Skip to content

aiohttpvspycurl

Apache 2 499 30 14,508
105.3 million (month) Jul 26 2019 3.9.3(2 months ago)
1,041 9 9 LGPL/MIT
7.45.3(a month ago) Feb 25 2003 1.2 million (month)

aiohttp is an asynchronous HTTP client/server framework for asyncio and Python. It provides a simple API for making HTTP requests and handling both client and server functionality. Like the requests package, aiohttp is designed to be easy to use and handle many of the low-level details of working with HTTP.

The main benefit of aiohttp over requests is that it is built on top of the asyncio library, which means that it can handle many requests at the same time without blocking the execution of your program. This can lead to significant performance improvements when making many small requests, or when dealing with slow or unreliable network connections.

aiohttp provides both client and server side functionality, so you can use it to create web servers and handle client requests in a non-blocking manner. It also supports WebSocket protocol, so it can be used for building real-time application like chat, game, etc.

aiohttp also provide several features for handling connection errors, managing timeouts, and client sessions. It also provide similar features like requests package like redirect handling, cookies, and support for several authentication modules.

You can install aiohttp via pip package manager:

pip install aiohttp

In terms of API design, aiohttp is similar to requests and thus should be familiar to anyone who has used the requests library, but it provides an async with block to manage the context of the connection and used await statement to wait for the result.

It''s worth noting that aiohttp is built on top of asyncio and is designed to be used in Python 3.5 and above. It provides the same functionality as httpx but it is specifically built for the asyncio framework.

PycURL is a Python interface to libcurl, a multi-protocol file transfer library written in C. PycURL allows developers to use a variety of network protocols in their Python programs, including HTTP, FTP, SMTP, POP3, and many more.

PycURL is often used in web scraping, data analysis, and automation tasks, as it allows developers to send and receive data over the internet. It can be used to perform various types of requests, such as GET, POST, PUT, and DELETE, and can also handle file uploads and downloads, cookies, and redirects.

One of the key features of PycURL is its support for SSL and proxy servers, which allows developers to securely transfer data over the internet and work around any network restrictions. PycURL also supports a wide range of authentication methods, such as Basic, Digest, and NTLM, and allows developers to easily set custom headers and query parameters.

Just like cURL itself, PycURL is also highly configurable and allows for fine-grained control over various aspects of the transfer, such as timeouts, retries, buffer sizes, and verbosity levels. Additionally, PycURL also provides easy access to the underlying libcurl library, which allows developers to access advanced functionality that is not exposed by the PycURL API.

It's important to note that PycURL is a wrapper around the libcurl library and therefore provides the same functionality and performance as libcurl.

Main strengths of PycURL is that it uses cURL which is one of the most feature rich low-level http clients. The downside is that it's a very low-level client (see the examples below) with complex API making use in web scraping very difficult and niche.

Highlights


asynciowebsocketshttp2http-servermulti-partresponse-streaminghttp-proxy
uses-curlhttp2multi-partresponse-streaminghttp-proxy

Example Use


import asyncio
from aiohttp import ClientSession, WSMsgType

# aiohttp only provides async client so we must use a coroutine:
async def run():
    async with ClientSession(headers={"User-Agent": "webscraping.fyi"}) as session:
        # we can use the session to make requests:
        response = await session.get("http://httpbin.org/headers")
        print(response.status)
        # note: to read the response body we must use await:
        print(await response.text())

        # aiohttp also comes with convenience methods for common requests:
        # POST json
        resp = await session.post("http://httpbin.org/post", json={"key": "value"})
        # POST form data
        resp = await session.post("http://httpbin.org/post", data={"key": "value"})
        # decode response as json
        resp = await session.get("http://httpbin.org/json")
        data = await resp.json()
        print(data)

        # aiohttp also supports websocket connections
        # which can be used to scrape websites that use websockets:
        async with session.ws_connect("http://example.org/ws") as ws:
            async for msg in ws:
                if msg.type == WSMsgType.TEXT:
                    if msg.data == "close cmd":
                        await ws.close()
                        break
                    else:
                        await ws.send_str(msg.data + "/answer")
                elif msg.type == WSMsgType.ERROR:
                    break


asyncio.run(run())
import pycurl
from io import BytesIO

buf = BytesIO()
headers = BytesIO()

c = pycurl.Curl()
c.setopt(c.HTTP_VERSION, c.CURL_HTTP_VERSION_2_0)  # set to use http2
# set proxy
c.setopt(c.PROXY, 'http://proxy.example.com:8080') 
c.setopt(c.PROXYUSERNAME, 'username')
c.setopt(c.PROXYPASSWORD, 'password')

# make a request
c.setopt(c.URL, 'https://httpbin.org/get')
c.setopt(c.WRITEFUNCTION, buf.write)  # where to save response body
c.setopt(c.HEADERFUNCTION, headers.write)  # where to save response headers
# to make post request enable POST option:
# c.setopt(c.POST, 1)
# c.setopt(c.POSTFIELDS, 'key1=value1&key2=value2')
c.perform()  # send request

# read response
data = buf.getvalue().decode()
headers = headers.getvalue().decode()  # headers as a string
headers = dict([h.split(': ') for h in headers.splitlines() if ': ' in h])  # headers as a dict
c.close()

# multiple concurrent requests can be made using CurlMulti object:
# Create a CurlMulti object
multi = pycurl.CurlMulti()
# Set the number of maximum connections
multi.setopt(pycurl.MAXCONNECTS, 5)

# Create a list to store the Curl objects
curls = []

# Add the first request
c1 = pycurl.Curl()
c1.setopt(c1.URL, 'https://httpbin.org/get')
c1.setopt(c1.WRITEFUNCTION, BytesIO().write)
multi.add_handle(c1)
curls.append(c1)

# Add the second request
c2 = pycurl.Curl()
c2.setopt(c2.URL, 'https://httpbin.org/')
c2.setopt(c2.WRITEFUNCTION, BytesIO().write)
multi.add_handle(c2)
curls.append(c2)

# Start the requests
while True:
    ret, _ = multi.perform()
    if ret != pycurl.E_CALL_MULTI_PERFORM:
        break

# Close the connections
for c in curls:
    multi.remove_handle(c)
    c.close()

Alternatives / Similar