dataclass decorator is broken in 1.7

[kittipatv]

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: 1.7
            pydantic compiled: True
                 install path: /Users/kittipat/pydantic_test/lib/python3.7/site-packages/pydantic
               python version: 3.7.9 (default, Sep 30 2020, 11:16:22)  [Clang 11.0.0 (clang-1100.0.33.17)]
                     platform: Darwin-19.6.0-x86_64-i386-64bit
     optional deps. installed: []

Prior to version 1.7, it was possible to subclass from stdlib dataclass. In version 1.7, doing so will result in error. Here is a minimal repro.

#!/usr/bin/env python3

from dataclasses import field, dataclass as vanilla_dataclass
from pydantic.dataclasses import dataclass

@vanilla_dataclass
class X:
    y: int = 0


@dataclass
class A(X):
    x: int = 1

@dataclass
class B:
    a: A = field(default_factory=lambda: A(x=2, y=3))


def main():
    print(B(**{"a": {"x": "4", "y": "5"}}))

if __name__ == "__main__":
    main()

Running with version 1.6

B(a=A(y=5, x=4))

Running with version 1.7

Traceback (most recent call last):
  File "dataclass.py", line 15, in <module>
    @dataclass
  File "pydantic/dataclasses.py", line 214, in pydantic.dataclasses.dataclass
    # name and type are filled in after the fact, not in __init__.
  File "pydantic/dataclasses.py", line 209, in pydantic.dataclasses.dataclass.wrap

  File "pydantic/dataclasses.py", line 138, in pydantic.dataclasses._process_class
    # +-------+-------+-------+--------+--------+
  File "dataclass.py", line 17, in <lambda>
    a: A = field(default_factory=lambda: A(x=2, y=3))
TypeError: __init__() got an unexpected keyword argument 'x'

[kittipatv]

It's possibly related to #744