Micro HTTP

A lightweight, efficient, and modular HTTP/1.1 server implementation built on top of tokio.

Features

• Full HTTP/1.1 protocol support
• Asynchronous I/O using tokio
• Streaming request and response bodies
• Chunked transfer encoding
• Keep-alive connections
• Expect-continue mechanism
• Efficient memory usage through zero-copy parsing
• Clean error handling
• Structured logging with tracing

Quick Start

Add this to your Cargo.toml:

[dependencies]
micro-http = "0.1"
tokio = { version = "1", features = ["rt-multi-thread", "net", "io-util", "macros", "sync", "signal", "test-util"] }
http = "1"
http-body = "1"
tracing = "0.1"
tracing-subscriber = "0.3"

Example

Here's a simple HTTP server that responds with "Hello World!":

use http::{Request, Response, StatusCode};
use http_body_util::BodyExt;
use std::error::Error;
use std::sync::Arc;

use tokio::net::TcpListener;

use tracing::{error, info, warn, Level};
use tracing_subscriber::FmtSubscriber;
use micro_http::connection::HttpConnection;
use micro_http::handler::make_handler;
use micro_http::protocol::body::ReqBody;

#[tokio::main]
async fn main() {
    // Initialize logging
    let subscriber = FmtSubscriber::builder()
        .with_max_level(Level::INFO)
        .finish();
    tracing::subscriber::set_global_default(subscriber)
        .expect("setting default subscriber failed");

    info!(port = 8080, "start listening");
    let tcp_listener = match TcpListener::bind("127.0.0.1:8080").await {
        Ok(tcp_listener) => tcp_listener,
        Err(e) => {
            error!(cause = %e, "bind server error");
            return;
        }
    };

    let handler = make_handler(simple_handler);
    let handler = Arc::new(handler);
    loop {
        let (tcp_stream, _remote_addr) = match tcp_listener.accept().await {
            Ok(stream_and_addr) => stream_and_addr,
            Err(e) => {
                warn!(cause = %e, "failed to accept");
                continue;
            }
        };

        let handler = handler.clone();

        tokio::spawn(async move {
            let (reader, writer) = tcp_stream.into_split();
            let connection = HttpConnection::new(reader, writer);
            match connection.process(handler).await {
                Ok(_) => {
                    info!("finished process, connection shutdown");
                }
                Err(e) => {
                    error!("service has error, cause {}, connection shutdown", e);
                }
            }
        });
    }
}

async fn simple_handler(request: Request<ReqBody>) -> Result<Response<String>, Box<dyn Error + Send + Sync>> {
    let path = request.uri().path().to_string();
    info!("request path {}", path);

    let (_header, body) = request.into_parts();

    let body_bytes = body.collect().await?.to_bytes();
    info!(body = std::str::from_utf8(&body_bytes[..]).unwrap(), "receiving request body");

    let response_body = "Hello World!\r\n";
    let response = Response::builder()
        .status(StatusCode::OK)
        .header(http::header::CONTENT_LENGTH, response_body.len())
        .body(response_body.to_string())
        .unwrap();

    Ok(response)
}

Architecture

The crate is organized into several key modules:

• connection: Core connection handling and lifecycle management
• protocol: Protocol types and abstractions
• codec: Protocol encoding/decoding implementation
• handler: Request handler traits and utilities

Performance Considerations

The implementation focuses on performance through:

• Zero-copy parsing where possible
• Efficient buffer management
• Streaming processing of bodies
• Concurrent request/response handling
• Connection keep-alive

Limitations

• HTTP/1.1 only (HTTP/2 or HTTP/3 not supported)
• No TLS support (use a reverse proxy for HTTPS)
• Maximum header size: 8KB
• Maximum number of headers: 64

Safety

The crate uses unsafe code in a few well-documented places for performance optimization, particularly in header parsing. All unsafe usage is carefully reviewed and tested.

License

This project is licensed under the MIT License or Apache-2.0 License, pick one.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.