Replace equality checks on sealed objects with is checks (fixes #129)

[kyay10]

See #129 for the cause of the crash. This changes the code gen to do comparison checks on sealed objects using is instead of using equality. Also adds relevant tests.

[kyay10]

Some weird errors are appearing. I'll investigate and see what's happening since I can't reproduce them locally.
Edit: Seems like the issue is related to the initialization ordering of SealedEnum.values, this can be fixed by making it a computed property, but I don't know if that's too costly.

We could also either make it by lazy, or by having a private val that stores a cached version of values (That might be the least compromise).

This also means that we can't use values inside of the init of a sealed class, but that's already the behavior of enum classes (see this simple playground example. Note that, if you remove the values()[] and just access the ordinal, there are no exceptions, so this PR brings the same constraints that enum class has in terms of initialization)

[alexvanyo]

+1 to matching the constraints of an enum class.

I think a lazy with a LazyThreadSafetyMode.PUBLICATION sounds good for the values with relatively little overhead.

[kyay10]

Is there any automated way to update the *fileName*Generated properties in the tests btw? I'm resorting to updating them manually, but with such a wide-reaching change, it takes a bit of time.

[kyay10]

I think a lazy with a LazyThreadSafetyMode.PUBLICATION sounds good for the values with relatively little overhead.

Ended up using that, and it seems to work pretty well. All tests are passing locally. Also moved the Flag tests to usecases. Hopefully, workflow tests should pass now

[alexvanyo]

Is there any automated way to update the fileNameGenerated properties in the tests btw? I'm resorting to updating them manually, but with such a wide-reaching change, it takes a bit of time.

Unfortunately no, I've just manually updated them as well.

[kyay10]

Is there any automated way to update the _fileName_Generated properties in the tests btw? I'm resorting to updating them manually, but with such a wide-reaching change, it takes a bit of time.

Unfortunately no, I've just manually updated them as well.

I've opened #131 to suggest an automatic way of doing that.

[alexvanyo]

Looks great to me, thank you for the change here and the contribution!