(📚) What do all the symbols mean in the output of `reveal_type`

[KotlinIsland]

Documentation

reveal_type("shaking")  # Literal["shaking"]?

reveal_type(int("1"))  # builtins.int*

from typing import TypeVar
T = TypeVar("T")
def foo(t: T) -> T:
    reveal_type(t)  # T`-1

What do all these wacky symbols mean *, ? and

`-1

I'm guessing there might be others as well.

I looked through the docs and couldn't find anything about it.

Also <: and :>, maybe also def(int) -> str and def().

[JelleZijlstra]

* means the type is inferred, though I'm not sure that's really a useful concept.

The -1 stuff on the TypeVar is just giving it a unique name.

Not sure about the ?.

[KotlinIsland]

Could the TypeVar ones be upgraded to something like T@foo or something. just to make it a little more readable.

[erictraut]

FWIW, T@foo is the convention we settled on in pyright. The name after @ represents the scope of the TypeVar (a class, function or type alias name). We chose this based on experimentation and user feedback.

Then again, mypy users might be used to the -1 convention or might dislike the added verboseness.

[KotlinIsland]

Looks like this ? is a defect?

This is the code that generates it, and that comment seems entirely unrelated to the situation.

def visit_instance(self, t: Instance) -> str:
    if t.last_known_value and not t.args:
        # Instances with a literal fallback should never be generic. If they are,
        # something went wrong so we fall back to showing the full Instance repr.
        s = '{}?'.format(t.last_known_value)