Generate ToMaybe in deriveEsqueletoRecord

[halogenandtoast]

replaces #370

• Bumped the version number.
• Documented new APIs with Haddock markup.
• Added @since declarations to the Haddock.
• Ran stylish-haskell and otherwise adhered to the style guide.

After submitting your PR:

• Update the Changelog.md file with a link to your PR.
• Check that CI passes (or if it fails, for reasons unrelated to your change, like CI timeouts).

[halogenandtoast]

Opening this up for conversation since the draft didn't get eyes and I think this is nearly there.

[parsonsmatt]

Could you write a test that shows it correctly working with Nothing values? ie where the DB gives us NULL back?

[parsonsmatt]

We probably need to gate this behind CPP for the GHC version since we support GHC that doesn't have this feature yet

[halogenandtoast]

I assume gating the whole file makes sense here?

[halogenandtoast]

@parsonsmatt did you mean something different from: https://github.com/bitemyapp/esqueleto/pull/378/files#diff-bef0a2133965554f3b49d699a015a29f235548c9ae5edc52ce80cd348332720dR282-R300

[halogenandtoast]

Confirmed this compiles and all existing tests run and pass on 8.10.7

[parsonsmatt]

The feature is supported in GHC <9.2, but you do need to use getField @"blah" explicitly instead of (.blah). So we need to gate use of OverloadedRecordDot behind CPP to support older GHC, but not the entire feature/tests.

Actually, it is probably easier to just getField directly? Then there's no CPP necessary

[halogenandtoast]

I don't have a machine I can run 8.6.5 on until November 4th so fixing this will have to wait until then unless someone else can do the update here.