madtypes 0.0.9

Python typing that raise TypeError at runtime

madtypes

• 💢 MadType is a Python metaclass that does type-validation at run-time.

• 🌐 Generate Json-Schema

• 📖 Type hints cheat sheet

• 💪 32 tests for the features and usage of MadType class

• 💪 18 tests for the features and usage of json-schema function

• PEP589 does not perform type-checking.

TypedDict type definitions could plausibly used to perform runtime type checking of dictionaries. For example, they could be used to validate that a JSON object conforms to the schema specified by a TypedDict type. This PEP doesn’t include such functionality, since the focus of this proposal is static type checking only, and other existing types do not support this, as discussed in Class-based syntax. Such functionality can be provided by a third-party library using the typing_inspect third-party module, for example.

from typing import TypedDict
from madtypes import MadType

def test_simple_dict_incorrect_setattr(): # 🤯 DOES NOT RAISE ERROR 🤯
    class Simple(TypedDict):
        name: str

    Simple(name=2)
    a: Simple = { "name": 2 }


class Person(dict, metaclass=MadType): # 💢 MadType does !
    name: str


def test_mad_dict_type_error_with_incorrect_creation():
    with pytest.raises(TypeError):
        Person(name=2)

	Min	Max	Mean	Min (+)	Max (+)	Mean (+)
Correct instantiation	0.000	0.000	0.000	0.000 (18.1x)	0.000 (23.8x)	0.000 (17.3x)
Incorrect instantiation	0.000	0.000	0.000	0.000 (2.6x)	0.000 (3.7x)	0.000 (2.9x)

• :warning: MadType instanciation is much slower than pure Python.
• :warning: Manually adding type-check inside a class is more effective than using MadType

MadType is appropriate to apply when :

• The described data is a business related element

• You are using MadType to assert valid data

• You are debugging

• The instantiation occurs rarely

• The schema has to be communicated with the team

• json-schema

def test_object_json_schema():
    class Item(dict, metaclass=MadType):
        name: str

    assert json_schema(Item) == {
        "type": "object",
        "properties": {"name": {"type": "string"}},
        "required": ["name"],
    }

• Further customization

It is possible to use the MadType metaclass customize primitives as well.

class SomeStringAttribute(str, metaclass=MadType):
   pass

SomeDescriptedAttribute(2) # raise type error

• Field description

It is possible to use this to describe a field.

class SomeDescriptedAttribute(str, metaclass=MadType):
    annotation = str
    description = "Some description"

using json_schema on SomeDescription will include the description attribute

class DescriptedString(str, metaclass=MadType):
    description = "Some description"
    annotation = str

class DescriptedItem(Schema):
    descripted: DescriptedString

assert json_schema(DescriptedItem) == {
    "type": "object",
    "properties": {
        "descripted": {
            "type": "string",
            "description": "Some description",
        },
    },
    "required": ["descripted"],
}

• Regular expression

Regex can be defined on an Annotated type using the pattern attribute.

:warning: be careful to respect the json-schema specifications when using json_schema At the moment it is not checked nor tested, and will probably render an invalid json-schema without warning nor error

def test_pattern_definition_allows_normal_usage():
    class PhoneNumber(str, metaclass=MadType):
        annotation = str
        pattern = r"\d{3}-\d{3}-\d{4}"

    PhoneNumber("000-000-0000")


def test_pattern_raise_type_error():
    class PhoneNumber(str, metaclass=MadType):
        annotation = str
        pattern = r"\d{3}-\d{3}-\d{4}"

    with pytest.raises(TypeError):
        PhoneNumber("oops")


def test_pattern_is_rendered_in_json_schema():
    class PhoneNumber(str, metaclass=MadType):
        annotation = str
        pattern = r"^\d{3}-\d{3}-\d{4}$"
        description = "A phone number in the format XXX-XXX-XXXX"

    class Contact(Schema):
        phone: PhoneNumber

    schema = json_schema(Contact)
    print(json.dumps(schema, indent=4))
    assert schema == {
        "type": "object",
        "properties": {
            "phone": {
                "pattern": "^\\d{3}-\\d{3}-\\d{4}$",
                "description": "A phone number in the format XXX-XXX-XXXX",
                "type": "string",
            }
        },
        "required": ["phone"],
    }

• Object validation

It is possible to define a is_valid method on a Schema object, which is during instantiation to allow restrictions based on multiple fields.

def test_object_validation():
    class Item(dict, metaclass=MadType):
        title: Optional[str]
        content: Optional[str]

        def is_valid(self, **kwargs):
            """title is mandatory if content is absent"""
            if not kwargs.get("content", None) and not kwargs.get(
                "title", None
            ):
                raise TypeError(
                    "Either `Title` or `Content` are mandatory for Item"
                )

    Item(
        title="foo"
    )  # we should be able to create with only one of title or content
    Item(content="foo")
    with pytest.raises(TypeError):
        Item()

• Multiple inheritance

It is possible to create a schema from existing schemas.

:warning: careful not to use MadType of sub-classes as this would trigger and infinite recursion.

def test_multiple_inheritance():
    class Foo(dict):
        foo: str

    class Bar(dict):
        bar: str

    class FooBar(Foo, Bar, metaclass=MadType):
        pass

    FooBar(foo="foo", bar="bar")
    with pytest.raises(TypeError):
        FooBar()

• Dynamicly remove a field

Fields can be removed.

def test_fields_can_be_removed():
    @subtract_fields("name")
    class Foo(dict, metaclass=MadType):
        name: str
        age: int

    Foo(age=2)

Installation

pip3 install madtypes