Skip to content

Modular, way of implementing rate-limiting in python with a few handy default implementations

License

Notifications You must be signed in to change notification settings

sigmavirus24/rush

Repository files navigation

rush: A library for throttles

Build status Test coverage Documentation

This library is a small collection of algorithms that can be reused when throttling user interactions with a resource (e.g., an API).

This library strives to allow any limiter and backing store to be used together without needing to be worried about potential compatibility.

Installation

pip install rush
pipenv install rush

Features

  • A basic periodic interval rate limiter - N accesses per period of time. An example would be the GitHub API that limits authenticated users to 5,000 requests per hour.
  • A leaky bucket rate limiter based off of the Generic Cell Ratelimiting Algorithm (a.k.a, GCRA).
  • A Redis storage backend for rate limit results so that users can have state persisted across machines and application restarts.
  • A in-memory dictionary storage backend for quick prototyping and testing.
  • Type annotations built into the library, verified with mypy, and distributed to users.

Quality

  • 100% test coverage
  • Code auto-formatted by Black (CI will check if formatting wasn't run prior to push)
  • Commit messages following a uniform Kernel-like style
  • Flake8, pylint, mypy, and bandit linting
  • Complete type annotations
  • Complete documentation linted by doclint and strictly compiled by Sphinx

Contributing

  • All contributors are expected to read and follow our Code of Conduct.

  • To reduce the initial friction of submitting a pull request:

    • Please run tox prior to submitting your pull request.
    • After a commit, please run tox -e commitlint.
  • To make it easier to support you, please gather the following information prior to filing an issue:

    • How you installed rush and the versions of its dependencies (if you're using the Redis store, please include rfc3986 and redis version information).
    • What stores and limiters are you using?
    • An example that reproduces your problem

About

Modular, way of implementing rate-limiting in python with a few handy default implementations

Topics

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Packages

No packages published

Languages