[py] Add support for http proxy authentication to remote_connection (#10358)

* Add support for proxy authentication to remote_connection

Currently, urllib3.ProxyManager is used when connecting to a Selenium
Server via http proxy. The proxy url is read from environment variables.

When using a proxy which requires authentication, credentials are
usually passed in the url like this:
'http://username:password@proxy.com:8080'
urllib3.ProxyManager does not support this, but instead provides a
proxy_header field which takes a basic auth header for authentication.

This commit implements support for this.
If credentials are given in the proxy url, they are seperated out.
Then, the url without credentials and a proxy authentication header,
which is created from the url credentials, are used to create the
ProxyManager instance.

* Fix flake8 style violations

Co-authored-by: David Burns <david.burns@theautomatedtester.co.uk>

Author: luispflamminger

py/selenium/webdriver/remote/remote_connection.py

@@ -128,6 +128,19 @@ def _get_proxy_url(self):
         elif self._url.startswith('http://'):
             return os.environ.get('http_proxy', os.environ.get('HTTP_PROXY'))
 
+    def _identify_http_proxy_auth(self):
+        url = self._proxy_url
+        url = url[url.find(":") + 3:]
+        return True if "@" in url and len(url[:url.find('@')]) > 0 else False
+
+    def _seperate_http_proxy_auth(self):
+        url = self._proxy_url
+        protocol = url[:url.find(":") + 3]
+        no_protocol = url[len(protocol):]
+        auth = no_protocol[:no_protocol.find('@')]
+        proxy_without_auth = protocol + no_protocol[len(auth) + 1:]
+        return proxy_without_auth, auth
+
     def _get_connection_manager(self):
         pool_manager_init_args = {
             'timeout': self._timeout
@@ -140,6 +153,10 @@ def _get_connection_manager(self):
             if self._proxy_url.lower().startswith('sock'):
                 from urllib3.contrib.socks import SOCKSProxyManager
                 return SOCKSProxyManager(self._proxy_url, **pool_manager_init_args)
+            elif self._identify_http_proxy_auth():
+                self._proxy_url, self._basic_proxy_auth = self._seperate_http_proxy_auth()
+                pool_manager_init_args['proxy_headers'] = urllib3.make_headers(
+                    proxy_basic_auth=self._basic_proxy_auth)
             return urllib3.ProxyManager(self._proxy_url, **pool_manager_init_args)
 
         return urllib3.PoolManager(**pool_manager_init_args)

py/test/unit/selenium/webdriver/remote/remote_connection_tests.py

@@ -72,6 +72,24 @@ def test_get_proxy_url_none(mock_proxy_settings_missing):
     assert proxy_url is None
 
 
+def test_get_proxy_url_http_auth(mock_proxy_auth_settings):
+    remote_connection = RemoteConnection('http://remote', keep_alive=False)
+    proxy_url = remote_connection._get_proxy_url()
+    raw_proxy_url, basic_auth_string = remote_connection._seperate_http_proxy_auth()
+    assert proxy_url == "http://user:password@http_proxy.com:8080"
+    assert raw_proxy_url == "http://http_proxy.com:8080"
+    assert basic_auth_string == "user:password"
+
+
+def test_get_proxy_url_https_auth(mock_proxy_auth_settings):
+    remote_connection = RemoteConnection('https://remote', keep_alive=False)
+    proxy_url = remote_connection._get_proxy_url()
+    raw_proxy_url, basic_auth_string = remote_connection._seperate_http_proxy_auth()
+    assert proxy_url == "https://user:password@https_proxy.com:8080"
+    assert raw_proxy_url == "https://https_proxy.com:8080"
+    assert basic_auth_string == "user:password"
+
+
 def test_get_connection_manager_without_proxy(mock_proxy_settings_missing):
     remote_connection = RemoteConnection('http://remote', keep_alive=False)
     conn = remote_connection._get_connection_manager()
@@ -102,6 +120,26 @@ def test_get_connection_manager_with_proxy(mock_proxy_settings):
     assert conn.proxy.port == 8080
 
 
+def test_get_connection_manager_with_auth_proxy(mock_proxy_auth_settings):
+    proxy_auth_header = urllib3.make_headers(
+        proxy_basic_auth="user:password"
+    )
+    remote_connection = RemoteConnection('http://remote', keep_alive=False)
+    conn = remote_connection._get_connection_manager()
+    assert type(conn) == urllib3.ProxyManager
+    assert conn.proxy.scheme == 'http'
+    assert conn.proxy.host == 'http_proxy.com'
+    assert conn.proxy.port == 8080
+    assert conn.proxy_headers == proxy_auth_header
+    remote_connection_https = RemoteConnection('https://remote', keep_alive=False)
+    conn = remote_connection_https._get_connection_manager()
+    assert type(conn) == urllib3.ProxyManager
+    assert conn.proxy.scheme == 'https'
+    assert conn.proxy.host == 'https_proxy.com'
+    assert conn.proxy.port == 8080
+    assert conn.proxy_headers == proxy_auth_header
+
+
 @pytest.mark.parametrize("url",
                          ["*", ".localhost", "localhost:80", "locahost", "127.0.0.1",
                           "LOCALHOST", "LOCALHOST:80", "http://localhost", "https://localhost",
@@ -167,6 +205,16 @@ def mock_proxy_settings(monkeypatch):
     monkeypatch.setenv("http_proxy", http_proxy)
 
 
+@pytest.fixture(scope="function")
+def mock_proxy_auth_settings(monkeypatch):
+    http_proxy = 'http://user:password@http_proxy.com:8080'
+    https_proxy = 'https://user:password@https_proxy.com:8080'
+    monkeypatch.setenv("HTTPS_PROXY", https_proxy)
+    monkeypatch.setenv("HTTP_PROXY", http_proxy)
+    monkeypatch.setenv("https_proxy", https_proxy)
+    monkeypatch.setenv("http_proxy", http_proxy)
+
+
 @pytest.fixture(scope="function")
 def mock_no_proxy_settings(monkeypatch):
     http_proxy = 'http://http_proxy.com:8080'