eventbusk - Event Bus Framework

Keywords: event-bus, distributed, stream, processing, data, queue, kafka, python

Install

pip install git+https://github.com/Airbase/eventbusk.git

Quick Start

from eventbusk import Event, EventBus
from dataclasses import dataclass

# create an app instance of the bus
bus = EventBus(broker="kafka://localhost:9092")


# define an event as a dataclass
@dataclass
class Foo(Event):
    foo: int


# register the event to a single topic
bus.register_event("topic_foo", Foo)


# Define an method that receives that event
@bus.receive(event_type=Foo)
def process_a(event):
    logger.info(f"Foo: {event}")


# Publish an event to the bus
foo = Foo(foo=1)
bus.send(foo)

Contributing

Setting up locally

• uv
• Docker Desktop

Set up the project locally:

git clone git@github.com:Airbase/eventbusk.git
cd eventbusk
uv sync --extra dev
pre-commit install

Run example project

You will need a Confluent Kafka server locally via Docker by following https://docs.confluent.io/platform/current/platform-quickstart.html

There's an examples/docker-compose.yml that we can use to run a Kafka broker in a separate terminal window

cd examples
docker-compose up

Now you can run the example project consumers. Ensure the topics in the example are created first.

# See examples/eventbus.py
uv run eventbusk worker -A eventbus:bus

You can also publish a sample message:

uv run python

>>> from eventbus import bus, Fooey, Barzy
>>> bus.send(Fooey(foo_val="lorem ipsum"))
>>> bus.send(Barzey(bar_val="dolor sit amet"))

Code quality

After making code changes you can run some basic sanity checks as follows.

Run the tests:

uv run task test

Run the linter:

uv run task ruff
uv run task pylint

Format the code:

uv run task format

Run type checks:

uv run task typecheck

You can also choose running pre-commit manually, which runs all all of the above, among other things.

uv run pre-commit run --all-files