Streams¶

Streams are high-level async/await-ready primitives to work with network connections. Streams allow sending and receiving data without using callbacks or low-level protocols and transports.

Here is an example of a TCP echo client written using asyncio streams:

import asyncio

async def tcp_echo_client(message):
    async with asyncio.connect('127.0.0.1', 8888) as stream:
        print(f'Send: {message!r}')
        await stream.write(message.encode())

        data = await stream.read(100)
        print(f'Received: {data.decode()!r}')

asyncio.run(tcp_echo_client('Hello World!'))

See also the Examples section below.

Stream Functions

The following top-level asyncio functions can be used to create and work with streams:

coroutine asyncio.connect(host=None, port=None, *, limit=2**16, ssl=None, family=0, proto=0, flags=0, sock=None, local_addr=None, server_hostname=None, ssl_handshake_timeout=None, happy_eyeballs_delay=None, interleave=None)¶

Connect to TCP socket on host : port address and return a Stream object of mode StreamMode.READWRITE.

limit determines the buffer size limit used by the returned Stream instance. By default the limit is set to 64 KiB.

The rest of the arguments are passed directly to loop.create_connection().

The function can be used with await to get a connected stream:

stream = await asyncio.connect('127.0.0.1', 8888)

The function can also be used as an async context manager:

async with asyncio.connect('127.0.0.1', 8888) as stream:
    ...

New in version 3.8.

coroutine asyncio.open_connection(host=None, port=None, *, loop=None, limit=None, ssl=None, family=0, proto=0, flags=0, sock=None, local_addr=None, server_hostname=None, ssl_handshake_timeout=None)¶

Establish a network connection and return a pair of (reader, writer) objects.

The returned reader and writer objects are instances of StreamReader and StreamWriter classes.

The loop argument is optional and can always be determined automatically when this function is awaited from a coroutine.

limit determines the buffer size limit used by the returned StreamReader instance. By default the limit is set to 64 KiB.

The rest of the arguments are passed directly to loop.create_connection().

New in version 3.7: The ssl_handshake_timeout parameter.

Deprecated since version 3.8, will be removed in version 3.10: open_connection() is deprecated in favor of connect().

coroutine asyncio.start_server(client_connected_cb, host=None, port=None, *, loop=None, limit=2**16, family=socket.AF_UNSPEC, flags=socket.AI_PASSIVE, sock=None, backlog=100, ssl=None, reuse_address=None, reuse_port=None, ssl_handshake_timeout=None, start_serving=True)¶

Start a socket server.

The client_connected_cb callback is called whenever a new client connection is established. It receives a (reader, writer) pair as two arguments, instances of the StreamReader and StreamWriter classes.

client_connected_cb can be a plain callable or a coroutine function; if it is a coroutine function, it will be automatically scheduled as a Task.

The loop argument is optional and can always be determined automatically when this method is awaited from a coroutine.

limit determines the buffer size limit used by the returned StreamReader instance. By default the limit is set to 64 KiB.

The rest of the arguments are passed directly to loop.create_server().

New in version 3.7: The ssl_handshake_timeout and start_serving parameters.

Deprecated since version 3.8, will be removed in version 3.10: start_server() is deprecated if favor of StreamServer

coroutine asyncio.connect_read_pipe(pipe, *, limit=2**16)¶

Takes a file-like object pipe to return a Stream object of the mode StreamMode.READ that has similar API of StreamReader. It can also be used as an async context manager.

limit determines the buffer size limit used by the returned Stream instance. By default the limit is set to 64 KiB.

New in version 3.8.

coroutine asyncio.connect_write_pipe(pipe, *, limit=2**16)¶

Takes a file-like object pipe to return a Stream object of the mode StreamMode.WRITE that has similar API of StreamWriter. It can also be used as an async context manager.

limit determines the buffer size limit used by the returned Stream instance. By default the limit is set to 64 KiB.

New in version 3.8.

Unix Sockets

asyncio.connect_unix(path=None, *, limit=2**16, ssl=None, sock=None, server_hostname=None, ssl_handshake_timeout=None)¶

Establish a Unix socket connection to socket with path address and return an awaitable Stream object of the mode StreamMode.READWRITE that can be used as a reader and a writer.

limit determines the buffer size limit used by the returned Stream instance. By default the limit is set to 64 KiB.

The rest of the arguments are passed directly to loop.create_unix_connection().

The function can be used with await to get a connected stream:

stream = await asyncio.connect_unix('/tmp/example.sock')

The function can also be used as an async context manager:

async with asyncio.connect_unix('/tmp/example.sock') as stream:
    ...

Availability: Unix.

New in version 3.8.

coroutine asyncio.open_unix_connection(path=None, *, loop=None, limit=None, ssl=None, sock=None, server_hostname=None, ssl_handshake_timeout=None)¶

Establish a Unix socket connection and return a pair of (reader, writer).

Similar to open_connection() but operates on Unix sockets.

See also the documentation of loop.create_unix_connection().

Availability: Unix.

New in version 3.7: The ssl_handshake_timeout parameter.

Changed in version 3.7: The path parameter can now be a path-like object

Deprecated since version 3.8, will be removed in version 3.10: open_unix_connection() is deprecated if favor of connect_unix().

coroutine asyncio.start_unix_server(client_connected_cb, path=None, *, loop=None, limit=None, sock=None, backlog=100, ssl=None, ssl_handshake_timeout=None, start_serving=True)¶

Start a Unix socket server.

Similar to start_server() but works with Unix sockets.

See also the documentation of loop.create_unix_server().

Availability: Unix.

New in version 3.7: The ssl_handshake_timeout and start_serving parameters.

Changed in version 3.7: The path parameter can now be a path-like object.

Deprecated since version 3.8, will be removed in version 3.10: start_unix_server() is deprecated in favor of UnixStreamServer.

StreamServer¶

class asyncio.StreamServer(client_connected_cb, /, host=None, port=None, *, limit=2**16, family=socket.AF_UNSPEC, flags=socket.AI_PASSIVE, sock=None, backlog=100, ssl=None, reuse_address=None, reuse_port=None, ssl_handshake_timeout=None, shutdown_timeout=60)¶

The client_connected_cb callback is called whenever a new client connection is established. It receives a Stream object of the mode StreamMode.READWRITE.

client_connected_cb can be a plain callable or a coroutine function; if it is a coroutine function, it will be automatically scheduled as a Task.

limit determines the buffer size limit used by the returned Stream instance. By default the limit is set to 64 KiB.

The rest of the arguments are passed directly to loop.create_server().

coroutine start_serving()¶: Binds to the given host and port to start the server.

coroutine serve_forever()¶

Start accepting connections until the coroutine is cancelled. Cancellation of serve_forever task causes the server to be closed.

This method can be called if the server is already accepting connections. Only one serve_forever task can exist per one Server object.

is_serving()¶: Returns True if the server is bound and currently serving.

bind()¶: Bind the server to the given host and port. This method is automatically called during __aenter__ when StreamServer is used as an async context manager.

is_bound()¶: Return True if the server is bound.

coroutine abort()¶: Closes the connection and cancels all pending tasks.

coroutine close()¶: Closes the connection. This method is automatically called during __aexit__ when StreamServer is used as an async context manager.

sockets¶: Returns a tuple of socket objects the server is bound to.

New in version 3.8.

UnixStreamServer¶

class asyncio.UnixStreamServer(client_connected_cb, /, path=None, *, limit=2**16, sock=None, backlog=100, ssl=None, ssl_handshake_timeout=None, shutdown_timeout=60)¶

The client_connected_cb callback is called whenever a new client connection is established. It receives a Stream object of the mode StreamMode.READWRITE.

client_connected_cb can be a plain callable or a coroutine function; if it is a coroutine function, it will be automatically scheduled as a Task.

limit determines the buffer size limit used by the returned Stream instance. By default the limit is set to 64 KiB.

The rest of the arguments are passed directly to loop.create_unix_server().

coroutine start_serving()¶: Binds to the given host and port to start the server.

is_serving()¶: Returns True if the server is bound and currently serving.

bind()¶: Bind the server to the given host and port. This method is automatically called during __aenter__ when UnixStreamServer is used as an async context manager.

is_bound()¶: Return True if the server is bound.

coroutine abort()¶: Closes the connection and cancels all pending tasks.

coroutine close()¶: Closes the connection. This method is automatically called during __aexit__ when UnixStreamServer is used as an async context manager.

sockets¶: Returns a tuple of socket objects the server is bound to.

Availability: Unix.

New in version 3.8.

Stream¶

class asyncio.Stream¶

Represents a Stream object that provides APIs to read and write data to the IO stream . It includes the API provided by StreamReader and StreamWriter.

Do not instantiate Stream objects directly; use API like connect() and StreamServer instead.

New in version 3.8.

StreamMode¶

class asyncio.StreamMode¶

A subclass of enum.Flag that defines a set of values that can be used to determine the mode of Stream objects.

READ¶

The stream object is readable and provides the API of StreamReader.

WRITE¶

The stream object is writeable and provides the API of StreamWriter.

READWRITE¶

The stream object is readable and writeable and provides the API of both StreamReader and StreamWriter.

New in version 3.8.

StreamReader¶

class asyncio.StreamReader¶

Represents a reader object that provides APIs to read data from the IO stream.

It is not recommended to instantiate StreamReader objects directly; use open_connection() and start_server() instead.

coroutine read(n=-1)¶

Read up to n bytes. If n is not provided, or set to -1, read until EOF and return all read bytes.

If EOF was received and the internal buffer is empty, return an empty bytes object.

coroutine readline()¶

Read one line, where “line” is a sequence of bytes ending with \n.

If EOF is received and \n was not found, the method returns partially read data.

If EOF is received and the internal buffer is empty, return an empty bytes object.

coroutine readexactly(n)¶

Read exactly n bytes.

Raise an IncompleteReadError if EOF is reached before n can be read. Use the IncompleteReadError.partial attribute to get the partially read data.

coroutine readuntil(separator=b'\n')¶

Read data from the stream until separator is found.

On success, the data and separator will be removed from the internal buffer (consumed). Returned data will include the separator at the end.

If the amount of data read exceeds the configured stream limit, a LimitOverrunError exception is raised, and the data is left in the internal buffer and can be read again.

If EOF is reached before the complete separator is found, an IncompleteReadError exception is raised, and the internal buffer is reset. The IncompleteReadError.partial attribute may contain a portion of the separator.

New in version 3.5.2.

at_eof()¶: Return True if the buffer is empty and feed_eof() was called.

StreamWriter¶

class asyncio.StreamWriter¶

Represents a writer object that provides APIs to write data to the IO stream.

It is not recommended to instantiate StreamWriter objects directly; use open_connection() and start_server() instead.

write(data)¶

The method attempts to write the data to the underlying socket immediately. If that fails, the data is queued in an internal write buffer until it can be sent.

Starting with Python 3.8, it is possible to directly await on the write() method:

await stream.write(data)

The await pauses the current coroutine until the data is written to the socket.

Below is an equivalent code that works with Python <= 3.7:

stream.write(data)
await stream.drain()

Changed in version 3.8: Support await stream.write(...) syntax.

writelines(data)¶

The method writes a list (or any iterable) of bytes to the underlying socket immediately. If that fails, the data is queued in an internal write buffer until it can be sent.

Starting with Python 3.8, it is possible to directly await on the write() method:

await stream.writelines(lines)

The await pauses the current coroutine until the data is written to the socket.

Below is an equivalent code that works with Python <= 3.7:

stream.writelines(lines)
await stream.drain()

Changed in version 3.8: Support await stream.writelines() syntax.

close()¶

The method closes the stream and the underlying socket.

Starting with Python 3.8, it is possible to directly await on the close() method:

await stream.close()

The await pauses the current coroutine until the stream and the underlying socket are closed (and SSL shutdown is performed for a secure connection).

Below is an equivalent code that works with Python <= 3.7:

stream.close()
await stream.wait_closed()

Changed in version 3.8: Support await stream.close() syntax.

can_write_eof()¶: Return True if the underlying transport supports the write_eof() method, False otherwise.

write_eof()¶: Close the write end of the stream after the buffered write data is flushed.

transport¶: Return the underlying asyncio transport.

get_extra_info(name, default=None)¶: Access optional transport information; see BaseTransport.get_extra_info() for details.

coroutine drain()¶

Wait until it is appropriate to resume writing to the stream. Example:

writer.write(data)
await writer.drain()

This is a flow control method that interacts with the underlying IO write buffer. When the size of the buffer reaches the high watermark, drain() blocks until the size of the buffer is drained down to the low watermark and writing can be resumed. When there is nothing to wait for, the drain() returns immediately.

is_closing()¶: Return True if the stream is closed or in the process of being closed.

New in version 3.7.

coroutine wait_closed()¶

Wait until the stream is closed.

Should be called after close() to wait until the underlying connection is closed.

New in version 3.7.

Examples¶

TCP echo client using streams¶

TCP echo client using the asyncio.connect() function:

import asyncio

async def tcp_echo_client(message):
    async with asyncio.connect('127.0.0.1', 8888) as stream:
        print(f'Send: {message!r}')
        await stream.write(message.encode())

        data = await stream.read(100)
        print(f'Received: {data.decode()!r}')

asyncio.run(tcp_echo_client('Hello World!'))

TCP echo server using streams¶

TCP echo server using the asyncio.StreamServer class:

import asyncio

async def handle_echo(stream):
    data = await stream.read(100)
    message = data.decode()
    addr = stream.get_extra_info('peername')

    print(f"Received {message!r} from {addr!r}")

    print(f"Send: {message!r}")
    await stream.write(data)

    print("Close the connection")
    await stream.close()

async def main():
    async with asyncio.StreamServer(
            handle_echo, '127.0.0.1', 8888) as server:
        addr = server.sockets[0].getsockname()
        print(f'Serving on {addr}')
        await server.serve_forever()

asyncio.run(main())

Get HTTP headers¶

Simple example querying HTTP headers of the URL passed on the command line:

import asyncio
import urllib.parse
import sys

async def print_http_headers(url):
    url = urllib.parse.urlsplit(url)
    if url.scheme == 'https':
        stream = await asyncio.connect(url.hostname, 443, ssl=True)
    else:
        stream = await asyncio.connect(url.hostname, 80)

    query = (
        f"HEAD {url.path or '/'} HTTP/1.0\r\n"
        f"Host: {url.hostname}\r\n"
        f"\r\n"
    )

    stream.write(query.encode('latin-1'))
    while (line := await stream.readline()):
        line = line.decode('latin1').rstrip()
        if line:
            print(f'HTTP header> {line}')

    # Ignore the body, close the socket
    await stream.close()

url = sys.argv[1]
asyncio.run(print_http_headers(url))

Usage:

python example.py http://example.com/path/page.html

or with HTTPS:

python example.py https://example.com/path/page.html

Register an open socket to wait for data using streams¶

Coroutine waiting until a socket receives data using the asyncio.connect() function:

import asyncio
import socket

async def wait_for_data():
    # Get a reference to the current event loop because
    # we want to access low-level APIs.
    loop = asyncio.get_running_loop()

    # Create a pair of connected sockets.
    rsock, wsock = socket.socketpair()

    # Register the open socket to wait for data.
    async with asyncio.connect(sock=rsock) as stream:
        # Simulate the reception of data from the network
        loop.call_soon(wsock.send, 'abc'.encode())

        # Wait for data
        data = await stream.read(100)

        # Got data, we are done: close the socket
        print("Received:", data.decode())

    # Close the second socket
    wsock.close()

asyncio.run(wait_for_data())

Streams¶

StreamServer¶

UnixStreamServer¶

Stream¶

StreamMode¶

StreamReader¶

StreamWriter¶

Examples¶

TCP echo client using streams¶

TCP echo server using streams¶

Get HTTP headers¶

Register an open socket to wait for data using streams¶

Table of Contents

Previous topic

Next topic

This Page