Branches

Apache Asyncband (Incubating)

[!IMPORTANT]
Apache Asyncband (incubating) is an effort undergoing incubation at the Apache Software Foundation (ASF), sponsored by the Apache Incubator PMC.
Please read the DISCLAIMER and a full explanation of “incubating”.
Asyncband was formerly published as MEA. The mea crate is deprecated and receives no further development. See the migration guide for migration instructions and details about the rename.

Overview

Asyncband is a focused collection of composable, runtime-agnostic concurrency building blocks for async Rust. It provides synchronization, initialization, task coordination, channels, resource reuse, and workload control without choosing an executor for the application.

Asyncband's async APIs are built on standard futures and wakers. The library does not spawn tasks, own worker threads, install timers, or require a reactor or I/O driver. Applications can poll its futures with Tokio, async-std, smol, a custom executor, or any other standards-based runtime, and compose runtime services such as deadlines around them.

Project scope

The project is not limited to small or stateless synchronization primitives. Stateful utilities such as a singleflight group or an object pool fit when they provide a generally reusable coordination mechanism and remain independent of executor policy.

The boundary is mechanism versus policy. Task placement, timers, deadlines, retries, periodic maintenance, and application lifecycle orchestration stay with the caller and its runtime. Potential future-concurrency or scheduling APIs are evaluated against the same boundary: they must remain executor-independent and compose with caller-owned execution and timing.

Getting started

The crate enables no APIs by default. Enable only the features your application uses:

cargo add asyncband --features mutex

use asyncband::mutex::Mutex;

async fn increment() {
    let counter = Mutex::new(0);
    *counter.lock().await += 1;
    assert_eq!(*counter.lock().await, 1);
}

Public paths stay direct—such as asyncband::mutex, asyncband::pool, and asyncband::once::OnceCell—while Cargo features keep unused implementations out of the build.

Examples

Runnable examples live in the examples workspace crate. They demonstrate how to choose and compose Asyncband primitives in complete programs.

API map

Area	API	Feature	Use
Locks and conditions	`Mutex`	`mutex`	Protect shared data with asynchronous mutual exclusion.
	`RwLock`	`rwlock`	Allow multiple readers or one writer.
	`Condvar`	`condvar`	Wait for notifications while releasing a mutex.
Coordination	`Semaphore`	`semaphore`	Limit concurrent work by acquiring permits.
	`Barrier`	`barrier`	Synchronize a fixed number of participants at a reusable rendezvous.
	`ManualResetEvent`	`event`	Signal current and future waits until explicitly reset.
	`Latch`	`latch`	Wait until a fixed one-way countdown reaches zero.
	`Phaser`	`phaser`	Coordinate repeated phases with a dynamic participant set.
	`WaitGroup`	`waitgroup`	Dynamically register participants and wait until all have completed.
	`Shutdown`	`shutdown`	Request shutdown and wait until all completion guards are dropped.
Work coalescing	`Once`	`once`	Complete one asynchronous initialization; cancelled or panicked attempts may be retried.
	`OnceCell`	`once-cell`	Store one value from an access-time initializer; failed, cancelled, or panicked attempts may be retried.
	`LazyCell`	`lazy-cell`	Initialize one value with a stored function and resume the same in-flight future after caller cancellation.
	`OnceMap`	`once-map`	Coalesce work per key and retain each successful value until explicitly removed.
	`Group`	`singleflight`	Coalesce overlapping work per key without retaining completed values.
Communication	`Completion`	`completion`	Publish one shared result to any number of current and future observers.
	`oneshot`	`oneshot`	Send one value from one sender to one receiver.
	`mpmc`	`mpmc`	Distribute each value to exactly one of multiple competing receivers.
	`mpsc`	`mpsc`	Send each value from multiple producers to one receiver with bounded backpressure or an unbounded queue.
	`broadcast`	`broadcast`	Deliver every value to receivers active at send time; retain an unbounded backlog until each consumes or drops.
	`watch`	`watch`	Publish cloneable latest state from one or more senders; receivers independently coalesce intermediate updates.
Object reuse	`pool`	`pool`	Reuse objects through bounded or unbounded pool variants.
Sync interop	`FutureExt`	`blocking`	Drive one runtime-agnostic future from a blocking thread.

Synchronous interoperability

The optional blocking module is a boundary adapter for synchronous callers. It parks the calling thread while driving one future; it is not a general-purpose executor.

cargo add asyncband --features blocking,oneshot

use std::thread;

use asyncband::blocking::FutureExt as _;
use asyncband::oneshot;

let (sender, receiver) = oneshot::channel();
thread::spawn(move || {
    let result = 6 * 7;
    sender.send(result).unwrap();
});

assert_eq!(receiver.block_on(), Ok(42));

Async first, blocking by adaptation

Async and synchronous synchronization primitives have different optimization constraints. Once an async operation is exposed as a runtime-agnostic future, synchronous code can usually drive that future through a block_on adapter. Asyncband therefore designs its primitives for async use and provides blocking interoperability at the boundary instead of duplicating synchronous methods across every type.

A sync-first implementation can exploit OS- or platform-specific facilities that an async implementation cannot assume. Libraries focused on synchronous code can therefore make different and sometimes better tradeoffs. Asyncband leaves those optimizations to dedicated libraries rather than treating blocking adaptation as a second family of primitives.

Execution constraints

The blocking module is a lightweight, thread-parking single-future executor, not a general-purpose async runtime. wait_timeout drops the future on timeout. Futures that depend on a runtime-specific timer or I/O driver still need that runtime's driver to make progress, and blocking an executor thread can cause starvation or deadlocks. See the blocking module documentation for the full contract.

Thread safety

Asyncband types implement Send and Sync only when the protected, transferred, or managed value satisfies the necessary bounds. See each API's documentation for its exact contract.

Minimum Supported Rust Version (MSRV)

This crate is built against the latest stable release, and its minimum supported rustc version is 1.86.0.

The policy is that the minimum Rust version required to use this crate can be increased in minor version updates. For example, if Asyncband 1.0 requires Rust 1.20.0, then Asyncband 1.0.z for all values of z will also require Rust 1.20.0 or newer. However, Asyncband 1.y for y > 0 may require a newer minimum version of Rust.

License and Trademarks

This project is licensed under Apache License, Version 2.0; the license file also records incorporated third-party works and their terms.

Apache Asyncband, Asyncband, and Apache are either registered trademarks or trademarks of The Apache Software Foundation in the United States and/or other countries.

This crate provides concurrency control and async coordination primitives that are runtime agnostic.