GC/nogc status in docs

[Adam D. Ruppe]

On Tuesday, 14 July 2020 at 13:35:12 UTC, Atila Neves wrote:
> I'm open to ideas on how to do it otherwise.

@nogc unittests are good, and the docs should probably show that 
too - if the declaration has a @nogc unittest attached, the doc 
generator could and show it is at least conditionally compatible.

A @nogc on the function itself would reject non- at nogc lambdas and 
such and that's unnecessarily limiting, but inference + unittest 
+ docs showing unittest annotations = win.

I'll prolly change adrdox later this week to do that, will be 
interesting to see how much of phobos will show conditional 
nogc/safe/pure as it is.