This fixes the caching error at #86 by ensuring that the faketi has the same metaclass_type as the TypeInfo that is being faked. I also adjusted the MRO so that builtins.object is correctly ignored by the mypy code. (It expected the MRO to end with builtins.object)
Note: This does not fix the fact that mypy 1.0.0 will show a:
error: Metaclass conflict: the metaclass of a derived class must be a (non-strict) subclass of the metaclasses of all its bases [misc]
This only ensures that error shows up every time so that it can be # type: ignored if you want.
This fixes the caching error at #86 by ensuring that the faketi has the same metaclass_type as the TypeInfo that is being faked. I also adjusted the MRO so that builtins.object is correctly ignored by the mypy code. (It expected the MRO to end with builtins.object)
Note: This does not fix the fact that mypy 1.0.0 will show a:
This only ensures that error shows up every time so that it can be
# type: ignore
d if you want.Closes #86