Kafkit¶

Kafkit helps you write Kafka producers and consumers in Python with asyncio:

Kafkit integrates aiokafka consumers and producers with the Confluent Schema Registry. The Deserializer class can deserialize messages with any schema that’s registered in a Confluent Schema Registry. The Serializer class can serialize Python objects against a single Avro schema, while the PolySerializer class is flexible enough to handle multiple schemas.
Kafkit provides Python APIs for working with the Confluent Schema Registry’s HTTP API. The RegistryApi client includes high-level methods that manage subjects and their schemas in a registry. These methods are cached so that the RegistryApi client can be an integral part of your application’s schema management. Additionally, RegistryApi includes low-level HTTP methods (GET, POST, PUT, PATCH, DELETE) so you can work directly with the Confluent Schema Registry API if you want.
kafkit.registry.aiohttp.RegistryApi is implemented with aiohttp, but that’s not the only implementation. Kafkit subscribes to the sans IO architecture (gidgethub is a popular example) meaning that you can subclass kafkit.registry.sansio.RegistryApi to integrate with your favorite HTTP client library. The kafkit.registry.sansio.MockRegistryApi is a mock client that you can use in your app’s unit tests.

Installation¶

Install Kafkit with aiohttp:

pip install kafkit[aiohttp]