如何通过SOCKS代理与python / http.client连接？

Question

通常，当使用httplib (python2.x)或http.client (python3.x)时，您将设置一个连接并发出如下请求：

from http.client import HTTPConnection
from urllib.parse import urlparse, urlencode

url = urlparse("http://final.destination.example.com:8888/")
conn = HTTPConnection(url.netloc)
request_path = "%s?%s" % (url.path, url.query)
conn.request("POST", request_path, post_data)

(注意这个示例代码有python3.x导入)

出于性能原因，我希望使用http.client，但我也希望使用SOCKS代理来远程访问集群。

有没有办法在httplib/http.client中使用SOCKS代理？

Answer 1

我在这里回答我自己的问题，希望下一个在这种情况下的人能够找到解决办法。

第一次安装PySocks时：

pip install PySocks

然后，您需要在实例化您的HTTPConnection并通知它它将使用代理之后手动设置SOCKS代理：

from http.client import HTTPConnection
from urllib.parse import urlparse, urlencode

import socks

url = urlparse("http://final.destination.example.com:8888/")

conn = HTTPConnection('127.0.0.1', 9000) # use socks proxy address
conn.set_tunnel(url.netloc, url.port) # remote host and port that you actually want to talk to
conn.sock = socks.socksocket() # manually set socket
conn.sock.set_proxy(socks.PROXY_TYPE_SOCKS5, "127.0.0.1", 9000) # use socks proxy address
conn.sock.connect((url.netloc, url.port)) # remote host and port that you actually want to talk to

request_path = "%s?%s" % (url.path, url.query)
conn.request("POST", request_path, post_data)