Add rule for ?? nil

[jpsim]

This is always redundant as the rhs of the nil coalescing operator is only ever evaluated if the lhs is nil. This should really be a warning in the Swift compiler...

[daniel-beard]

I'd like to take this one on. Working on it now.

[jpsim]

Hmmm, actually there are cases where this is valid, for example when lhs is a double optional: Optional<Optional<T>>. Maybe we shouldn't make such a rule. Or if we do, it should be a warning & opt-in.

[daniel-beard]

I'll make it an opt-in rule for now

[BluMist]

nitpick: I believe coalesing should be changed to coalescing, you actually spelled it correctly once in the description

description: "nil coalescing operator is only evaluated if the lhs is nil " + ", coalesing operator with nil as rhs is redundant",

[marcelofabri]

This was implemented in #773 and could be closed

[norio-nomura]

Thanks!