event-collector
===============

.. image:: https://img.shields.io/badge/version-0.2.3-blue
.. image:: https://img.shields.io/badge/license-MIT-blue

*event-collector* is a toolkit for creating lightweight event processing pipelines.

This library implements an in-memory `competing consumer
<https://learn.microsoft.com/en-us/azure/architecture/patterns/competing-consumers>`__
queue using only Python standard library functions and ``asyncio``
primitives. The "collector queue" runs as ``asyncio`` background tasks
and can autoscale according to load.  Using collector queues you can
set up data processing workflows quickly without having to deploy a
heavy stack of external services.

Because the collector queue runs alongside your data-processing
application or service -- i.e., in the same process -- it has the same
fault tolerance and latency guarantees as the rest of your code.

Repository
----------

* `codeberg.org/kchan/event-collector <https://codeberg.org/kchan/event-collector>`__

Requirements
------------

* >= Python 3.13

*event-collector* uses the following packages:

* ``orjson`` (for ``structlog`` JSON logging)
* ``structlog`` (for JSON logging)

How to install
--------------

Using ``pip``:

.. code-block::

    pip install event-collector


Or, if using ``uv``:

.. code-block::

    uv add event-collector


Contents
--------

.. toctree::
   :maxdepth: 2

   concepts
   getting-started

   event-collector-workflow
   tuning-the-queue
   api/index

.. note::

   This project is under active development.