• Map and validate incoming request data to Pydantic (v1 and v2) models.
• framework-agnostic and comes with first-party support for Flask and AioHttp.
• Async ready.

# Add this in your app initialization code.
setup_mapper(
    integration=FlaskIntegration(
        # Automatically decorate views to avoid adding @map_request to every view.
        # Will not incur any performance overhead for views that don't use request-mapper.

        # If map_views is enabled, setup_mapper must be called after 
        # all views have been registered!
        map_views=True,
        # Register an error handler that returns 422 alongside pydantic validation errors
        # when the request cannot be mapped.
        add_error_handler=True,
    )
)


# Define models using Pydantic v1 or v2.
class PostCreateRequest(BaseModel):
    title: str
    content: str


class PostFilterQuery(BaseModel):
    status: PostStatus | None = None


@app.get("/posts")
# Map data from the current request.
def post_list_all(query: FromQuery[PostFilterQuery]) -> PaginatedResponse[Post]:
    # "query" is a valid pydantic model at this point.
    return PaginatedResponse(...)


@app.post("/posts")
def post_create(body: FromBody[PostCreateRequest]) -> PostCreateResponse:
    # "body" is a valid pydantic model at this point.
    return PostCreateResponse(...)

• pip install request-mapper.
• In your application setup, call mapper.setup_mapper with the integration of your choice.
• Decorate targets with @map_request (Optional when using flask integration).
• Map request data using one of the provided annotated types.
  • FromQuery[T] or Annotated[T, QueryStringMapping]
  • FromBody[T] or Annotated[T, RequestBodyMapping]
• Response is converted back to dict using Pydantic.
  • Override behavior by passing a custom response converter.

• Pull data from flask's request args, json and form data.
• Optional: Automatically map views without having to decorate them.
  • Views which do not use request mapper will not incur any performance penalty.
  • When using this feature the call to setup_mapper must be AFTER all views are registered.
• Optional: Set up an error handler for validation errors which makes the api respond with a 422.

• Async integration.
• Pull data from the current request.
• Note: When using function-based views, request must still be present as the first argument as required by aio.
• Class-based views support

  class HomeView(web.View):
      @map_request
      async def get(self, request: FromQuery[QueryDummyModel]) -> web.Response:
          return web.json_response({"query": request.query})

You can create your own integration by inheriting BaseIntegration and supplying an instance of that to the setup_mapper call.

Note that you must decorate your views with @map_request

A demo flask application is available at maldoinc/wireup-demo

Check out Wireup, a Dependency Injection library with a focus on developer experience, type safety and ease of use.