CLVM derive support for enums

[Rigidity]

This PR adds support for deriving ToClvm and FromClvm on enums:

• All 3 layouts are supported, and can be set individually for each variant if desired.
• A discriminant (the numeric value of the enum) is used as the first value in each variant, to distinguish between the values.
• The discriminant defaults to isize and starts at 0 incrementing by 1 per variant unless overridden, similar to the defaults used by rustc.
• You can set the integer type used for the discriminant using the repr attribute (e.g. #[repr(u16)]).
• For a simple integer encoding, use unit variants with the tuple layout.
• For the list and tuple enums, you can use untagged to indicate that each of the variants are to be parsed individually and with no discriminator. This is useful for matching one of a set of possible values.

You can also derive these traits on unit structs, although this just encodes the value as if it had no fields and isn't too useful.

The quality of the code was also improved, and some of the duplicated logic was refactored out of each individual macro into the helpers file.

This should have no noticeable breaking changes in the way other derives work as far as I know.

[coveralls-official]

Pull Request Test Coverage Report for Build 7071376448

Warning: This coverage report may be inaccurate.

We've detected an issue with your CI configuration that might affect the accuracy of this pull request's coverage report.
To ensure accuracy in future PRs, please see these guidelines.
A quick fix for this PR: rebase it; your next report should be accurate.

• 553 of 565 (97.88%) changed or added relevant lines in 5 files are covered.
• 80 unchanged lines in 4 files lost coverage.
• Overall coverage increased (+0.3%) to 86.134%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
clvm-derive/src/helpers.rs	24	26	92.31%
clvm-derive/src/from_clvm.rs	247	252	98.02%
clvm-derive/src/to_clvm.rs	168	173	97.11%

Files with Coverage Reduction	New Missed Lines	%
clvm-derive/src/helpers.rs	1	81.03%
src/gen/conditions.rs	4	99.71%
clvm-traits/src/error.rs	6	12.5%
chia-protocol/src/program.rs	69	71.02%

Totals
Change from base Build 7020975872:	0.3%
Covered Lines:	10684
Relevant Lines:	12404

---

💛 - Coveralls

[arvidn]

I'm not all the way through this review yet. But here's what I've got so far.
Generally I think it would be good with more comments explaining what's going on. Macro code is especially unintuitive sometimes.

I appreciate the refactoring of fields() and repr_macros() to handle the different kinds of types in a uniform way (although I think they could both use comments describing what they do).

As for the named "untagged", maybe it would make sense to give it a name that describes the behavior instead. i.e. that it will perform a match instead. I don't know, untagged isn't so bad either (as along as it's well documented).

The exact behavior would ideally also be demonstrated in a test. regarding match-first, match-best or ambiguity being an error.