def pac_context_for_url(url="https://www.actlab.org/", proxy_auth=None):
"""
This context manager provides a simple way to add rudimentary PAC functionality
to code that cannot be modified to use :class:`PACSession`,
but obeys the ``HTTP_PROXY`` and ``HTTPS_PROXY`` environment variables.
Upon entering this context, PAC discovery occurs with default parameters.
If a PAC is found, then it's asked for the proxy to use for the given URL.
The proxy environment variables are then set accordingly.
Note that this provides a very simplified PAC experience that's insufficient for some scenarios.
:param url: Consult the PAC for the proxy to use for this URL. Default is "https://www.actlab.org/"
:param requests.auth.HTTPProxyAuth proxy_auth: Username and password proxy authentication.
"""
prev_http_proxy, prev_https_proxy = os.environ.get(
'HTTP_PROXY'), os.environ.get('HTTPS_PROXY')
pac = pypac.get_pac()
if pac:
resolver = pypac.ProxyResolver(pac, proxy_auth=proxy_auth)
proxies = resolver.get_proxy_for_requests(url)
# Cannot set None for environ. (#27)
os.environ['HTTP_PROXY'] = proxies.get('http') or ''
os.environ['HTTPS_PROXY'] = proxies.get('https') or ''
yield
if prev_http_proxy:
os.environ['HTTP_PROXY'] = prev_http_proxy
elif 'HTTP_PROXY' in os.environ:
del os.environ['HTTP_PROXY']
if prev_https_proxy:
os.environ['HTTPS_PROXY'] = prev_https_proxy
elif 'HTTPS_PROXY' in os.environ:
del os.environ['HTTPS_PROXY']
def __init__(self, pac_url: str = None, js: str = None, **kwargs: Any) -> None:
super().__init__(**kwargs)
if pac_url and "://" not in pac_url:
pac_url = "http://" + pac_url
self.pac_url = pac_url
self._js = js
self._pac_file = get_pac(
url=pac_url,
js=js,
allowed_content_types=[
"application/octet-stream",
"application/x-ns-proxy-autoconfig",
"application/x-javascript-config",
],
)
self._resolver = ProxyResolver(self._pac_file)
def __init__(self,
pac_url: str = None,
js: str = None,
**kwargs: Any) -> None:
super().__init__(**kwargs)
if pac_url and "://" not in pac_url:
pac_url = "http://" + pac_url
self.pac_url = pac_url
self._pac_file = get_pac(
url=pac_url,
js=js,
allowed_content_types=[
"application/octet-stream",
"application/x-ns-proxy-autoconfig",
"application/x-javascript-config",
],
)
self._resolver = ProxyResolver(self._pac_file)
def __init__(self, *, url: str = "", pac_url: str = "") -> None:
args: Dict[str, Any] = {}
if pac_url:
# Load the PAC file as PyPAC won't do it for us
if pac_url.startswith("file:"):
with open(pac_url.replace("file://", "")) as pac:
args["js"] = pac.read()
else:
args["url"] = pac_url
args["allowed_content_types"] = [
"application/octet-stream",
"application/x-ns-proxy-autoconfig",
"application/x-javascript-config",
]
self.pac_url = pac_url
self._pac_file = get_pac(**args)
self._resolver = ProxyResolver(self._pac_file)
def load_proxy():
# Return if proxies specified in Px config
# Check if need to refresh
if (State.proxy_mode == MODE_CONFIG or
(State.proxy_refresh is not None and time.time() - State.proxy_refresh
< State.config.getint("settings", "proxyreload"))):
dprint("Skip proxy refresh")
return
# Reset proxy server list
State.proxy_mode = MODE_NONE
State.proxy_server = []
# First use pypac for WPAD or PAC config
if "pypac" in sys.modules:
State.pac = pypac.get_pac()
# Fallback for file:/// URLs
if State.pac is None:
acu = pypac.windows.autoconfig_url_from_registry()
if acu:
acu = acu.replace("file:///", "")
if ":" not in acu:
acu = "c:\\" + acu
if os.path.isfile(acu):
dprint("Loading " + acu)
with open(acu, "r") as f:
State.pac = pypac.parser.PACFile(f.read())
# Fall back to any manual proxies defined in Internet Options
if State.pac is None:
parse_proxy(",".join([
urlparse.urlparse(i).netloc
for i in set(urlrequest.getproxies().values())
]))
if State.proxy_server:
State.proxy_mode = MODE_MANUAL
else:
State.proxy_mode = MODE_PAC
dprint("Proxy mode = " + str(State.proxy_mode))
State.proxy_refresh = time.time()
def _get_proxy_from_pac(url: str, hostname: str) -> Optional[str]:
"""
Returns:
proxy url to use
"" if no proxy should be used
`None` if no config was found
"""
# Hack to disable check on content-type (given server might not be well
# configured, and `get_pac` silently fails on wrong content-type by returning None)
# Also disable WPAD protocol to retrieve PAC from DNS given it produce annoying
# WARNING logs
get_pac_kwargs: dict = {"from_dns": False, "allowed_content_types": {""}}
force_proxy_pac_url = os.environ.get("http_proxy_pac") or os.environ.get(
"HTTP_PROXY_PAC")
if force_proxy_pac_url:
get_pac_kwargs["url"] = force_proxy_pac_url
logger.debug(
"Retrieving .PAC proxy config url from `HTTP_PROXY_PAC` env var",
pac_url=force_proxy_pac_url,
)
else:
logger.debug("Retrieving .PAC proxy config url from system")
try:
pacfile = pypac.get_pac(**get_pac_kwargs)
except Exception as exc:
logger.warning("Error while retrieving .PAC proxy config",
exc_info=exc)
return None
if not pacfile:
return None
try:
proxies = pacfile.find_proxy_for_url(url, hostname)
logger.debug("Found proxies info in .PAC proxy config",
target_url=url,
proxies=proxies)
proxies = [p.strip() for p in proxies.split(";")]
if len(proxies) > 1:
logger.warning(
"Found multiple proxies info in .PAC proxy config for target, only the first one is going to be used !",
target_url=url,
proxies=proxies,
)
# We don't handle multiple proxies so keep the first correct one and pray !
for proxy in proxies:
if proxy in ("DIRECT", ""):
# PAC explicitly told us not to use a proxy
return ""
elif proxy:
# Should be of style `PROXY 8.8.8.8:9999`
proxy_type, proxy_netloc = proxy.strip().split()
proxy_type = proxy_type.upper()
if proxy_type in ("PROXY", "HTTP"):
return f"http://{proxy_netloc}"
elif proxy_type == "HTTPS":
return f"https://{proxy_netloc}"
else:
logger.warning(
"Unsupported proxy type requested by .PAC proxy config",
proxy=proxy,
target_url=url,
)
else:
return None
except Exception as exc:
logger.warning("Error while using .PAC proxy config",
target_url=url,
exc_info=exc)
return None
def target_session(proxy):
pac = get_pac(url=proxy, allowed_content_types=['text/plain'])
return PACSession(pac)
import os
import pypac
# PyPAC découvrira automatiquement les paramètres du fichier PAC
pac = pypac.get_pac(url='http://127.0.0.1:9000/localproxy.pac')
# On récupère le proxy associé à Cloudant (idem, on aurait pu utiliser n'importe quelle autre URL)
proxies = pac.find_proxy_for_url('https://rest.apisandbox.zuora.com',
'zuora.com')
#proxies = pac.find_proxy_for_url('https://www.google.com', 'google.com')
# Ce qui retourne quelque chose comme : 'PROXY 4.5.6.7:8080; PROXY 7.8.9.10:8080'
# On prend la premier résultat :
proxy = 'http://{}/'.format(proxies.split()[1].rstrip(';'))
# On injecte les valeurs dans les variables d'environnement
os.environ['HTTP_PROXY'] = os.environ['HTTPS_PROXY'] = proxy