no-proxy

Standalone no_proxy/NO_PROXY parser and evaluator.

Usage

from no_proxy import bypass # More specific functions and types available
from urllib.parse import urlparse

bypass(no_proxy="*", host="example.com") # True
bypass("127.0.0.0/8,::1", "127.0.0.1") # True

my_host = urlparse("//example.com:5000/path?x=y").hostname # example.com
bypass("www.example.com,example.org", my_host) # False

Specification

A no-proxy string is either * or a comma-separated list of no-proxy entries. An IP-address or hostname matches a no-proxy string if the string is * or if any of its entries matches.

A no-proxy entry is either

• an IP-address, which matches itself,
• an IP-range, which matches included IP-addresses,
• a domain with leading ., which matches all of its subdomains, or
• a hostname, which matches itself.

Hostnames are not validated and there is no special handling of whitespace, trailing dots, ports, etc.