Void as a "real" type

[SirOlaf]

Make a bunch of new things possible

type G[T] = object
  x: T
if G[void]().x is void: # true
  echo "It's a void"

proc `$`(x: void): string = "()"
var x: void
echo x # '()'
echo typeof(x) # 'void'

Also an alternative to #20385 and #20603

[Varriount]

So, what is a void-typed variable? How is it represented in backend code?

[SirOlaf]

Should stay as tyVoid until it disappears in cgen

[beef331]

Does this have any repercussions on offsetOf or sizeOf magics?

[SirOlaf]

If I do it right by some miracle, it's just a compiletime construct, so doesn't do anything but let you compare types (or use them in weird ways).
Best case I make it work, worst case it's out there so someone else might get inspired to try as well.

Edit:
addr for this says x and z are 8 bytes apart (64bit system), so it's like y doesn't exist while letting you "use" it

type H = object
    x: int
    y: void
    z: int

[Araq]

This is covered by nim-lang/RFCs#437

However it says

"Void type" -- We should optimize away the "empty tuple" type in the backend and remove this feature.

But fixing it instead is the better solution surely.

[SirOlaf]

Probably a controversial thing, these all work. Void gets treated as so optional you can skip it entirely

proc foo(voidparam: void, intparam: int) = echo intparam
foo(intparam=2)
foo(55)
var x: void
foo(x, 22)
foo(x, intparam=33)
foo(voidparam=x, 44)

[github-actions]

This pull request is stale because it has been open for 1 year with no activity. Contribute more commits on the pull request and rebase it on the latest devel, or it will be closed in 30 days. Thank you for your contributions.

[github-actions]

This pull request has been marked as stale and closed due to inactivity after 395 days.