Validación

Request validation — schemas Pydantic

from pydantic import BaseModel, ConfigDict, EmailStr, Field


class ThingCreateSchema(BaseModel):
    name: str = Field(..., min_length=1, max_length=200)
    email: EmailStr
    age: int = Field(..., ge=0, le=120)
    model_config = ConfigDict(extra="ignore")

FastAPI valida automáticamente. Errores → RequestValidationError → handled por validation_exception_handler del app/utils/exception_handlers.py.

Response validation

from_attributes=True en BaseSchema permite serializar SQLAlchemy rows directo:

class ThingResponseSchema(BaseSchema):   # extends BaseSchema → from_attributes=True
    id: uuid.UUID
    name: str
    created_at: datetime

return self.format_response(ThingResponseSchema.model_validate(thing_orm_instance))

id type matches DB type

Si tu PK es Mapped[UUID], el schema DEBE tener id: uuid.UUID. Con id: str Pydantic falla silenciosamente.

Duplicate check — duplicate_check_fields

class UserService(BaseService):
    duplicate_check_fields = ["email"]

service.create(payload) consulta repo.get_by_filters({"email": payload.email}) antes de insertar. Si existe, lanza DatabaseIntegrityException(message="Registro ya existe").

Validation errors → BaseResponse

Handler global convierte RequestValidationError a:

{
  "data": [
    { "type": "missing", "loc": ["body", "email"], "msg": "Field required" }
  ],
  "message": "Error de validación",
  "status": "VALIDATION_ERROR"
}

Custom validators

Pydantic v2 field validators:

from pydantic import field_validator

class ThingCreateSchema(BaseModel):
    slug: str

    @field_validator("slug")
    @classmethod
    def slug_must_be_lowercase(cls, v: str) -> str:
        if v != v.lower():
            raise ValueError("slug must be lowercase")
        return v

Nested validation

class AddressSchema(BaseModel):
    city: str
    state: str


class UserCreateSchema(BaseModel):
    email: EmailStr
    address: AddressSchema   # nested

Pydantic valida recursivamente. Errores reportados con loc profundo: ["body", "address", "city"].