test_dataclasses.py

import dataclasses
import pickle
from collections.abc import Hashable
from datetime import datetime
from pathlib import Path
from typing import Callable, ClassVar, Dict, FrozenSet, List, Optional

import pytest

import pydantic
from pydantic import BaseModel, ValidationError, validator


def test_simple():
    @pydantic.dataclasses.dataclass
    class MyDataclass:
        a: int
        b: float

    d = MyDataclass('1', '2.5')
    assert d.a == 1
    assert d.b == 2.5
    d = MyDataclass(b=10, a=20)
    assert d.a == 20
    assert d.b == 10


def test_model_name():
    @pydantic.dataclasses.dataclass
    class MyDataClass:
        model_name: str

    d = MyDataClass('foo')
    assert d.model_name == 'foo'
    d = MyDataClass(model_name='foo')
    assert d.model_name == 'foo'


def test_value_error():
    @pydantic.dataclasses.dataclass
    class MyDataclass:
        a: int
        b: int

    with pytest.raises(ValidationError) as exc_info:
        MyDataclass(1, 'wrong')

    assert exc_info.value.errors() == [
        {'loc': ('b',), 'msg': 'value is not a valid integer', 'type': 'type_error.integer'}
    ]


def test_frozen():
    @pydantic.dataclasses.dataclass(frozen=True)
    class MyDataclass:
        a: int

    d = MyDataclass(1)
    assert d.a == 1

    with pytest.raises(AttributeError):
        d.a = 7


def test_validate_assignment():
    class Config:
        validate_assignment = True

    @pydantic.dataclasses.dataclass(config=Config)
    class MyDataclass:
        a: int

    d = MyDataclass(1)
    assert d.a == 1

    d.a = '7'
    assert d.a == 7


def test_validate_assignment_error():
    class Config:
        validate_assignment = True

    @pydantic.dataclasses.dataclass(config=Config)
    class MyDataclass:
        a: int

    d = MyDataclass(1)

    with pytest.raises(ValidationError) as exc_info:
        d.a = 'xxx'
    assert exc_info.value.errors() == [
        {'loc': ('a',), 'msg': 'value is not a valid integer', 'type': 'type_error.integer'}
    ]


def test_not_validate_assignment():
    @pydantic.dataclasses.dataclass
    class MyDataclass:
        a: int

    d = MyDataclass(1)
    assert d.a == 1

    d.a = '7'
    assert d.a == '7'


def test_validate_assignment_value_change():
    class Config:
        validate_assignment = True

    @pydantic.dataclasses.dataclass(config=Config, frozen=False)
    class MyDataclass:
        a: int

        @validator('a')
        def double_a(cls, v):
            return v * 2

    d = MyDataclass(2)
    assert d.a == 4

    d.a = 3
    assert d.a == 6


def test_validate_assignment_extra():
    class Config:
        validate_assignment = True

    @pydantic.dataclasses.dataclass(config=Config, frozen=False)
    class MyDataclass:
        a: int

    d = MyDataclass(1)
    assert d.a == 1

    d.extra_field = 1.23
    assert d.extra_field == 1.23
    d.extra_field = 'bye'
    assert d.extra_field == 'bye'


def test_post_init():
    post_init_called = False

    @pydantic.dataclasses.dataclass
    class MyDataclass:
        a: int

        def __post_init__(self):
            nonlocal post_init_called
            post_init_called = True

    d = MyDataclass('1')
    assert d.a == 1
    assert post_init_called


def test_post_init_inheritance_chain():
    parent_post_init_called = False
    post_init_called = False

    @pydantic.dataclasses.dataclass
    class ParentDataclass:
        a: int

        def __post_init__(self):
            nonlocal parent_post_init_called
            parent_post_init_called = True

    @pydantic.dataclasses.dataclass
    class MyDataclass(ParentDataclass):
        b: int

        def __post_init__(self):
            super().__post_init__()
            nonlocal post_init_called
            post_init_called = True

    d = MyDataclass(a=1, b=2)
    assert d.a == 1
    assert d.b == 2
    assert parent_post_init_called
    assert post_init_called


def test_post_init_post_parse():
    post_init_post_parse_called = False

    @pydantic.dataclasses.dataclass
    class MyDataclass:
        a: int

        def __post_init_post_parse__(self):
            nonlocal post_init_post_parse_called
            post_init_post_parse_called = True

    d = MyDataclass('1')
    assert d.a == 1
    assert post_init_post_parse_called


def test_post_init_post_parse_types():
    @pydantic.dataclasses.dataclass
    class CustomType(object):
        b: int

    @pydantic.dataclasses.dataclass
    class MyDataclass:
        a: CustomType

        def __post_init__(self):
            assert type(self.a) == dict

        def __post_init_post_parse__(self):
            assert type(self.a) == CustomType

    d = MyDataclass(**{'a': {'b': 1}})
    assert d.a.b == 1


def test_post_init_assignment():
    from dataclasses import field

    # Based on: https://docs.python.org/3/library/dataclasses.html#post-init-processing
    @pydantic.dataclasses.dataclass
    class C:
        a: float
        b: float
        c: float = field(init=False)

        def __post_init__(self):
            self.c = self.a + self.b

    c = C(0.1, 0.2)
    assert c.a == 0.1
    assert c.b == 0.2
    assert c.c == 0.30000000000000004


def test_inheritance():
    @pydantic.dataclasses.dataclass
    class A:
        a: str = None

    @pydantic.dataclasses.dataclass
    class B(A):
        b: int = None

    b = B(a='a', b=12)
    assert b.a == 'a'
    assert b.b == 12

    with pytest.raises(ValidationError):
        B(a='a', b='b')


def test_validate_long_string_error():
    class Config:
        max_anystr_length = 3

    @pydantic.dataclasses.dataclass(config=Config)
    class MyDataclass:
        a: str

    with pytest.raises(ValidationError) as exc_info:
        MyDataclass('xxxx')

    assert exc_info.value.errors() == [
        {
            'loc': ('a',),
            'msg': 'ensure this value has at most 3 characters',
            'type': 'value_error.any_str.max_length',
            'ctx': {'limit_value': 3},
        }
    ]


def test_validate_assigment_long_string_error():
    class Config:
        max_anystr_length = 3
        validate_assignment = True

    @pydantic.dataclasses.dataclass(config=Config)
    class MyDataclass:
        a: str

    d = MyDataclass('xxx')
    with pytest.raises(ValidationError) as exc_info:
        d.a = 'xxxx'

    assert issubclass(MyDataclass.__pydantic_model__.__config__, BaseModel.Config)
    assert exc_info.value.errors() == [
        {
            'loc': ('a',),
            'msg': 'ensure this value has at most 3 characters',
            'type': 'value_error.any_str.max_length',
            'ctx': {'limit_value': 3},
        }
    ]


def test_no_validate_assigment_long_string_error():
    class Config:
        max_anystr_length = 3
        validate_assignment = False

    @pydantic.dataclasses.dataclass(config=Config)
    class MyDataclass:
        a: str

    d = MyDataclass('xxx')
    d.a = 'xxxx'

    assert d.a == 'xxxx'


def test_nested_dataclass():
    @pydantic.dataclasses.dataclass
    class Nested:
        number: int

    @pydantic.dataclasses.dataclass
    class Outer:
        n: Nested

    navbar = Outer(n=Nested(number='1'))
    assert isinstance(navbar.n, Nested)
    assert navbar.n.number == 1

    navbar = Outer(n=('2',))
    assert isinstance(navbar.n, Nested)
    assert navbar.n.number == 2

    navbar = Outer(n={'number': '3'})
    assert isinstance(navbar.n, Nested)
    assert navbar.n.number == 3

    with pytest.raises(ValidationError) as exc_info:
        Outer(n='not nested')
    assert exc_info.value.errors() == [
        {
            'loc': ('n',),
            'msg': 'instance of Nested, tuple or dict expected',
            'type': 'type_error.dataclass',
            'ctx': {'class_name': 'Nested'},
        }
    ]

    with pytest.raises(ValidationError) as exc_info:
        Outer(n=('x',))
    assert exc_info.value.errors() == [
        {'loc': ('n', 'number'), 'msg': 'value is not a valid integer', 'type': 'type_error.integer'}
    ]


def test_arbitrary_types_allowed():
    class Button:
        def __init__(self, href: str):
            self.href = href

    class Config:
        arbitrary_types_allowed = True

    @pydantic.dataclasses.dataclass(config=Config)
    class Navbar:
        button: Button

    btn = Button(href='a')
    navbar = Navbar(button=btn)
    assert navbar.button.href == 'a'

    with pytest.raises(ValidationError) as exc_info:
        Navbar(button=('b',))
    assert exc_info.value.errors() == [
        {
            'loc': ('button',),
            'msg': 'instance of Button expected',
            'type': 'type_error.arbitrary_type',
            'ctx': {'expected_arbitrary_type': 'Button'},
        }
    ]


def test_nested_dataclass_model():
    @pydantic.dataclasses.dataclass
    class Nested:
        number: int

    class Outer(BaseModel):
        n: Nested

    navbar = Outer(n=Nested(number='1'))
    assert navbar.n.number == 1


def test_fields():
    @pydantic.dataclasses.dataclass
    class User:
        id: int
        name: str = 'John Doe'
        signup_ts: datetime = None

    user = User(id=123)
    fields = user.__pydantic_model__.__fields__

    assert fields['id'].required is True
    assert fields['id'].default is None

    assert fields['name'].required is False
    assert fields['name'].default == 'John Doe'

    assert fields['signup_ts'].required is False
    assert fields['signup_ts'].default is None


def test_default_factory_field():
    @pydantic.dataclasses.dataclass
    class User:
        id: int
        aliases: Dict[str, str] = dataclasses.field(default_factory=lambda: {'John': 'Joey'})

    user = User(id=123)
    fields = user.__pydantic_model__.__fields__

    assert fields['id'].required is True
    assert fields['id'].default is None

    assert fields['aliases'].required is False
    assert fields['aliases'].default_factory() == {'John': 'Joey'}


def test_default_factory_singleton_field():
    class MySingleton:
        pass

    class MyConfig:
        arbitrary_types_allowed = True

    MY_SINGLETON = MySingleton()

    @pydantic.dataclasses.dataclass(config=MyConfig)
    class Foo:
        singleton: MySingleton = dataclasses.field(default_factory=lambda: MY_SINGLETON)

    # Returning a singleton from a default_factory is supported
    assert Foo().singleton is Foo().singleton


def test_schema():
    @pydantic.dataclasses.dataclass
    class User:
        id: int
        name: str = 'John Doe'
        aliases: Dict[str, str] = dataclasses.field(default_factory=lambda: {'John': 'Joey'})
        signup_ts: datetime = None
        age: Optional[int] = dataclasses.field(
            default=None, metadata=dict(title='The age of the user', description='do not lie!')
        )
        height: Optional[int] = pydantic.Field(None, title='The height in cm', ge=50, le=300)

    user = User(id=123)
    assert user.__pydantic_model__.schema() == {
        'title': 'User',
        'type': 'object',
        'properties': {
            'id': {'title': 'Id', 'type': 'integer'},
            'name': {'title': 'Name', 'default': 'John Doe', 'type': 'string'},
            'aliases': {
                'title': 'Aliases',
                'type': 'object',
                'additionalProperties': {'type': 'string'},
            },
            'signup_ts': {'title': 'Signup Ts', 'type': 'string', 'format': 'date-time'},
            'age': {
                'title': 'The age of the user',
                'description': 'do not lie!',
                'type': 'integer',
            },
            'height': {
                'title': 'The height in cm',
                'minimum': 50,
                'maximum': 300,
                'type': 'integer',
            },
        },
        'required': ['id'],
    }


def test_nested_schema():
    @pydantic.dataclasses.dataclass
    class Nested:
        number: int

    @pydantic.dataclasses.dataclass
    class Outer:
        n: Nested

    assert Outer.__pydantic_model__.schema() == {
        'title': 'Outer',
        'type': 'object',
        'properties': {'n': {'$ref': '#/definitions/Nested'}},
        'required': ['n'],
        'definitions': {
            'Nested': {
                'title': 'Nested',
                'type': 'object',
                'properties': {'number': {'title': 'Number', 'type': 'integer'}},
                'required': ['number'],
            }
        },
    }


def test_initvar():
    InitVar = dataclasses.InitVar

    @pydantic.dataclasses.dataclass
    class TestInitVar:
        x: int
        y: InitVar

    tiv = TestInitVar(1, 2)
    assert tiv.x == 1
    with pytest.raises(AttributeError):
        tiv.y


def test_derived_field_from_initvar():
    InitVar = dataclasses.InitVar

    @pydantic.dataclasses.dataclass
    class DerivedWithInitVar:
        plusone: int = dataclasses.field(init=False)
        number: InitVar[int]

        def __post_init__(self, number):
            self.plusone = number + 1

    derived = DerivedWithInitVar(1)
    assert derived.plusone == 2
    with pytest.raises(TypeError):
        DerivedWithInitVar('Not A Number')


def test_initvars_post_init():
    @pydantic.dataclasses.dataclass
    class PathDataPostInit:
        path: Path
        base_path: dataclasses.InitVar[Optional[Path]] = None

        def __post_init__(self, base_path):
            if base_path is not None:
                self.path = base_path / self.path

    path_data = PathDataPostInit('world')
    assert 'path' in path_data.__dict__
    assert 'base_path' not in path_data.__dict__
    assert path_data.path == Path('world')

    with pytest.raises(TypeError) as exc_info:
        PathDataPostInit('world', base_path='/hello')
    assert str(exc_info.value) == "unsupported operand type(s) for /: 'str' and 'str'"


def test_initvars_post_init_post_parse():
    @pydantic.dataclasses.dataclass
    class PathDataPostInitPostParse:
        path: Path
        base_path: dataclasses.InitVar[Optional[Path]] = None

        def __post_init_post_parse__(self, base_path):
            if base_path is not None:
                self.path = base_path / self.path

    path_data = PathDataPostInitPostParse('world')
    assert 'path' in path_data.__dict__
    assert 'base_path' not in path_data.__dict__
    assert path_data.path == Path('world')

    assert PathDataPostInitPostParse('world', base_path='/hello').path == Path('/hello/world')


def test_classvar():
    @pydantic.dataclasses.dataclass
    class TestClassVar:
        klassvar: ClassVar = "I'm a Class variable"
        x: int

    tcv = TestClassVar(2)
    assert tcv.klassvar == "I'm a Class variable"


def test_frozenset_field():
    @pydantic.dataclasses.dataclass
    class TestFrozenSet:
        set: FrozenSet[int]

    test_set = frozenset({1, 2, 3})
    object_under_test = TestFrozenSet(set=test_set)

    assert object_under_test.set == test_set


def test_inheritance_post_init():
    post_init_called = False

    @pydantic.dataclasses.dataclass
    class Base:
        a: int

        def __post_init__(self):
            nonlocal post_init_called
            post_init_called = True

    @pydantic.dataclasses.dataclass
    class Child(Base):
        b: int

    Child(a=1, b=2)
    assert post_init_called


def test_hashable_required():
    @pydantic.dataclasses.dataclass
    class MyDataclass:
        v: Hashable

    MyDataclass(v=None)
    with pytest.raises(ValidationError) as exc_info:
        MyDataclass(v=[])
    assert exc_info.value.errors() == [
        {'loc': ('v',), 'msg': 'value is not a valid hashable', 'type': 'type_error.hashable'}
    ]
    with pytest.raises(TypeError) as exc_info:
        MyDataclass()
    assert str(exc_info.value) == "__init__() missing 1 required positional argument: 'v'"


@pytest.mark.parametrize('default', [1, None, ...])
def test_hashable_optional(default):
    @pydantic.dataclasses.dataclass
    class MyDataclass:
        v: Hashable = default

    MyDataclass()
    MyDataclass(v=None)


def test_override_builtin_dataclass():
    @dataclasses.dataclass
    class File:
        hash: str
        name: Optional[str]
        size: int
        content: Optional[bytes] = None

    FileChecked = pydantic.dataclasses.dataclass(File)
    f = FileChecked(hash='xxx', name=b'whatever.txt', size='456')
    assert f.name == 'whatever.txt'
    assert f.size == 456

    with pytest.raises(ValidationError) as e:
        FileChecked(hash=[1], name='name', size=3)
    assert e.value.errors() == [{'loc': ('hash',), 'msg': 'str type expected', 'type': 'type_error.str'}]


def test_override_builtin_dataclass_2():
    @dataclasses.dataclass
    class Meta:
        modified_date: Optional[datetime]
        seen_count: int

    @pydantic.dataclasses.dataclass
    @dataclasses.dataclass
    class File(Meta):
        filename: str

    f = File(filename=b'thefilename', modified_date='2020-01-01T00:00', seen_count='7')
    assert f.filename == 'thefilename'
    assert f.modified_date == datetime(2020, 1, 1, 0, 0)
    assert f.seen_count == 7


def test_override_builtin_dataclass_nested():
    @dataclasses.dataclass
    class Meta:
        modified_date: Optional[datetime]
        seen_count: int

    @dataclasses.dataclass
    class File:
        filename: str
        meta: Meta

    class Foo(BaseModel):
        file: File

    FileChecked = pydantic.dataclasses.dataclass(File)
    f = FileChecked(filename=b'thefilename', meta=Meta(modified_date='2020-01-01T00:00', seen_count='7'))
    assert f.filename == 'thefilename'
    assert f.meta.modified_date == datetime(2020, 1, 1, 0, 0)
    assert f.meta.seen_count == 7

    with pytest.raises(ValidationError) as e:
        FileChecked(filename=b'thefilename', meta=Meta(modified_date='2020-01-01T00:00', seen_count=['7']))
    assert e.value.errors() == [
        {'loc': ('meta', 'seen_count'), 'msg': 'value is not a valid integer', 'type': 'type_error.integer'}
    ]

    foo = Foo.parse_obj(
        {
            'file': {
                'filename': b'thefilename',
                'meta': {'modified_date': '2020-01-01T00:00', 'seen_count': '7'},
            },
        }
    )
    assert foo.file.filename == 'thefilename'
    assert foo.file.meta.modified_date == datetime(2020, 1, 1, 0, 0)
    assert foo.file.meta.seen_count == 7


def test_override_builtin_dataclass_nested_schema():
    @dataclasses.dataclass
    class Meta:
        modified_date: Optional[datetime]
        seen_count: int

    @dataclasses.dataclass
    class File:
        filename: str
        meta: Meta

    FileChecked = pydantic.dataclasses.dataclass(File)
    assert FileChecked.__pydantic_model__.schema() == {
        'definitions': {
            'Meta': {
                'properties': {
                    'modified_date': {'format': 'date-time', 'title': 'Modified ' 'Date', 'type': 'string'},
                    'seen_count': {'title': 'Seen Count', 'type': 'integer'},
                },
                'required': ['modified_date', 'seen_count'],
                'title': 'Meta',
                'type': 'object',
            }
        },
        'properties': {
            'filename': {'title': 'Filename', 'type': 'string'},
            'meta': {'$ref': '#/definitions/Meta'},
        },
        'required': ['filename', 'meta'],
        'title': 'File',
        'type': 'object',
    }


def test_inherit_builtin_dataclass():
    @dataclasses.dataclass
    class Z:
        z: int

    @dataclasses.dataclass
    class Y(Z):
        y: int

    @pydantic.dataclasses.dataclass
    class X(Y):
        x: int

    pika = X(x='2', y='4', z='3')
    assert pika.x == 2
    assert pika.y == 4
    assert pika.z == 3


def test_dataclass_arbitrary():
    class ArbitraryType:
        def __init__(self):
            ...

    @dataclasses.dataclass
    class Test:
        foo: ArbitraryType
        bar: List[ArbitraryType]

    class TestModel(BaseModel):
        a: ArbitraryType
        b: Test

        class Config:
            arbitrary_types_allowed = True

    TestModel(a=ArbitraryType(), b=(ArbitraryType(), [ArbitraryType()]))


def test_forward_stdlib_dataclass_params():
    @dataclasses.dataclass(frozen=True)
    class Item:
        name: str

    class Example(BaseModel):
        item: Item
        other: str

        class Config:
            arbitrary_types_allowed = True

    e = Example(item=Item(name='pika'), other='bulbi')
    e.other = 'bulbi2'
    with pytest.raises(dataclasses.FrozenInstanceError):
        e.item.name = 'pika2'


def test_pydantic_callable_field():
    """pydantic callable fields behaviour should be the same as stdlib dataclass"""

    def foo(arg1, arg2):
        return arg1, arg2

    def bar(x: int, y: float, z: str) -> bool:
        return str(x + y) == z

    class PydanticModel(BaseModel):
        required_callable: Callable
        required_callable_2: Callable[[int, float, str], bool]

        default_callable: Callable = foo
        default_callable_2: Callable[[int, float, str], bool] = bar

    @pydantic.dataclasses.dataclass
    class PydanticDataclass:
        required_callable: Callable
        required_callable_2: Callable[[int, float, str], bool]

        default_callable: Callable = foo
        default_callable_2: Callable[[int, float, str], bool] = bar

    @dataclasses.dataclass
    class StdlibDataclass:
        required_callable: Callable
        required_callable_2: Callable[[int, float, str], bool]

        default_callable: Callable = foo
        default_callable_2: Callable[[int, float, str], bool] = bar

    pyd_m = PydanticModel(required_callable=foo, required_callable_2=bar)
    pyd_dc = PydanticDataclass(required_callable=foo, required_callable_2=bar)
    std_dc = StdlibDataclass(required_callable=foo, required_callable_2=bar)

    assert (
        pyd_m.required_callable
        is pyd_m.default_callable
        is pyd_dc.required_callable
        is pyd_dc.default_callable
        is std_dc.required_callable
        is std_dc.default_callable
    )
    assert (
        pyd_m.required_callable_2
        is pyd_m.default_callable_2
        is pyd_dc.required_callable_2
        is pyd_dc.default_callable_2
        is std_dc.required_callable_2
        is std_dc.default_callable_2
    )


def test_pickle_overriden_builtin_dataclass(create_module):
    module = create_module(
        # language=Python
        """\
import dataclasses
import pydantic


@dataclasses.dataclass
class BuiltInDataclassForPickle:
    value: int

class ModelForPickle(pydantic.BaseModel):
    # pickle can only work with top level classes as it imports them

    dataclass: BuiltInDataclassForPickle

    class Config:
        validate_assignment = True
        """
    )
    obj = module.ModelForPickle(dataclass=module.BuiltInDataclassForPickle(value=5))

    pickled_obj = pickle.dumps(obj)
    restored_obj = pickle.loads(pickled_obj)

    assert restored_obj.dataclass.value == 5
    assert restored_obj == obj

    # ensure the restored dataclass is still a pydantic dataclass
    with pytest.raises(ValidationError, match='value\n +value is not a valid integer'):
        restored_obj.dataclass.value = 'value of a wrong type'