pypac4http2

Python package which builds on pypac to add pac file support for httplib2 Http

This package adds proxy auto configure pac support to the httplib2 library Http classby sub-classing the class and adding a new constructor parameter for a resolver class.

Installation

pip install pypac4http2

Usage

HttpPac Class

The HttpPac class extends httplib2.Http and automatically resolves the proxy for each request using a PAC file.

from pypac4http2 import HttpPac

# Automatically discover PAC using OS settings or WPAD
http = HttpPac()
response, content = http.request("http://example.org")

# Specify a PAC URL
http = HttpPac(pac_url="http://internal.corp/proxy.pac")
response, content = http.request("http://example.org")

# Provide PAC as a JavaScript string
pac_js = "function FindProxyForURL(url, host) { return 'PROXY proxy:8080'; }"
http = HttpPac(pac_js=pac_js)
response, content = http.request("http://example.org")

# The library also supports the PAC_URL environment variable:
# export PAC_URL=http://internal.corp/proxy.pac

Usage with Google API Client

HttpPac is designed to be a drop-in replacement for httplib2.Http, making it ideal for use with the Google API Python Client.

In a typical enterprise environment, you may also need your authentication flow (which often uses google-auth and requests) to respect the same PAC settings. You can use pypac.PACSession alongside HttpPac.

import os
from googleapiclient.discovery import build
import google.auth
from google.auth.transport.requests import Request
from pypac import PACSession, get_pac
from pypac4http2 import HttpPac

# 1. Use PACSession for authentication flows that use 'requests'
# pypac.PACSession doesn't natively check PAC_URL, so we get the PAC file manually
pac_url = os.environ.get("PAC_URL")
pac = get_pac(url=pac_url) if pac_url else get_pac()
auth_session = PACSession(pac=pac)
auth_request = Request(session=auth_session)

# 2. Get default credentials using the proxy-aware request
credentials, project = google.auth.default(request=auth_request)

# 3. Refresh credentials using the proxy-aware session
credentials.refresh(auth_request)

# 4. Use HttpPac for the Google API service calls (which use 'httplib2')
# Both will respect PAC settings (e.g. via PAC_URL environment variable)
http = HttpPac()
service = build('drive', 'v3', http=http, credentials=credentials)

# All API calls will now automatically resolve proxies via PAC
files = service.files().list().execute()

CLI Tool

The package includes a CLI tool pypac4http2 to resolve proxies for a given URL.

# Resolve proxy using OS auto-discovery (or PAC_URL env var if set)
pypac4http2 https://google.com

# Resolve proxy using a specific PAC URL
pypac4http2 --pac-url http://example.com/proxy.pac https://google.com

# Resolve proxy using a PAC JavaScript string
pypac4http2 --pac-js "function FindProxyForURL(url, host) { return 'DIRECT'; }" https://google.com

Output Example:

Proxy choice: PROXY proxy.example.com:8080
{
  "target_url": "https://google.com",
  "pac_result": "PROXY proxy.example.com:8080",
  "proxy_info": {
    "proxy_type": "HTTP",
    "proxy_host": "proxy.example.com",
    "proxy_port": 8080
  }
}