Features

• Asynchronous HTTP/2 client and server

• Multiplexing streams of data with managed flow and priority control

• Optional tickless health check

• More to come

Non-features:

• Request/Response wrappers

• Web server, dispatcher, cookie, etc

• HTTP/2 upgrade

Example

A server saying hello:

# Server request handler
async def on_connected(proto):
    try:
        while True:
            # Receive a request from queue
            stream_id, headers = await proto.recv_request()

            # Send response headers
            await proto.start_response(stream_id, {':status': '200'})

            # Response body starts with "hello"
            await proto.send_data(stream_id, b'hello, ')

            # Read all request body as a name from client side
            name = await proto.read_stream(stream_id, -1)

            # Amend response body with the name
            await proto.send_data(stream_id, name)

            # Send trailers with the length of the name
            await proto.send_trailers(stream_id, {'len': str(len(name))})
    except asyncio.CancelledError:
        # do cleanup here
        pass

# Start server on random port, with maximum concurrent requests of 3
server = await aioh2.start_server(on_connected, port=0, concurrency=3)
port = server.sockets[0].getsockname()[1]

And a client to try out:

# Open client connection
client = await aioh2.open_connection('0.0.0.0', port,
                                     functional_timeout=0.1)

# Optionally wait for an ack of tickless ping - a.k.a. until functional
await asyncio.sleep(0.1)  # simulate client being busy with something else
rtt = await client.wait_functional()
if rtt:
    print('Round-trip time: %.1fms' % (rtt * 1000))

# Start request with headers
stream_id = await client.start_request(
    {':method': 'GET', ':path': '/index.html'})

# Send my name "world" as whole request body
await client.send_data(stream_id, b'world', end_stream=True)

# Receive response headers
headers = await client.recv_response(stream_id)
print('Response headers:', headers)

# Read all response body
resp = await client.read_stream(stream_id, -1)
print('Response body:', resp)

# Read response trailers
trailers = await client.recv_trailers(stream_id)
print('Response trailers:', trailers)

client.close_connection()
await asyncio.sleep(.1)

Above example can be found at examples/core.py.

Credits

A big thanks to the great library hyper-h2 from Cory Benfield.

DecentFoX Studio is a software outsourcing company delivering high-quality web-based products and mobile apps for global customers with agile methodology, focusing on bleeding-edge technologies and fast-developing scalable architectures.

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

0.2.1 (2017-12-03)

• Updated to h2 3.0 (Contributed by Arthur Darcet and Peter Wu)

• Dropped support for Python 3.3

0.1.0 (2016-2-6)

• First release on PyPI.