Inconsistency in `reveal_type(some_type)`

[denballakh]

Consider this example:

t1 = int
t2: type[int] = t1

reveal_type(int) # huge overload
reveal_type(t1) # huge overload
reveal_type(t2) # Type[builtins.int]

I expect get the same result in all three reveal_type's, but in two cases i get huge overload: Overload(def (Union[builtins.str, builtins.bytes, array.array[Any], mmap.mmap, ctypes._CData, pickle.PickleBuffer, typing.SupportsInt, typing_extensions.SupportsIndex, _typeshed.SupportsTrunc] =) -> builtins.int, def (Union[builtins.str, builtins.bytes], base: typing_extensions.SupportsIndex) -> builtins.int), and in one case i get Type[builtins.int]

My Environment

• mypy 0.961 (compiled: no)
• no command-line flags, no config file
• CPython 3.10.4
• Windows 10

[JukkaL]

The reason for the overload is that in the first instances type object references have the type Overloaded with is_type_object() true, while for t2 the type is TypeType. These types don't behave exactly the same, so displaying the exact same type wouldn't be perfect either. The current behavior is not very good either, since we lose information that the type refers to a type object specifically, not just any callable. I'm not sure what's the best option here -- one option would be to use type[Overload(...)] instead of just Overload(...), but that could be confusing as well.