Going to and from json

[alonks1234]

Is there a way to extract json from .model_dump_json() and reconstruct / validate with .model_validate_json() ?

I am getting errors:

def array_validator(array: npt.NDArray) -> npt.NDArray:
>       if dimensions and (array_dimensions := len(array.shape)) != dimensions:
E       AttributeError: 'dict' object has no attribute 'shape'

Because the jsonified dict is not understood when validating.

        from pydantic_numpy.model import NumpyModel
        import pydantic_numpy as pnd
        from pydantic import Field
        import numpy as np
        
        class A(NumpyModel):
            numpy_arr: pnd.NpNDArrayFp16 = Field(frozen=True, strict=True, kw_only=False)
            
        a = A(numpy_arr=np.zeros(10))
        json_str = a.model_dump_json()
        a.model_validate_json(json_str)

Replication of what I would expect to be able to do

[caniko]

Fixed with #23

[alonks1234]

Thank you!