Stream-Framework: Build Activity Feeds with Python

Overview

Stream-Framework provides building blocks for applications that publish activities to user feeds, such as social networks, newsfeeds, and notification systems. It addresses the work of distributing new activities to followers and retrieving feed entries, with Redis and/or Cassandra as storage options.

It is intended for Python teams that want to manage feed behavior themselves, including follower lookup and feed definitions. Feed distribution can run as asynchronous Celery tasks, making the framework a fit when write-heavy fanout is acceptable and read paths need to stay straightforward.

Key Features

• Models activities and stores them in configurable feed classes.
• Supports Redis and Cassandra storage backends.
• Provides a manager abstraction for publishing user activities and distributing them to followers.
• Supports asynchronous fanout through Celery tasks.
• Allows applications to define follower lookup and fanout priorities.
• Offers optional Redis and Cassandra dependency sets for installation.

Use Cases

• A Python social application can publish each user's posts to the feeds of their followers.
• A product team can build a Twitter-, Instagram-, or Pinterest-style feed with application-specific activity types.
• An application can maintain personal activity streams, such as a GitHub-style history of events.
• A service can distribute notification-like activities to user feeds when it can manage Redis or Cassandra and background task workers.

Project Facts

• Language: Python
• License: NOASSERTION
• Stars: 4.7k
• Forks: 528
• Topics: activity-feed, activity-stream, big-data, cassandra, feed, news, news-feed, redis
• Archived: false

Getting Started

Install the base package with pip:

pip install stream-framework

Add the redis and/or cassandra extras for the backends you plan to use. See the README and documentation for configuration and usage details.

Considerations

• Redis and Cassandra dependencies are optional and must be installed for the corresponding backends. The example also relies on Celery for background fanout.
• This framework is for teams willing to operate and configure feed storage and task-processing infrastructure. The project authors also offer a hosted feed service, but that is a separate product.
• The README describes Cassandra 2.1 as the performance target and says Cassandra 2.2 and 3.3 pass the test suite without production experience. Check backend and dependency compatibility against your environment before adopting it.