app = FastAPI(lifespan=lifespan)


@app.get("/items/{item_id}")
async def read_items(item_id: str):
    return items[item_id]


def test_read_items():
    # Before the lifespan starts, "items" is still empty
    assert items == {}

    with TestClient(app) as client:
        # Inside the "with TestClient" block, the lifespan starts and items added

In some cases, you might still have to use Pydantic's version of `dataclasses`. For example, if you have errors with the automatically generated API documentation.

In that case, you can simply swap the standard `dataclasses` with `pydantic.dataclasses`, which is a drop-in replacement:

{* ../../docs_src/dataclasses_/tutorial003_py310.py hl[1,4,7:10,13:16,22:24,27] *}

1. We still import `field` from standard `dataclasses`.

For example, imagine that you want to return an HTTP status code of "OK" `200` by default.

But if the data didn't exist, you want to create it, and return an HTTP status code of "CREATED" `201`.

But you still want to be able to filter and convert the data you return with a `response_model`.

For those cases, you can use a `Response` parameter.

## Use a `Response` parameter { #use-a-response-parameter }

Hiding your documentation user interfaces in production *shouldn't* be the way to protect your API.

That doesn't add any extra security to your API, the *path operations* will still be available where they are.

If there's a security flaw in your code, it will still exist.

        example: Annotated[
            Optional[Any],
            deprecated(
                "Deprecated in OpenAPI 3.1.0 that now uses JSON Schema 2020-12, "
                "although still supported. Use examples instead."
            ),
        ] = _Unset,
        openapi_examples: Optional[dict[str, Example]] = None,
        deprecated: Union[deprecated, str, bool, None] = None,

app = FastAPI(
    servers=[
        {"url": "/", "description": "Default, relative server"},
        {
            "url": "http://staging.localhost.tiangolo.com:8000",
            "description": "Staging but actually localhost still",
        },
        {"url": "https://prod.example.com"},
    ]
)


@app.get("/foo")
def foo():
    return {"message": "Hello World"}


client = TestClient(app)


def test_app():

When you return a `Response` directly its data is not validated, converted (serialized), or documented automatically.

But you can still document it as described in [Additional Responses in OpenAPI](additional-responses.md){.internal-link target=_blank}.

{* ../../docs_src/query_params_str_validations/tutorial002_an_py310.py hl[9] *}

Notice that the default value is still `None`, so the parameter is still optional.

But now, having `Query(max_length=50)` inside of `Annotated`, we are telling FastAPI that we want it to have **additional validation** for this value, we want it to have maximum 50 characters. 😎

/// tip

But it is still recommended to use the ideas above, using multiple classes, instead of these parameters.

This is because the JSON Schema generated in your app's OpenAPI (and the docs) will still be the one for the complete model, even if you use `response_model_include` or `response_model_exclude` to omit some attributes.

test.db
log.txt
Pipfile.lock
env3.*
env
docs_build
site_build
venv
docs.zip
archive.zip

# vim temporary files
*~
.*.sw?
.cache

# macOS
.DS_Store

# Ignore while the setup still depends on requirements.txt files
uv.lock