mirror of
https://github.com/python/cpython.git
synced 2025-07-24 19:54:21 +00:00

It consists of code from urllib, urllib2, urlparse, and robotparser. The old modules have all been removed. The new package has five submodules: urllib.parse, urllib.request, urllib.response, urllib.error, and urllib.robotparser. The urllib.request.urlopen() function uses the url opener from urllib2. Note that the unittests have not been renamed for the beta, but they will be renamed in the future. Joint work with Senthil Kumaran.
459 lines
16 KiB
Python
459 lines
16 KiB
Python
#!/usr/bin/env python
|
|
|
|
import email
|
|
import threading
|
|
import urllib.parse
|
|
import urllib.request
|
|
import http.server
|
|
import unittest
|
|
import hashlib
|
|
from test import support
|
|
|
|
# Loopback http server infrastructure
|
|
|
|
class LoopbackHttpServer(http.server.HTTPServer):
|
|
"""HTTP server w/ a few modifications that make it useful for
|
|
loopback testing purposes.
|
|
"""
|
|
|
|
def __init__(self, server_address, RequestHandlerClass):
|
|
http.server.HTTPServer.__init__(self,
|
|
server_address,
|
|
RequestHandlerClass)
|
|
|
|
# Set the timeout of our listening socket really low so
|
|
# that we can stop the server easily.
|
|
self.socket.settimeout(1.0)
|
|
|
|
def get_request(self):
|
|
"""HTTPServer method, overridden."""
|
|
|
|
request, client_address = self.socket.accept()
|
|
|
|
# It's a loopback connection, so setting the timeout
|
|
# really low shouldn't affect anything, but should make
|
|
# deadlocks less likely to occur.
|
|
request.settimeout(10.0)
|
|
|
|
return (request, client_address)
|
|
|
|
class LoopbackHttpServerThread(threading.Thread):
|
|
"""Stoppable thread that runs a loopback http server."""
|
|
|
|
def __init__(self, request_handler):
|
|
threading.Thread.__init__(self)
|
|
self._stop_server = False
|
|
self.ready = threading.Event()
|
|
request_handler.protocol_version = "HTTP/1.0"
|
|
self.httpd = LoopbackHttpServer(("127.0.0.1", 0),
|
|
request_handler)
|
|
#print "Serving HTTP on %s port %s" % (self.httpd.server_name,
|
|
# self.httpd.server_port)
|
|
self.port = self.httpd.server_port
|
|
|
|
def stop(self):
|
|
"""Stops the webserver if it's currently running."""
|
|
|
|
# Set the stop flag.
|
|
self._stop_server = True
|
|
|
|
self.join()
|
|
|
|
def run(self):
|
|
self.ready.set()
|
|
while not self._stop_server:
|
|
self.httpd.handle_request()
|
|
|
|
# Authentication infrastructure
|
|
|
|
class DigestAuthHandler:
|
|
"""Handler for performing digest authentication."""
|
|
|
|
def __init__(self):
|
|
self._request_num = 0
|
|
self._nonces = []
|
|
self._users = {}
|
|
self._realm_name = "Test Realm"
|
|
self._qop = "auth"
|
|
|
|
def set_qop(self, qop):
|
|
self._qop = qop
|
|
|
|
def set_users(self, users):
|
|
assert isinstance(users, dict)
|
|
self._users = users
|
|
|
|
def set_realm(self, realm):
|
|
self._realm_name = realm
|
|
|
|
def _generate_nonce(self):
|
|
self._request_num += 1
|
|
nonce = hashlib.md5(str(self._request_num).encode("ascii")).hexdigest()
|
|
self._nonces.append(nonce)
|
|
return nonce
|
|
|
|
def _create_auth_dict(self, auth_str):
|
|
first_space_index = auth_str.find(" ")
|
|
auth_str = auth_str[first_space_index+1:]
|
|
|
|
parts = auth_str.split(",")
|
|
|
|
auth_dict = {}
|
|
for part in parts:
|
|
name, value = part.split("=")
|
|
name = name.strip()
|
|
if value[0] == '"' and value[-1] == '"':
|
|
value = value[1:-1]
|
|
else:
|
|
value = value.strip()
|
|
auth_dict[name] = value
|
|
return auth_dict
|
|
|
|
def _validate_auth(self, auth_dict, password, method, uri):
|
|
final_dict = {}
|
|
final_dict.update(auth_dict)
|
|
final_dict["password"] = password
|
|
final_dict["method"] = method
|
|
final_dict["uri"] = uri
|
|
HA1_str = "%(username)s:%(realm)s:%(password)s" % final_dict
|
|
HA1 = hashlib.md5(HA1_str.encode("ascii")).hexdigest()
|
|
HA2_str = "%(method)s:%(uri)s" % final_dict
|
|
HA2 = hashlib.md5(HA2_str.encode("ascii")).hexdigest()
|
|
final_dict["HA1"] = HA1
|
|
final_dict["HA2"] = HA2
|
|
response_str = "%(HA1)s:%(nonce)s:%(nc)s:" \
|
|
"%(cnonce)s:%(qop)s:%(HA2)s" % final_dict
|
|
response = hashlib.md5(response_str.encode("ascii")).hexdigest()
|
|
|
|
return response == auth_dict["response"]
|
|
|
|
def _return_auth_challenge(self, request_handler):
|
|
request_handler.send_response(407, "Proxy Authentication Required")
|
|
request_handler.send_header("Content-Type", "text/html")
|
|
request_handler.send_header(
|
|
'Proxy-Authenticate', 'Digest realm="%s", '
|
|
'qop="%s",'
|
|
'nonce="%s", ' % \
|
|
(self._realm_name, self._qop, self._generate_nonce()))
|
|
# XXX: Not sure if we're supposed to add this next header or
|
|
# not.
|
|
#request_handler.send_header('Connection', 'close')
|
|
request_handler.end_headers()
|
|
request_handler.wfile.write(b"Proxy Authentication Required.")
|
|
return False
|
|
|
|
def handle_request(self, request_handler):
|
|
"""Performs digest authentication on the given HTTP request
|
|
handler. Returns True if authentication was successful, False
|
|
otherwise.
|
|
|
|
If no users have been set, then digest auth is effectively
|
|
disabled and this method will always return True.
|
|
"""
|
|
|
|
if len(self._users) == 0:
|
|
return True
|
|
|
|
if "Proxy-Authorization" not in request_handler.headers:
|
|
return self._return_auth_challenge(request_handler)
|
|
else:
|
|
auth_dict = self._create_auth_dict(
|
|
request_handler.headers["Proxy-Authorization"]
|
|
)
|
|
if auth_dict["username"] in self._users:
|
|
password = self._users[ auth_dict["username"] ]
|
|
else:
|
|
return self._return_auth_challenge(request_handler)
|
|
if not auth_dict.get("nonce") in self._nonces:
|
|
return self._return_auth_challenge(request_handler)
|
|
else:
|
|
self._nonces.remove(auth_dict["nonce"])
|
|
|
|
auth_validated = False
|
|
|
|
# MSIE uses short_path in its validation, but Python's
|
|
# urllib2 uses the full path, so we're going to see if
|
|
# either of them works here.
|
|
|
|
for path in [request_handler.path, request_handler.short_path]:
|
|
if self._validate_auth(auth_dict,
|
|
password,
|
|
request_handler.command,
|
|
path):
|
|
auth_validated = True
|
|
|
|
if not auth_validated:
|
|
return self._return_auth_challenge(request_handler)
|
|
return True
|
|
|
|
# Proxy test infrastructure
|
|
|
|
class FakeProxyHandler(http.server.BaseHTTPRequestHandler):
|
|
"""This is a 'fake proxy' that makes it look like the entire
|
|
internet has gone down due to a sudden zombie invasion. It main
|
|
utility is in providing us with authentication support for
|
|
testing.
|
|
"""
|
|
|
|
digest_auth_handler = DigestAuthHandler()
|
|
|
|
def log_message(self, format, *args):
|
|
# Uncomment the next line for debugging.
|
|
# sys.stderr.write(format % args)
|
|
pass
|
|
|
|
def do_GET(self):
|
|
(scm, netloc, path, params, query, fragment) = urllib.parse.urlparse(
|
|
self.path, "http")
|
|
self.short_path = path
|
|
if self.digest_auth_handler.handle_request(self):
|
|
self.send_response(200, "OK")
|
|
self.send_header("Content-Type", "text/html")
|
|
self.end_headers()
|
|
self.wfile.write(bytes("You've reached %s!<BR>" % self.path,
|
|
"ascii"))
|
|
self.wfile.write(b"Our apologies, but our server is down due to "
|
|
b"a sudden zombie invasion.")
|
|
|
|
# Test cases
|
|
|
|
class ProxyAuthTests(unittest.TestCase):
|
|
URL = "http://localhost"
|
|
|
|
USER = "tester"
|
|
PASSWD = "test123"
|
|
REALM = "TestRealm"
|
|
|
|
def setUp(self):
|
|
FakeProxyHandler.digest_auth_handler.set_users({
|
|
self.USER : self.PASSWD
|
|
})
|
|
FakeProxyHandler.digest_auth_handler.set_realm(self.REALM)
|
|
|
|
self.server = LoopbackHttpServerThread(FakeProxyHandler)
|
|
self.server.start()
|
|
self.server.ready.wait()
|
|
proxy_url = "http://127.0.0.1:%d" % self.server.port
|
|
handler = urllib.request.ProxyHandler({"http" : proxy_url})
|
|
self._digest_auth_handler = urllib.request.ProxyDigestAuthHandler()
|
|
self.opener = urllib.request.build_opener(
|
|
handler, self._digest_auth_handler)
|
|
|
|
def tearDown(self):
|
|
self.server.stop()
|
|
|
|
def test_proxy_with_bad_password_raises_httperror(self):
|
|
self._digest_auth_handler.add_password(self.REALM, self.URL,
|
|
self.USER, self.PASSWD+"bad")
|
|
FakeProxyHandler.digest_auth_handler.set_qop("auth")
|
|
self.assertRaises(urllib.error.HTTPError,
|
|
self.opener.open,
|
|
self.URL)
|
|
|
|
def test_proxy_with_no_password_raises_httperror(self):
|
|
FakeProxyHandler.digest_auth_handler.set_qop("auth")
|
|
self.assertRaises(urllib.error.HTTPError,
|
|
self.opener.open,
|
|
self.URL)
|
|
|
|
def test_proxy_qop_auth_works(self):
|
|
self._digest_auth_handler.add_password(self.REALM, self.URL,
|
|
self.USER, self.PASSWD)
|
|
FakeProxyHandler.digest_auth_handler.set_qop("auth")
|
|
result = self.opener.open(self.URL)
|
|
while result.read():
|
|
pass
|
|
result.close()
|
|
|
|
def test_proxy_qop_auth_int_works_or_throws_urlerror(self):
|
|
self._digest_auth_handler.add_password(self.REALM, self.URL,
|
|
self.USER, self.PASSWD)
|
|
FakeProxyHandler.digest_auth_handler.set_qop("auth-int")
|
|
try:
|
|
result = self.opener.open(self.URL)
|
|
except urllib.error.URLError:
|
|
# It's okay if we don't support auth-int, but we certainly
|
|
# shouldn't receive any kind of exception here other than
|
|
# a URLError.
|
|
result = None
|
|
if result:
|
|
while result.read():
|
|
pass
|
|
result.close()
|
|
|
|
|
|
def GetRequestHandler(responses):
|
|
|
|
class FakeHTTPRequestHandler(http.server.BaseHTTPRequestHandler):
|
|
|
|
server_version = "TestHTTP/"
|
|
requests = []
|
|
headers_received = []
|
|
port = 80
|
|
|
|
def do_GET(self):
|
|
body = self.send_head()
|
|
if body:
|
|
self.wfile.write(body)
|
|
|
|
def do_POST(self):
|
|
content_length = self.headers["Content-Length"]
|
|
post_data = self.rfile.read(int(content_length))
|
|
self.do_GET()
|
|
self.requests.append(post_data)
|
|
|
|
def send_head(self):
|
|
FakeHTTPRequestHandler.headers_received = self.headers
|
|
self.requests.append(self.path)
|
|
response_code, headers, body = responses.pop(0)
|
|
|
|
self.send_response(response_code)
|
|
|
|
for (header, value) in headers:
|
|
self.send_header(header, value % self.port)
|
|
if body:
|
|
self.send_header("Content-type", "text/plain")
|
|
self.end_headers()
|
|
return body
|
|
self.end_headers()
|
|
|
|
def log_message(self, *args):
|
|
pass
|
|
|
|
|
|
return FakeHTTPRequestHandler
|
|
|
|
|
|
class TestUrlopen(unittest.TestCase):
|
|
"""Tests urllib2.urlopen using the network.
|
|
|
|
These tests are not exhaustive. Assuming that testing using files does a
|
|
good job overall of some of the basic interface features. There are no
|
|
tests exercising the optional 'data' and 'proxies' arguments. No tests
|
|
for transparent redirection have been written.
|
|
"""
|
|
|
|
def setUp(self):
|
|
self.server = None
|
|
|
|
def tearDown(self):
|
|
if self.server is not None:
|
|
self.server.stop()
|
|
|
|
def urlopen(self, url, data=None):
|
|
f = urllib.request.urlopen(url, data)
|
|
result = f.read()
|
|
f.close()
|
|
return result
|
|
|
|
def start_server(self, responses=None):
|
|
if responses is None:
|
|
responses = [(200, [], b"we don't care")]
|
|
handler = GetRequestHandler(responses)
|
|
|
|
self.server = LoopbackHttpServerThread(handler)
|
|
self.server.start()
|
|
self.server.ready.wait()
|
|
port = self.server.port
|
|
handler.port = port
|
|
return handler
|
|
|
|
def test_redirection(self):
|
|
expected_response = b"We got here..."
|
|
responses = [
|
|
(302, [("Location", "http://localhost:%s/somewhere_else")], ""),
|
|
(200, [], expected_response)
|
|
]
|
|
|
|
handler = self.start_server(responses)
|
|
data = self.urlopen("http://localhost:%s/" % handler.port)
|
|
self.assertEquals(data, expected_response)
|
|
self.assertEquals(handler.requests, ["/", "/somewhere_else"])
|
|
|
|
def test_404(self):
|
|
expected_response = b"Bad bad bad..."
|
|
handler = self.start_server([(404, [], expected_response)])
|
|
|
|
try:
|
|
self.urlopen("http://localhost:%s/weeble" % handler.port)
|
|
except urllib.error.URLError as f:
|
|
data = f.read()
|
|
f.close()
|
|
else:
|
|
self.fail("404 should raise URLError")
|
|
|
|
self.assertEquals(data, expected_response)
|
|
self.assertEquals(handler.requests, ["/weeble"])
|
|
|
|
def test_200(self):
|
|
expected_response = b"pycon 2008..."
|
|
handler = self.start_server([(200, [], expected_response)])
|
|
data = self.urlopen("http://localhost:%s/bizarre" % handler.port)
|
|
self.assertEquals(data, expected_response)
|
|
self.assertEquals(handler.requests, ["/bizarre"])
|
|
|
|
def test_200_with_parameters(self):
|
|
expected_response = b"pycon 2008..."
|
|
handler = self.start_server([(200, [], expected_response)])
|
|
data = self.urlopen("http://localhost:%s/bizarre" % handler.port,
|
|
b"get=with_feeling")
|
|
self.assertEquals(data, expected_response)
|
|
self.assertEquals(handler.requests, ["/bizarre", b"get=with_feeling"])
|
|
|
|
def test_sending_headers(self):
|
|
handler = self.start_server()
|
|
req = urllib.request.Request("http://localhost:%s/" % handler.port,
|
|
headers={"Range": "bytes=20-39"})
|
|
urllib.request.urlopen(req)
|
|
self.assertEqual(handler.headers_received["Range"], "bytes=20-39")
|
|
|
|
def test_basic(self):
|
|
handler = self.start_server()
|
|
open_url = urllib.request.urlopen("http://localhost:%s" % handler.port)
|
|
for attr in ("read", "close", "info", "geturl"):
|
|
self.assert_(hasattr(open_url, attr), "object returned from "
|
|
"urlopen lacks the %s attribute" % attr)
|
|
try:
|
|
self.assert_(open_url.read(), "calling 'read' failed")
|
|
finally:
|
|
open_url.close()
|
|
|
|
def test_info(self):
|
|
handler = self.start_server()
|
|
try:
|
|
open_url = urllib.request.urlopen(
|
|
"http://localhost:%s" % handler.port)
|
|
info_obj = open_url.info()
|
|
self.assert_(isinstance(info_obj, email.message.Message),
|
|
"object returned by 'info' is not an instance of "
|
|
"email.message.Message")
|
|
self.assertEqual(info_obj.get_content_subtype(), "plain")
|
|
finally:
|
|
self.server.stop()
|
|
|
|
def test_geturl(self):
|
|
# Make sure same URL as opened is returned by geturl.
|
|
handler = self.start_server()
|
|
open_url = urllib.request.urlopen("http://localhost:%s" % handler.port)
|
|
url = open_url.geturl()
|
|
self.assertEqual(url, "http://localhost:%s" % handler.port)
|
|
|
|
def test_bad_address(self):
|
|
# Make sure proper exception is raised when connecting to a bogus
|
|
# address.
|
|
self.assertRaises(IOError,
|
|
# SF patch 809915: In Sep 2003, VeriSign started
|
|
# highjacking invalid .com and .net addresses to
|
|
# boost traffic to their own site. This test
|
|
# started failing then. One hopes the .invalid
|
|
# domain will be spared to serve its defined
|
|
# purpose.
|
|
urllib.request.urlopen,
|
|
"http://www.python.invalid./")
|
|
|
|
def test_main():
|
|
support.run_unittest(ProxyAuthTests)
|
|
support.run_unittest(TestUrlopen)
|
|
|
|
if __name__ == "__main__":
|
|
test_main()
|