Configuration for a Starlite application that features:
- SQLAlchemy 2.0
- SAQ async worker
- Lots of features!
This will install starlite-saqlalchemy
with minimal dependencies.
poetry add starlite-saqlalchemy
You can also install additional dependencies depending on the features you need:
# Repository implementation, DTOs
poetry add starlite-saqlalchemy[sqlalchemy]
# Async worker using saq
poetry add starlite-saqlalchemy[worker]
# Redis cache backend
poetry add starlite-saqlalchemy[cache]
# Sentry integration for starlite
poetry add starlite-saqlalchemy[sentry]
# or to install them all:
poetry add starlite-saqlalchemy[all]
from starlite import Starlite, get
from starlite_saqlalchemy import ConfigureApp
@get("/example")
def example_handler() -> dict:
"""Hello, world!"""
return {"hello": "world"}
app = Starlite(route_handlers=[example_handler], on_app_init=[ConfigureApp()])
The application configured in the above example includes the following configuration.
Receives and logs any unhandled exceptions raised out of route handling.
Integrates a Redis cache backend with Starlite first-class cache support.
Support filtering collection routes by created and updated timestamps, list of ids, and limit/offset pagination.
Includes an aggregate filters
dependency to easily inject all filters into a route handler, e.g,:
from starlite import get
from starlite_saqlalchemy.dependencies import FilterTypes
@get()
async def get_collection(filters: list[FilterTypes]) -> list[...]:
...
Configures Starlite's built-in Gzip compression support.
Exception handlers that translate non-Starlite repository and service object exception types into Starlite's HTTP exceptions.
A health check route handler that returns some basic application info.
Configures logging for the application including:
- Queue listener handler, appropriate for asyncio applications
- Health check route filter so that health check requests don't clog your logs
- An informative log format
- Configuration for dependency logs
Configures OpenAPI docs for the application, including config by environment to allow for easy personalization per application.
A response class that can handle serialization of SQLAlchemy/Postgres UUID types.
Just supply the DSN via environment, and Sentry is configured for you.
Engine, logging, pooling etc all configurable via environment. We configure starlite and include a
custom before_send
wrapper that inspects the outgoing status code to determine whether the
transaction that represents the request should be committed, or rolled back.
A customized SAQ queue and worker that is started and shutdown using the Starlite lifecycle event hooks - no need to run your worker in another process, we attach it to the same event loop as the Starlite app uses. Be careful not to do things in workers that will block the loop!
In addition to application config, the library include:
An abstract repository object type and a SQLAlchemy repository implementation.
A factory for building pydantic models from SQLAlchemy 2.0 style declarative classes. Use these to
annotate the data
parameter and return type of routes to control the data that can be modified per
route, and the information included in route responses.
http.Client
is a wrapper around httpx.AsyncClient
with some extra features including unwrapping
enveloped data, and closing the underlying client during shutdown of the Starlite application.
A SQLAlchemy declarative base class that includes:
- a mapping of the builtin
UUID
type to the postgresql dialect UUID type. - an
id
column - a
created
timestamp column - an
updated
timestamp column - an automated
__tablename__
attribute - a
from_dto()
class method, to ease construction of model types from DTO objects.
We also add:
- a
before_flush
event listener that ensures that theupdated
timestamp is touched on instances on their way into the database. - a constraint naming convention so that index and constraint names are automatically generated.
A Service object that integrates with the Repository ABC and provides standard logic for typical operations.
Configuration by environment.
All contributions big or small are welcome and appreciated! Please check out CONTRIBUTING.md
for
specific information about configuring your environment and workflows used by this project.