Add docs for GHC-04584

[TauOmicronMu]

Closes #278

Think this reads alright :)

[BinderDavid]

Looks very good :) I think there are just 2 remaining small changes in capitalization.

[david-christiansen]

There's a general style thing to make it consistent with the rest of the docs. The example is really well-chosen.

[david-christiansen]

Here, I'd use a more specific text for this specific variation.

Suggested change

	Pattern matching allows constructors and literals to be used to take apart values, exposing their contents. The syntax of patterns is very similar to that of the expressions that create values, but many constructs that are allowed in expression contexts are not meaningful patterns.
	[Tuple sections](https://ghc.gitlab.haskell.org/ghc/doc/users_guide/exts/tuple_sections.html) are a concise syntax for functions that result in tuples. The missing expressions in the tuple become arguments to the function. Because tuple sections always result in functions, and functions cannot be matched in patterns, tuple sections are not allowed in patterns.

[TauOmicronMu]

Have removed the GHC-09646 stuff from this PR and put it onto a new branch - didn't mean for it to come through here. I'll make the requested changes before sending the new PR over though.

[david-christiansen]

Thanks!

[david-christiansen]

I would explain the intent behind the example more. E.g.

Suggested change

	When pattern matching, tuple sections are not allowed as patterns to be matched against. In the example below, we can fix this by adding the wildcard _ to match on the first element of the pair.
	When pattern matching, tuple sections are not allowed as patterns to be matched against. In the example below, one of the components of a tuple pattern was forgotten. We can fix this by adding the wildcard _ to match on the first element of the pair.

[david-christiansen]

Looks like we had a review race condition!

[david-christiansen]

@BinderDavid I'll let you approve before a merge