[red-knot] understand more special forms in annotation expressions

[carljm]

Feel free to create and link a sub-task to claim one of these to work on:

• Unions via |
• [red-knot] Understand typing.Union special form in annotations #14498
• [red-knot] Understand typing.Optional special form in annotations #14354
• [red-knot] support typing.Any in annotations #14544
• [red-knot] support typing.NoReturn and typing.Never in annotations #14558
• [red-knot] understand type[a.X] in a type expression #14545
• [red-knot] understand type[A | B] where A and B are both classes #14703
• [red-knot] Understand type[Union[A, B]] where A and B are both classes #14834
• [red-knot] support type[Any] #14546
• [red-knot] Understand Annotated #14922
• [red-knot] Understand typing.Type #14891
• [red-knot] Understand typing.Tuple #14916

[Glyphack]

To understand better: Do we want to follow the same approach for typing.Any and infer it as a known instance?
Any is assigned an object. If we leave it as is then we just need to infer the KnownClass::Object in the annotations.

[AlexWaygood]

To understand better: Do we want to follow the same approach for typing.Any and infer it as a known instance?

I think so. It's important that we distinguish internally between "members of the type Any" and "the singleton symbol that actually represents the Any type itself". Inferring the symbol typing.Any as a known instance allows us to do that.