This work has been overtaken by the httpx
project: https://github.com/encode/httpx
We now recommend using httpx.AsyncClient()
for async/await support with a requests-compatible API.
Note: Use ipython
to try this from the console, since it supports await
.
>>> import httpx
>>> client = httpx.AsyncClient()
>>> r = await client.get('https://www.example.org/')
>>> r.status_code
200
>>> r.text
'<!doctype html>\n<html>\n<head>\n<title>Example Domain</title>...'
Brings support for async
/await
syntax to Python's fabulous requests
library.
- Python 3.6+
$ pip install requests-async
Just use the standard requests API, but use await
for making requests.
Note: Use ipython
to try this from the console, since it supports await
.
import requests_async as requests
response = await requests.get('https://example.org')
print(response.status_code)
print(response.text)
Or use explicit sessions, with an async context manager.
import requests_async as requests
async with requests.Session() as session:
response = await session.get('https://example.org')
print(response.status_code)
print(response.text)
The requests_async
package subclasses requests
, so you're getting all the
standard behavior and API you'd expect.
The iter_content()
and iter_lines()
methods are async iterators.
response = await requests.get('https://example.org', stream=True)
async for chunk in response.iter_content():
...
The method signatures remain the same as the standard requests
API:
iter_content(chunk_size=1, decode_unicode=False)
iter_lines(chunk_size=512, decode_unicode=False, delimiter=None)
The methods will yield text if decode_unicode
is set and the response includes
an encoding. Otherwise the methods will yield bytes.
You can also stream request bodies. To do this you should use an asynchronous generator that yields bytes.
async def stream_body():
...
response = await requests.post('https://example.org', data=stream_body())
In some situations, such as when you're testing a web application, you may not want to make actual outgoing network requests, but would prefer instead to mock out the endpoints.
You can do this using the ASGISession
, which allows you to plug into
any ASGI application, instead of making actual network requests.
import requests_async
# Create a mock service, with Starlette, Responder, Quart, FastAPI, Bocadillo,
# or any other ASGI web framework.
mock_app = ...
if TESTING:
# Issue requests to the mocked application.
requests = requests_async.ASGISession(mock_app)
else:
# Make live network requests.
requests = requests_async.Session()
You can also use ASGISession
as a test client for any ASGI application.
You'll probably want to install pytest
and pytest-asyncio
, or something
equivalent, to allow you to write async
test cases.
from requests_async import ASGISession
from myproject import app
import pytest
@pytest.mark.asyncio
async def test_homepage():
client = ASGISession(app)
response = await client.get("/")
assert response.status_code == 200
- The
httpx
package both sync and async HTTP clients, with a requests-compatible API. - The
aiohttp
package provides an alternative client for making async HTTP requests.