fast-twitch.middlewares.rate-limit

Usage and execution boundary

Consumes tokens per key before the handler. Allowed requests receive rate-limit metadata. A rejected request returns 429 with Retry-After by default. Stores are process-local atoms, not a distributed limiter.

(ns application.middleware
  (:require [fast-twitch.middlewares.rate-limit :as middleware]))

(defn handler [_]
  {:status 200 :headers {"content-type" "text/plain"} :body "Hello"})

(def app (middleware/wrap-rate-limit handler {:requests-per-second 10 :burst 20}))

This defines the Ring handler portion. Adapt app through routing/ft-handler or pass it to server/listen!; startup instructions and a complete walkthrough show the surrounding project.

Settings and defaults

Wrapper order and all middleware families. The function reference below preserves the upstream signatures and source links; helper functions do not all share the wrapper's arity.

Upstream function reference

Applies token-bucket rate limiting to requests.

client-key

(client-key request)

Function.

Returns the default rate limit key for a request. Source

default-error-response

The default response returned when a client exceeds its rate limit. Source

memory-store

(memory-store)

Function.

Creates an atom-backed in-memory rate limit store. Source

rate-limit-request

(rate-limit-request request result)

Function.

Associates rate limit metadata with a request. Source

rate-limit-response

(rate-limit-response request result options)

Function.

Builds the response returned when a request exceeds its rate limit. Source

rate-limit-result

(rate-limit-result store key rate burst)

Function.

Returns a map describing whether the request is allowed by the rate limit. Source

wrap-rate-limit

(wrap-rate-limit handler)
(wrap-rate-limit handler options)

Function.

Wraps a handler with per-key token-bucket rate limiting. Source