Apply formatting and linting

[dragospe]

Following some discussion at Zurihac, I wanted to give my take on the formatting/linting question.

This change will impact practically every file in the codebase, which can be difficult for contributors with open PRs. I've split up this PR into two separate commits:

• 3f03d37: this is the commit that adds the formatting/linting utilities, and updates the Makefile with relevant commands and the README with some documentation
• 01213d2: runs the make format and make lint commands. THIS BREAKS THINGS.

I'm marking this currently as draft, because there are a few things I still need to work out.

• Add this to CI. Having formatting/linting goes out of sync quickly if its not enforced.
• Determine what type(s) of breakage come from using these tools. Currently, all of the haskell compiles, but some of the tests don't run. This is may actually highlight some issues with Agda2hs itself, too; or it may just indicate that hlint/fourmolu/cabal-fmt don't work well with the generated code.

I think there was someone else who also had ideas here. Happy to chat and compare approaches -- the setup here is just my default when I'm starting new projects.

[jespercockx]

In general I think this is a good idea and we should investigate further. My main concern with the output of the auto-formatter is the extra indentation it sometimes creates, which is something I am personally quite sensitive to.

Also, we should exclude the files in the golden directory from the auto-formatter, though we could consider applying auto-formatting to the output of agda2hs when generating the Haskell code.

[jespercockx]

This seems to create a lot of extra indentation that I'd rather avoid.

[jespercockx]

We should not apply the linting to the files in the golden directory, since these are the expected outputs of the test suite. However, we could consider applying the auto-formatter to the code generated by agda2hs (which would then lead to an update of the golden files too).

[anka-213]

..., though we could consider applying auto-formatting to the output of agda2hs when generating the Haskell code.

I do agree with this part, since, even though the formatting is generally quite decent, the formatting of the output from agda2hs can sometimes be fairly ugly, especially for things like large records or long type-lines.