Fix missing type casts when using bitfields with remapped types (and some other cases)

[Exanite]

This PR fixes #618 and some other issues that I identified as I was looking to understand the code.

#618 explains the two main cases fixed (missing type casts when remapping and incorrect sbyte case).

I also changed the uint/int case, mainly as an implementation detail. It doesn't matter if the cast happens before or after the left shift. It just has to happen before the right shift. This keeps the behavior consistent across cases.
Edit: I only changed the comment. The previous behavior already matches the current and the comment.

  // backing  int, current int            (value << cns) >> cns
  // backing  int, current uint           (uint)((value >> cns) & msk)

- // backing uint, current int            ((int)value << cns) >> cns
+ // backing uint, current int            (int)(value << cns) >> cns
  // backing uint, current uint           (value >> cns) & msk

  // backing uint, current byte           (byte)((value >> cns) & msk)
- // backing uint, current sbyte          (sbyte)((value << cns) >> cns)
+ // backing uint, current sbyte          (sbyte)((sbyte)(value << cns) >> cns)

You'll notice the code has changed a significant amount. This is because I rewrote most of the conditions and grouped the code together to make it a bit more readable. Because it was hard to track scopes, such as where each matching opening/closing parentheses were located, I added variables identifying the purpose and conditions for each scope.

Remaining tasks

• Separate tests into Unix/Windows variants. I've tested on both platforms by manually accounting for the platform differences (different bitfield packing; different enum backing types), but I haven't added tests for both platforms yet.
• Continue verifying that my new implementation works. It currently passes all test cases, but because of the larger changes made and how many conditions are required, I'm not yet confident that every case works.
  • Looks like CI tests failed. Windows tests failing is expected. Linux ARM tests failing is unexpected, but shouldn't be too hard to fix. I just need to research how C/C++ on Linux ARM behaves. Specifically, if char is a byte or sbyte on Linux ARM.
• Update branch with changes from main

[Exanite]

Ready for review.
All test cases pass.
Comparing current behavior with previous behavior shows that only the two targeted cases have been changed.

Since I changed the indentation, it's easier to see the changes if you ignore whitespace in your diff viewer.

[tannergooding]

Why the comparison against typeBacking rather than buildinTypeBacking? Won't that be meaningful for certain typedefs or enums?

[Exanite]

I'm going to take a look at this in more detail. I don't remember deciding against builtinTypeBacking, so it's likely something I missed when I was writing the condition. Changing typeBacking to builtinTypeBacking also doesn't affect the tests on either Linux nor Windows, so this probably is an untested case.

I'll draft the PR for now and undraft when I have an answer.

[tannergooding]

Overall LGTM. Just a minor nit and a question that I don't remember the answer to off the top of my head

[Exanite]

Good catch. Using builtinTypeBacking instead of typeBacking is in fact important.
The difference with typeBacking is that an unnecessary type cast gets inserted.

I've added a test case and changed the code to use builtinTypeBacking instead.