Python 3.9.8 causes TypeError

[ahopkins]

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

Our CI upgraded to v3.9.9. It comes back with the following error. It appears the problem was introduced in 3.9.8. It seems okay in 3.9.7.

pydantic/main.py:299: in pydantic.main.ModelMetaclass.__new__
pydantic/fields.py:411: in pydantic.fields.ModelField.infer
pydantic/fields.py:342: in pydantic.fields.ModelField.__init__
pydantic/fields.py:456: in pydantic.fields.ModelField.prepare
pydantic/fields.py:670: in pydantic.fields.ModelField.populate_validators
pydantic/validators.py:675: in find_validators
pydantic/dataclasses.py:266: in make_dataclass_validator
pydantic/dataclasses.py:255: in pydantic.dataclasses.dataclass
pydantic/dataclasses.py:250: in pydantic.dataclasses.dataclass.wrap
pydantic/dataclasses.py:148: in pydantic.dataclasses._process_class
pydantic/typing.py:287: in pydantic.typing.resolve_annotations
/usr/lib/python3.9/typing.py:292: in _eval_type
    return t._evaluate(globalns, localns, recursive_guard)
/usr/lib/python3.9/typing.py:553: in _evaluate
    type_ = _type_check(
/usr/lib/python3.9/typing.py:158: in _type_check
    raise TypeError(f"{arg} is not valid as type argument")
E   TypeError: typing.ClassVar[typing.Pattern[str]] is not valid as type argument

pydantic==1.8.2

[hellocoldworld]

This might be related to #3401 and #3403

[ahopkins]

@hellocoldworld Perfect! Thanks. Closing in favor of the alternative.