Skip to main content
Join the official 2019 Python Developers SurveyStart the survey!

Some function and classes to help you deal with aiohttp sessions

Project description

Travis status

Automatically add session management to a class

Some function and classes to help you deal with aiohttp client sessions. This is made after this discussion. This works for decorating both coroutines and asynchronous generators methods.

Installation

pip install aiohttp-asynctools

TL;DR

Add an aiohttp.ClientSession object to your class in a fast and clean way with the following 4 steps:

  1. Import asynctools
  2. Extend asynctools.AbstractSessionContainer
  3. (optional) Cutomize the session instanciation in your __init__ method.
  4. Decorate asynchronous methods/generators with @asynctools.attach_session to attach a session argument.

Here is what it looks like for a simple example using a math API (http://api.mathjs.org/v4):

import asyncio
import asynctools  # 1.

class MathRequests(asynctools.AbstractSessionContainer):  # 2.
    def __init__(self):
        super().__init__(raise_for_status=True)  # 3. (optional)

    @asynctools.attach_session  # 4.
    async def get_text(self, url, params, session=None):
        async with session.get(url, params=params) as response:
            return await response.text()
    async def get_square(self, value):
        return await self.get_text("http://api.mathjs.org/v4", params={'expr' : '{}^2'.format(value)})

Note that the __init__ method has to be here, if it is empty, just use the pass keyword as its content.

You are now ready to instantiate a MathRequests context manager and start requesting the math service using a single aiohttp session (the session is hidden from the MathRequests user):

from aiohttp import web
routes = web.RouteTableDef()

@routes.get('/squares')
async def index(request):
    tasks = []
    data = request.query
    values = data['values'].split(',')
    # Use the object as a context manager (async with <context_manager> as <name>)
    async with MathRequests() as maths:
        results = await asyncio.gather(*[ maths.get_square(v) for v in values ])
    return web.json_response({ 'values':values, 'results':results })

maths_app = web.Application()
maths_app.add_routes(routes)

Alternatively you can start/close sessions explicitly:

# ...
    maths.start_session()
    results = await asyncio.gather(*[ maths.get_square(v) for v in values ])
    maths.close_session()
# ...

You can now start your server and test the service (example here with gunicorn):

gunicorn math_requests:maths_app --bind localhost:8085 --worker-class aiohttp.GunicornWebWorker --reload
curl 'http://localhost:8085/squares?values=1,2,3,4,5,6,7,8,9,10'

Which should output:

{
  "values": ["1", "2", "3", "4", "5", "6", "7", "8", "9", "10"],
  "results": ["1", "4", "9", "16", "25", "36", "49", "64", "81", "100"]
}

Details and explanation

What?

The goal is to help aiohttp users to build classes that will contain sessions object in an efficient/clean way.

Why?

If you want to build class that will make requests using aiohttp client, at some point you'll have to deal with sessions. The quickstart guide for aiohttp client has an important note about sessions.

import aiohttp

async with aiohttp.ClientSession() as session:
   async with session.get('http://httpbin.org/get') as resp:
       print(resp.status)
       print(await resp.text())

...

Note

Don’t create a session per request. Most likely you need a session per application which performs all requests altogether.

More complex cases may require a session per site, e.g. one for Github and other one for Facebook APIs. Anyway making a session for every request is a very bad idea.

A session contains a connection pool inside. Connection reusage and keep-alives (both are on by default) may speed up total performance.

The goal is to have a single session attached to a given object, this is where this module will give you a simple way of achieving this with only 4 (very simple) lines of code.

How?

The module provides an abstract class AbstractSessionContainer and a method decorator attach_session that you'll have to use to automatically add session management to an existing class.

Say you have a class MathRequests that has a single method get_square that returns the square value of the given parameter using an aiohttp.get request to the math API service located at http://api.mathjs.org/v4. Here is what your class look like for now:

import asyncio, aiohttp
routes = aiohttp.web.RouteTableDef()

class MathRequests:
    async def get_text(self, url, params):
        # Remember "making a session for every request is a very bad idea"
        async with aiohttp.ClientSession() as session:
            async with session.get(url, params=params) as response:
                return await response.text()
    async def get_square(self, value):
        return await self.get_text("http://api.mathjs.org/v4", params={'expr' : '{}^2'.format(value)})

@routes.get('/squares')
async def index(request):
    tasks = []
    data = request.query
    values = data['values'].split(',')
    maths = MathRequests()
    results = await asyncio.gather(*[ maths.get_square(v) for v in values ])
    return web.json_response({ 'values':values, 'results':results })

maths_app = aiohttp.web.Application()
maths_app.add_routes(routes)

As aiohttp documentation says, this is a bad idea to implement MathRequests this way, we need to share a single session for all get_square requests.

A simple solution to this would be to store a client session object within MathRequests, which you could initiate in the __init__ method. Saddly this is not a very clean solution as aiohttp sessions should be instantiated in a synchronous way (outside the even loop). See aiohttp#1468 for more information about creation a session outside of coroutine.

Here is the final solution using the provided module asynctools:

import asyncio
import asynctools # 1) Import

# 2) Extends the abstract class that will handle the aiohttp session for you:
class MathRequests(asynctools.AbstractSessionContainer):
    def __init__(self):
        # 2') (optional) initilise with any 'aiohttp.ClientSession' argument
        super().__init__(raise_for_status=True)
    # 3) This decorator will automatically fill the session argument:
    @asynctools.attach_session
    async def get_text(self, url, params, session=None):  # 4) Add the 'session' argument
        async with session.get(url, params=params) as response:
            return await response.text()
    async def get_square(self, value):
        return await self.get_text("http://api.mathjs.org/v4", params={'expr' : '{}^2'.format(value)})

from aiohttp import web
routes = web.RouteTableDef()

@routes.get('/squares')
async def index(request):
    tasks = []
    data = request.query
    values = data['values'].split(',')
    async with MathRequests() as maths: # Use the object as a context manager (async with <context_manager> as <name>)
        results = await asyncio.gather(*[ maths.get_square(v) for v in values ])
    return web.json_response({ 'values':values, 'results':results })

maths_app = web.Application()
maths_app.add_routes(routes)

Using the MathRequests as a context manager is the best option (as it will make sure the session is correctly started and closed), but it's not the only option, you can also keep your code as it was before:

@routes.get('/squares')
async def index(request):
    tasks = []
    data = request.query
    values = data['values'].split(',')
    maths = MathRequests()
    results = await asyncio.gather(*[ maths.get_square(v) for v in values ])
    return web.json_response({ 'values':values, 'results':results })

In this case, no session is attached to the maths object and every call to get_square will use a different session (which is as bad as it was with the old version of MathRequests). What you can do to avoid that is to explicitly open a "math session" which will make all get_square calls to use the same session (also, don't forget to close the session when you are done):

@routes.get('/maths')
async def index(request):
    tasks = []
    data = request.query
    values = data['values'].split(',')
    maths = MathRequests()
    maths.start_session()
    results = await asyncio.gather(*[ maths.get_square(v) for v in values ])
    maths.close_session()
    return web.json_response({ 'values':values, 'results':results })

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for aiohttp-asynctools, version 0.1.2
Filename, size File type Python version Upload date Hashes
Filename, size aiohttp_asynctools-0.1.2-py2.py3-none-any.whl (8.1 kB) File type Wheel Python version py2.py3 Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page