How to Build a Data Validation Schema in Python

import re
from dataclasses import dataclass, field
from typing import Any, Callable


@dataclass
class Field:
    name: str
    validator: Callable[[Any], bool]
    required: bool = True

    def validate(self, value: Any) -> bool:
        if not self.required and value is None:
            return True
        return self.validator(value)


@dataclass
class Schema:
    fields: list[Field] = field(default_factory=list)

    def add(self, name: str, validator: Callable[[Any], bool], required: bool = True) -> None:
        self.fields.append(Field(name, validator, required))

    def validate(self, data: dict) -> dict[str, str]:
        errors = {}
        for field_def in self.fields:
            value = data.get(field_def.name)
            if not field_def.validate(value):
                errors[field_def.name] = f"Invalid value: {value!r}"
        return errors


def is_email(value: str) -> bool:
    return bool(re.match(r"^[^@]+@[^@]+\.[^@]+$", value))


if __name__ == "__main__":
    schema = Schema()
    schema.add("email", is_email)
    schema.add("age", lambda v: isinstance(v, int) and 0 <= v <= 120, required=False)
    schema.add("name", lambda v: isinstance(v, str) and len(v) > 0)

    sample = {"email": "user@example.com", "age": 30, "name": "Alice"}
    print("Valid data errors:", schema.validate(sample))

    bad_sample = {"email": "not-an-email", "age": -5, "name": ""}
    print("Invalid data errors:", schema.validate(bad_sample))