This release adds support for Sanic v21

Fixes returning "500 Internal Server Error" responses to requests with malformed json when running with ASGI integration.

This release adds python_name to the Info type.

Fix the Pydantic conversion method for Enum values, and add a mechanism to specify an interface type when converting from Pydantic. The Pydantic interface is really a base dataclass for the subclasses to extend. When you do the conversion, you have to use strawberry.experimental.pydantic.interface to let us know that this type is an interface. You also have to use your converted interface type as the base class for the sub types as normal.

Fixes a bug with the selected_fields property on info when an operation
variable is not defined.

Issue #1248.

Fix issues (#1158 and #1104) where Generics using LazyTypes
and Enums would not be properly resolved

These now function as expected:

Enum

T = TypeVar("T")

@strawberry.enum
class VehicleMake(Enum):
    FORD = 'ford'
    TOYOTA = 'toyota'
    HONDA = 'honda'

@strawberry.type
class GenericForEnum(Generic[T]):
    generic_slot: T

@strawberry.type
class SomeType:
    field: GenericForEnum[VehicleMake]

LazyType

another_file.py

@strawberry.type
class TypeFromAnotherFile:
    something: bool

this_file.py

T = TypeVar("T")

@strawberry.type
class GenericType(Generic[T]):
    item: T

@strawberry.type
class RealType:
    lazy: GenericType[LazyType["TypeFromAnotherFile", "another_file.py"]]

This release adds fragment and input variable information to the
selected_fields attribute on the Info object.

Fixes a bug in the Pydantic conversion code around Union values.

Fixes a bug in the export-schema command around the handling of local modules.

Fixes a bug in the Pydantic conversion code around complex Optional values.

This release adds a new exception called InvalidFieldArgument which is raised when a Union or Interface is used as an argument type.
For example this will raise an exception:

import strawberry

@strawberry.type
class Noun:
    text: str

@strawberry.type
class Verb:
    text: str

Word = strawberry.union("Word", types=(Noun, Verb))

@strawberry.field
def add_word(word: Word) -> bool:
	...

Fix type resolution when inheriting from types from another module using deferred annotations.

This release adds support for Pyright and Pylance, improving the
integration with Visual Studio Code!

Change the version constraint of opentelemetry-sdk and opentelemetry-api to <2

This release adds support for enabling subscriptions in GraphiQL
on Django by setting a flag subscriptions_enabled on the BaseView class.

from strawberry.django.views import AsyncGraphQLView

from .schema import schema

urlpatterns = [path("graphql", AsyncGraphQLView.as_view(schema=schema, graphiql=True, subscriptions_enabled=True))]

This release fixes an issue with the MyPy plugin that prevented using
TextChoices from django in strawberry.enum.

This release improves how we deal with custom scalars. Instead of being global
they are now scoped to the schema. This allows you to have multiple schemas in
the same project with different scalars.

Also you can now override the built in scalars with your own custom
implementation. Out of the box Strawberry provides you with custom scalars for
common Python types like datetime and Decimal. If you require a custom
implementation of one of these built in scalars you can now pass a map of
overrides to your schema:

from datetime import datetime, timezone
import strawberry

EpochDateTime = strawberry.scalar(
    datetime,
    serialize=lambda value: int(value.timestamp()),
    parse_value=lambda value: datetime.fromtimestamp(int(value), timezone.utc),
)

@strawberry.type
class Query:
    @strawberry.field
    def current_time(self) -> datetime:
        return datetime.now()

schema = strawberry.Schema(
  Query,
  scalar_overrides={
    datetime: EpochDateTime,
  }
)
result = schema.execute_sync("{ currentTime }")
assert result.data == {"currentTime": 1628683200}

This release allows to install Strawberry along side click version 8.

This release add full support for async directives and fixes and issue when
using directives and async extensions.

@strawberry.type
class Query:
    name: str = "Banana"

@strawberry.directive(
    locations=[DirectiveLocation.FIELD], description="Make string uppercase"
)
async def uppercase(value: str):
    return value.upper()

schema = strawberry.Schema(query=Query, directives=[uppercase])

Fix issue where strawberry.Private fields on converted Pydantic types were not added to the resulting dataclass.