Still passes, even with :underspec. Reverting back to the tuple {:some, t} didn’t make any difference.
While the email specifies in mathematical detail what underspec is supposed to do, I still cannot find a real life scenario where I would want to use it.
I do think you are correct in the sense that [] | [t, ...] is the issue. I don’t see how though, since [] means empty list, and [t, ....] means a list with at least 1 element.
In my code, nil is neither of those cases.
The following code does raise a warning though:
@type option(t) :: some(t) | nothing
@type some(t) :: [t, ...]
@type nothing :: nil
@spec validate_name(String.t()) :: option(String.t())
def validate_name(name) do
if String.length(name) > 0 do
[name]
else
0
end
end
Warning
Extra type:
nil
Success typing:
[binary(), ...] | 0
But this is confusing. There is 1 path that leads to success. Dialyzer should not complain.
Yet, here it does because nil !== 0.
