Checks

• I added a descriptive title to this issue
• I have searched (google, github) for similar issues and couldn't find anything
• I have read and followed the docs and still think this is a bug

Bug

Output of python -c "import pydantic.utils; print(pydantic.utils.version_info())":

             pydantic version: 1.8.1
            pydantic compiled: False
                 install path: /nix/store/1f6nk37c7df54cm8b3rlnn6pidaa3mrp-python3.8-pydantic-1.8.1/lib/python3.8/site-packages/pydantic
               python version: 3.8.8 (default, Feb 19 2021, 11:04:50)  [GCC 9.3.0]
                     platform: Linux-5.4.114-x86_64-with
     optional deps. installed: ['typing-extensions']

Here's the smallest program I could come up with that fails under from __future__ import annotations but works without it:

from __future__ import annotations

from pydantic import BaseModel, PositiveInt
from typing import NamedTuple

class Tup(NamedTuple):
    v: PositiveInt

class Mod(BaseModel):
    t: Tup

print(Mod.parse_obj({"t": [3]}))

The call to parse_obj raises this exception: pydantic.errors.ConfigError: field "v" not yet prepared so type is still a ForwardRef, you might need to call Tup.update_forward_refs(). But there is no update_forward_refs on NamedTuple instances.

Just using a base int in the NamedTuple doesn't exhibit the same problem, but of course it also doesn't check the constraints I want.

My current workaround is to just not use from __future__ import annotations in the module where I declare and use the NamedTuple, but I was hoping to use the new style consistently throughout the project.