Struct MultiRateLimiter

pub struct MultiRateLimiter<K> { /* private fields */ }

MultiRateLimiter enables key-based rate limiting, where each key has its own RateLimiter.

This behavior is useful when you want to throttle a set of keys independently, for example you may have a web crawler that wants to throttle its requests to each domain independently.

Examples

use async_throttle::MultiRateLimiter;
use std::sync::Arc;

#[tokio::main]
async fn main() {
   let period = std::time::Duration::from_secs(5);
   let rate_limiter = MultiRateLimiter::new(period);
    
   // This completes instantly
   rate_limiter.throttle("foo", || computation()).await;

   // This completes instantly
   rate_limiter.throttle("bar", || computation()).await;

   // This takes 5 seconds to complete because the key "foo" is rate limited
   rate_limiter.throttle("foo", || computation()).await;
}

async fn computation() { }

Implementations

impl<K: Eq + Hash + Clone> MultiRateLimiter<K>

pub fn new(period: Duration) -> Self

Creates a new MultiRateLimiter.

pub async fn throttle<Fut, F, T>(&self, key: K, f: F) -> T

where Fut: Future<Output = T>, F: FnOnce() -> Fut,

Throttles the execution of a function based on a key. Throttling is key-specific, so multiple keys can be throttled independently.

Examples

use async_throttle::MultiRateLimiter;
use anyhow::Result;
use std::sync::Arc;

async fn do_work() { /* some computation */ }

async fn throttle_by_key(the_key: u32, limiter: Arc<MultiRateLimiter<u32>>) {
   limiter.throttle(the_key, || do_work()).await
}