CollectionViewSet — API Reference

from fastapi_viewsets.collection_viewset import CollectionViewSet

Class hierarchy

AsyncCollectionViewSet[K, T]  (ImplMixin[K, T], ABC)
└── CollectionViewSet[K, T]

AsyncCollectionViewSet

Abstract base. Subclasses must implement the five container_* methods.

class AsyncCollectionViewSet(Generic[K, T], ImplMixin[K, T], ABC):
    async def container_list(self) -> list[T]: ...
    async def container_retrieve(self, pk: K) -> T: ...
    async def container_add(self, data: T) -> T: ...
    async def container_update(self, pk: K, data: T, partial: bool) -> T: ...
    async def container_delete(self, pk: K) -> None: ...

All perform_* methods from ImplMixin are implemented here by delegating to the container_* methods.

CollectionViewSet

Concrete implementation backed by a Python collection.

Constructor

CollectionViewSet(container, pk_field="id")

Parameter	Type	Default	Description
container	Iterable[T] | MutableSequence[T] | MutableSet[T] | MutableMapping[K, T]	—	Backing collection
pk_field	str	"id"	Attribute/key name used as primary key

Supported container types

Type	Read	Write
Iterable (non-mutable)	✅	❌
MutableSequence (e.g. list)	✅	✅
MutableSet (e.g. set)	✅	✅
MutableMapping (e.g. dict)	✅	✅

A write to a non-mutable container raises ReadOnlyCollectionError (405, read_only). A partial update replaces a set's item with a copy (model_copy, _replace or dataclasses.replace) and raises TypeError for an item that is none of a pydantic model, a named tuple or a dataclass.

Auto-increment PK

Fields with json_schema_extra={"autoinc_int": True} are automatically assigned the next integer value on create when the field is None or 0.

class Item(BaseModel):
    id: int | None = Field(default=None, json_schema_extra={"autoinc_int": True})
    name: str

Errors

Raises NotFoundError(pk) (HTTP 404) when a record is not found during retrieve, update, or delete.