automatically add fields_validators based on hint type with pydantic

I’d like to define once for all fields_validators function in a BaseModel class, and inherit this class in my model, and the validators should apply to the updated class attributes.

MWE

def to_int(v: Union[str, int]) -> int:
    if isinstance(v, str):
        if v.startswith("0x"):
            return int(v, 16)
        return int(v)
    return v


def to_bytes(v: Union[str, bytes, list[int]]) -> bytes:
    if isinstance(v, bytes):
        return v
    elif isinstance(v, str):
        if v.startswith("0x"):
            return bytes.fromhex(v[2:])
        return v.encode()
    else:
        return bytes(v)


class BaseModelCamelCase(BaseModel):
    model_config = ConfigDict(
        populate_by_name=True,
        alias_generator=AliasGenerator(
            validation_alias=lambda name: AliasChoices(to_camel(name), name)
        ),
    )

    # FIXME: should apply to int type from get_type_hints only
    @field_validator("*", mode="before")
    def to_int(cls, v: Union[str, int]) -> int:
        return to_int(v)

    # FIXME: should apply to bytes type from get_type_hints only
    @field_validator("*", mode="before")
    def to_bytes(cls, v: Union[str, bytes, list[int]]) -> bytes:
        return to_bytes(v)

class BaseTransactionModel(BaseModelCamelCase):
    nonce: int
    gas: int = Field(validation_alias=AliasChoices("gasLimit", "gas_limit", "gas"))
    to: Optional[bytes]
    value: int
    data: bytes
    r: int = 0
    s: int = 0

I tried to use model_validate but then I lost the alias parsing

>Solution :

To automatically add field validators based on type hints in Pydantic, you can use the Annotated type from typing_extensions along with Pydantic’s validator functions. This allows you to bind validation logic directly to a type, making your code more modular and reusable.

Here’s an example of how you can achieve this:

from typing import Any, List
from typing_extensions import Annotated
from pydantic import BaseModel, ValidationError
from pydantic.functional_validators import AfterValidator

# Define a validator function
def check_positive(value: int) -> int:
    if value <= 0:
        raise ValueError(f'{value} is not a positive number')
    return value

# Create an annotated type with the validator
PositiveInt = Annotated[int, AfterValidator(check_positive)]

# Use the annotated type in a Pydantic model
class MyModel(BaseModel):
    positive_number: PositiveInt

# Example usage
try:
    model = MyModel(positive_number=10)
    print(model)
except ValidationError as e:
    print(e)

try:
    model = MyModel(positive_number=-5)
except ValidationError as e:
    print(e)

Explanation:
Validator Function: check_positive ensures the value is a positive integer.
Annotated Type: PositiveInt combines the int type with the check_positive validator using Annotated.
Pydantic Model: MyModel uses PositiveInt for the positive_number field, automatically applying the validator.
Benefits:
Modularity: Validators are tied to types, making them reusable across different models.
Clarity: The validation logic is separated from the model definition, improving readability.