case Testing.

That said, is it easy to follow for people doing the exercises? That should be the main criteria to apply here, right?
If it is, I think this is actually better in every way. It generalizes over all the domain of input. Unless the exercise itself changes, the canonical testcases will always be behind this exercise's test suite.

[jpreese]

Yeah, I would agree. This is a nice case to have where we can expose developers to property based testing (the tests themselves even use FsCheck).

[robkeim]

I agree we have a good reason for diverging from the canonical data here, and I think it would be sad to give up the opportunity of introducing/using property based testing just to conform with the canonical data.

When we go to solve #434 we'll need to have another status that is explicitly ignoring the canonical data.

[ErikSchierboom]

@robkeim Good idea!