std::net

Struct TcpListener

1.0.0 · Source

pub struct TcpListener(/* private fields */);

Expand description

A TCP socket server, listening for connections.

After creating a TcpListener by binding it to a socket address, it listens for incoming TCP connections. These can be accepted by calling accept or by iterating over the Incoming iterator returned by incoming.

The socket will be closed when the value is dropped.

The Transmission Control Protocol is specified in IETF RFC 793.

§Examples

use std::net::{TcpListener, TcpStream};

fn handle_client(stream: TcpStream) {
    // ...
}

fn main() -> std::io::Result<()> {
    let listener = TcpListener::bind("127.0.0.1:80")?;

    // accept connections and process them serially
    for stream in listener.incoming() {
        handle_client(stream?);
    }
    Ok(())
}

Implementations§

Source §

impl TcpListener

1.0.0 · Source

pub fn bind<A>(addr: A) -> Result<TcpListener, Error>
where A: ToSocketAddrs,

Creates a new TcpListener which will be bound to the specified address.

The returned listener is ready for accepting connections.

Binding with a port number of 0 will request that the OS assigns a port to this listener. The port allocated can be queried via the TcpListener::local_addr method.

The address type can be any implementor of ToSocketAddrs trait. See its documentation for concrete examples.

If addr yields multiple addresses, bind will be attempted with each of the addresses until one succeeds and returns the listener. If none of the addresses succeed in creating a listener, the error returned from the last attempt (the last address) is returned.

§Examples

Creates a TCP listener bound to 127.0.0.1:80:

use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:80").unwrap();

Creates a TCP listener bound to 127.0.0.1:80. If that fails, create a TCP listener bound to 127.0.0.1:443:

use std::net::{SocketAddr, TcpListener};

let addrs = [
    SocketAddr::from(([127, 0, 0, 1], 80)),
    SocketAddr::from(([127, 0, 0, 1], 443)),
];
let listener = TcpListener::bind(&addrs[..]).unwrap();

Creates a TCP listener bound to a port assigned by the operating system at 127.0.0.1.

use std::net::TcpListener;

let socket = TcpListener::bind("127.0.0.1:0").unwrap();

1.0.0 · Source

pub fn local_addr(&self) -> Result<SocketAddr, Error>

Returns the local socket address of this listener.

§Examples

use std::net::{Ipv4Addr, SocketAddr, SocketAddrV4, TcpListener};

let listener = TcpListener::bind("127.0.0.1:8080").unwrap();
assert_eq!(listener.local_addr().unwrap(),
           SocketAddr::V4(SocketAddrV4::new(Ipv4Addr::new(127, 0, 0, 1), 8080)));

1.0.0 · Source

pub fn try_clone(&self) -> Result<TcpListener, Error>

Creates a new independently owned handle to the underlying socket.

The returned TcpListener is a reference to the same socket that this object references. Both handles can be used to accept incoming connections and options set on one listener will affect the other.

§Examples

use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:8080").unwrap();
let listener_clone = listener.try_clone().unwrap();

1.0.0 · Source

pub fn accept(&self) -> Result<(TcpStream, SocketAddr), Error>

Accept a new incoming connection from this listener.

This function will block the calling thread until a new TCP connection is established. When established, the corresponding TcpStream and the remote peer’s address will be returned.

§Errors

Some errors this function returns do not indicate a problem with the listener itself, and a program serving a long-lived listener will usually want to handle them and keep accepting connections rather than treat them as fatal. These include, but are not limited to:

An error specific to a single incoming connection that failed before it could be accepted, such as one aborted by the peer (ConnectionAborted). A later call may succeed immediately.
An error from reaching the per-process or system-wide open file descriptor limit. The call can be retried once other file descriptors have been closed, typically after a short delay.
An error from failing to allocate memory while accepting a connection (OutOfMemory).

Which errors can occur is platform-specific. On Unix, Interrupted errors are retried internally rather than being returned.

§Examples

use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:8080").unwrap();
match listener.accept() {
    Ok((_socket, addr)) => println!("new client: {addr:?}"),
    Err(e) => println!("couldn't get client: {e:?}"),
}

1.0.0 · Source

pub fn incoming(&self) -> Incoming<'_> ⓘ

Returns an iterator over the connections being received on this listener.

The returned iterator will never return None and will also not yield the peer’s SocketAddr structure. Iterating over it is equivalent to calling TcpListener::accept in a loop.

§Errors

Each connection yielded by the iterator can fail for the same reasons as TcpListener::accept; see its documentation for details.

§Examples

use std::net::{TcpListener, TcpStream};

fn handle_connection(stream: TcpStream) {
   //...
}

fn main() -> std::io::Result<()> {
    let listener = TcpListener::bind("127.0.0.1:80")?;

    for stream in listener.incoming() {
        match stream {
            Ok(stream) => {
                handle_connection(stream);
            }
            Err(e) => { /* connection failed */ }
        }
    }
    Ok(())
}

Source

pub fn into_incoming(self) -> IntoIncoming ⓘ

🔬This is a nightly-only experimental API. (tcplistener_into_incoming)

Turn this into an iterator over the connections being received on this listener.

The returned iterator will never return None and will also not yield the peer’s SocketAddr structure. Iterating over it is equivalent to calling TcpListener::accept in a loop.

§Errors

Each connection yielded by the iterator can fail for the same reasons as TcpListener::accept; see its documentation for details.

§Examples

#![feature(tcplistener_into_incoming)]
use std::net::{TcpListener, TcpStream};

fn listen_on(port: u16) -> impl Iterator<Item = TcpStream> {
    let listener = TcpListener::bind(("127.0.0.1", port)).unwrap();
    listener.into_incoming()
        .filter_map(Result::ok) /* Ignore failed connections */
}

fn main() -> std::io::Result<()> {
    for stream in listen_on(80) {
        /* handle the connection here */
    }
    Ok(())
}

1.9.0 · Source

pub fn set_ttl(&self, ttl: u32) -> Result<(), Error>

Sets the value for the IP_TTL option on this socket.

This value sets the time-to-live field that is used in every packet sent from this socket.

§Examples

use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:80").unwrap();
listener.set_ttl(100).expect("could not set TTL");

1.9.0 · Source

pub fn ttl(&self) -> Result<u32, Error>

Gets the value of the IP_TTL option for this socket.

For more information about this option, see TcpListener::set_ttl.

§Examples

use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:80").unwrap();
listener.set_ttl(100).expect("could not set TTL");
assert_eq!(listener.ttl().unwrap_or(0), 100);

1.9.0 · Source

pub fn set_only_v6(&self, only_v6: bool) -> Result<(), Error>

👎Deprecated since 1.16.0:

this option can only be set before the socket is bound

1.9.0 · Source

pub fn only_v6(&self) -> Result<bool, Error>

👎Deprecated since 1.16.0:

this option can only be set before the socket is bound

1.9.0 · Source

pub fn take_error(&self) -> Result<Option<Error>, Error>

Gets the value of the SO_ERROR option on this socket.

This will retrieve the stored error in the underlying socket, clearing the field in the process. This can be useful for checking errors between calls.

§Examples

use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:80").unwrap();
listener.take_error().expect("No error was expected");

1.9.0 · Source

pub fn set_nonblocking(&self, nonblocking: bool) -> Result<(), Error>

Moves this TCP stream into or out of nonblocking mode.

This will result in the accept operation becoming nonblocking, i.e., immediately returning from their calls. If the IO operation is successful, Ok is returned and no further action is required. If the IO operation could not be completed and needs to be retried, an error with kind io::ErrorKind::WouldBlock is returned.

On most Unix platforms, calling this method corresponds to calling ioctl FIONBIO. On Windows, calling this method corresponds to calling ioctlsocket FIONBIO.

§Examples

Bind a TCP listener to an address, listen for connections, and read bytes in nonblocking mode:

use std::io;
use std::net::TcpListener;

let listener = TcpListener::bind("127.0.0.1:7878").unwrap();
listener.set_nonblocking(true).expect("Cannot set non-blocking");

for stream in listener.incoming() {
    match stream {
        Ok(s) => {
            // do something with the TcpStream
            handle_connection(s);
        }
        Err(ref e) if e.kind() == io::ErrorKind::WouldBlock => {
            // wait until network socket is ready, typically implemented
            // via platform-specific APIs such as epoll or IOCP
            wait_for_fd();
            continue;
        }
        Err(e) => panic!("encountered IO error: {e}"),
    }
}