blob: b2cec7eeb61a6cee03fcfe1723d88a6c4883a85d [file] [log] [blame]
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001import unittest
Benjamin Petersonee8712c2008-05-20 21:35:26 +00002from test import support
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00003
Christian Heimes05e8be12008-02-23 18:30:17 +00004import os
Guido van Rossum34d19282007-08-09 01:03:29 +00005import io
Georg Brandlf78e02b2008-06-10 17:40:04 +00006import socket
Senthil Kumaran7bc0d872010-12-19 10:49:52 +00007import array
Senthil Kumaran4de00a22011-05-11 21:17:57 +08008import sys
Jeremy Hyltone3e61042001-05-09 15:50:25 +00009
Jeremy Hylton1afc1692008-06-18 20:49:58 +000010import urllib.request
Ronald Oussorene72e1612011-03-14 18:15:25 -040011# The proxy bypass method imported below has logic specific to the OSX
12# proxy config data structure but is testable on all platforms.
13from urllib.request import Request, OpenerDirector, _proxy_bypass_macosx_sysconf
guido@google.coma119df92011-03-29 11:41:02 -070014import urllib.error
Jeremy Hyltone3e61042001-05-09 15:50:25 +000015
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +000016# XXX
17# Request
18# CacheFTPHandler (hard to write)
Thomas Wouters477c8d52006-05-27 19:21:47 +000019# parse_keqv_list, parse_http_list, HTTPDigestAuthHandler
Jeremy Hyltone3e61042001-05-09 15:50:25 +000020
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +000021class TrivialTests(unittest.TestCase):
Senthil Kumaran6c5bd402011-11-01 23:20:31 +080022
23 def test___all__(self):
24 # Verify which names are exposed
25 for module in 'request', 'response', 'parse', 'error', 'robotparser':
26 context = {}
27 exec('from urllib.%s import *' % module, context)
28 del context['__builtins__']
Florent Xicluna3dbb1f12011-11-04 22:15:37 +010029 if module == 'request' and os.name == 'nt':
30 u, p = context.pop('url2pathname'), context.pop('pathname2url')
31 self.assertEqual(u.__module__, 'nturl2path')
32 self.assertEqual(p.__module__, 'nturl2path')
Senthil Kumaran6c5bd402011-11-01 23:20:31 +080033 for k, v in context.items():
34 self.assertEqual(v.__module__, 'urllib.%s' % module,
35 "%r is exposed in 'urllib.%s' but defined in %r" %
36 (k, module, v.__module__))
37
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +000038 def test_trivial(self):
39 # A couple trivial tests
Guido van Rossume2ae77b2001-10-24 20:42:55 +000040
Jeremy Hylton1afc1692008-06-18 20:49:58 +000041 self.assertRaises(ValueError, urllib.request.urlopen, 'bogus url')
Tim Peters861adac2001-07-16 20:49:49 +000042
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +000043 # XXX Name hacking to get this to work on Windows.
Jeremy Hylton1afc1692008-06-18 20:49:58 +000044 fname = os.path.abspath(urllib.request.__file__).replace('\\', '/')
Senthil Kumarand587e302010-01-10 17:45:52 +000045
Senthil Kumarand587e302010-01-10 17:45:52 +000046 if os.name == 'nt':
47 file_url = "file:///%s" % fname
48 else:
49 file_url = "file://%s" % fname
50
Jeremy Hylton1afc1692008-06-18 20:49:58 +000051 f = urllib.request.urlopen(file_url)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +000052
53 buf = f.read()
54 f.close()
Tim Petersf5f32b42005-07-17 23:16:17 +000055
Georg Brandle1b13d22005-08-24 22:20:32 +000056 def test_parse_http_list(self):
Jeremy Hylton1afc1692008-06-18 20:49:58 +000057 tests = [
58 ('a,b,c', ['a', 'b', 'c']),
59 ('path"o,l"og"i"cal, example', ['path"o,l"og"i"cal', 'example']),
60 ('a, b, "c", "d", "e,f", g, h',
61 ['a', 'b', '"c"', '"d"', '"e,f"', 'g', 'h']),
62 ('a="b\\"c", d="e\\,f", g="h\\\\i"',
63 ['a="b"c"', 'd="e,f"', 'g="h\\i"'])]
Georg Brandle1b13d22005-08-24 22:20:32 +000064 for string, list in tests:
Florent Xicluna419e3842010-08-08 16:16:07 +000065 self.assertEqual(urllib.request.parse_http_list(string), list)
Georg Brandle1b13d22005-08-24 22:20:32 +000066
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +000067
Thomas Wouters00ee7ba2006-08-21 19:07:27 +000068def test_request_headers_dict():
69 """
70 The Request.headers dictionary is not a documented interface. It should
71 stay that way, because the complete set of headers are only accessible
72 through the .get_header(), .has_header(), .header_items() interface.
73 However, .headers pre-dates those methods, and so real code will be using
74 the dictionary.
75
76 The introduction in 2.4 of those methods was a mistake for the same reason:
77 code that previously saw all (urllib2 user)-provided headers in .headers
78 now sees only a subset (and the function interface is ugly and incomplete).
79 A better change would have been to replace .headers dict with a dict
80 subclass (or UserDict.DictMixin instance?) that preserved the .headers
81 interface and also provided access to the "unredirected" headers. It's
82 probably too late to fix that, though.
83
84
85 Check .capitalize() case normalization:
86
87 >>> url = "http://example.com"
88 >>> Request(url, headers={"Spam-eggs": "blah"}).headers["Spam-eggs"]
89 'blah'
90 >>> Request(url, headers={"spam-EggS": "blah"}).headers["Spam-eggs"]
91 'blah'
92
93 Currently, Request(url, "Spam-eggs").headers["Spam-Eggs"] raises KeyError,
94 but that could be changed in future.
95
96 """
97
98def test_request_headers_methods():
99 """
100 Note the case normalization of header names here, to .capitalize()-case.
101 This should be preserved for backwards-compatibility. (In the HTTP case,
102 normalization to .title()-case is done by urllib2 before sending headers to
Georg Brandl24420152008-05-26 16:32:26 +0000103 http.client).
Thomas Wouters00ee7ba2006-08-21 19:07:27 +0000104
105 >>> url = "http://example.com"
106 >>> r = Request(url, headers={"Spam-eggs": "blah"})
107 >>> r.has_header("Spam-eggs")
108 True
109 >>> r.header_items()
110 [('Spam-eggs', 'blah')]
111 >>> r.add_header("Foo-Bar", "baz")
Guido van Rossumcc2b0162007-02-11 06:12:03 +0000112 >>> items = sorted(r.header_items())
Thomas Wouters00ee7ba2006-08-21 19:07:27 +0000113 >>> items
114 [('Foo-bar', 'baz'), ('Spam-eggs', 'blah')]
115
116 Note that e.g. r.has_header("spam-EggS") is currently False, and
117 r.get_header("spam-EggS") returns None, but that could be changed in
118 future.
119
120 >>> r.has_header("Not-there")
121 False
Guido van Rossum7131f842007-02-09 20:13:25 +0000122 >>> print(r.get_header("Not-there"))
Thomas Wouters00ee7ba2006-08-21 19:07:27 +0000123 None
124 >>> r.get_header("Not-there", "default")
125 'default'
126
127 """
128
129
Thomas Wouters477c8d52006-05-27 19:21:47 +0000130def test_password_manager(self):
131 """
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000132 >>> mgr = urllib.request.HTTPPasswordMgr()
Thomas Wouters477c8d52006-05-27 19:21:47 +0000133 >>> add = mgr.add_password
134 >>> add("Some Realm", "http://example.com/", "joe", "password")
135 >>> add("Some Realm", "http://example.com/ni", "ni", "ni")
136 >>> add("c", "http://example.com/foo", "foo", "ni")
137 >>> add("c", "http://example.com/bar", "bar", "nini")
138 >>> add("b", "http://example.com/", "first", "blah")
139 >>> add("b", "http://example.com/", "second", "spam")
140 >>> add("a", "http://example.com", "1", "a")
141 >>> add("Some Realm", "http://c.example.com:3128", "3", "c")
142 >>> add("Some Realm", "d.example.com", "4", "d")
143 >>> add("Some Realm", "e.example.com:3128", "5", "e")
144
145 >>> mgr.find_user_password("Some Realm", "example.com")
146 ('joe', 'password')
147 >>> mgr.find_user_password("Some Realm", "http://example.com")
148 ('joe', 'password')
149 >>> mgr.find_user_password("Some Realm", "http://example.com/")
150 ('joe', 'password')
151 >>> mgr.find_user_password("Some Realm", "http://example.com/spam")
152 ('joe', 'password')
153 >>> mgr.find_user_password("Some Realm", "http://example.com/spam/spam")
154 ('joe', 'password')
155 >>> mgr.find_user_password("c", "http://example.com/foo")
156 ('foo', 'ni')
157 >>> mgr.find_user_password("c", "http://example.com/bar")
158 ('bar', 'nini')
159
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000160 Actually, this is really undefined ATM
161## Currently, we use the highest-level path where more than one match:
Thomas Wouters477c8d52006-05-27 19:21:47 +0000162
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000163## >>> mgr.find_user_password("Some Realm", "http://example.com/ni")
164## ('joe', 'password')
Thomas Wouters477c8d52006-05-27 19:21:47 +0000165
166 Use latest add_password() in case of conflict:
167
168 >>> mgr.find_user_password("b", "http://example.com/")
169 ('second', 'spam')
170
171 No special relationship between a.example.com and example.com:
172
173 >>> mgr.find_user_password("a", "http://example.com/")
174 ('1', 'a')
175 >>> mgr.find_user_password("a", "http://a.example.com/")
176 (None, None)
177
178 Ports:
179
180 >>> mgr.find_user_password("Some Realm", "c.example.com")
181 (None, None)
182 >>> mgr.find_user_password("Some Realm", "c.example.com:3128")
183 ('3', 'c')
184 >>> mgr.find_user_password("Some Realm", "http://c.example.com:3128")
185 ('3', 'c')
186 >>> mgr.find_user_password("Some Realm", "d.example.com")
187 ('4', 'd')
188 >>> mgr.find_user_password("Some Realm", "e.example.com:3128")
189 ('5', 'e')
190
191 """
192 pass
193
194
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000195def test_password_manager_default_port(self):
196 """
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000197 >>> mgr = urllib.request.HTTPPasswordMgr()
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000198 >>> add = mgr.add_password
199
200 The point to note here is that we can't guess the default port if there's
201 no scheme. This applies to both add_password and find_user_password.
202
203 >>> add("f", "http://g.example.com:80", "10", "j")
204 >>> add("g", "http://h.example.com", "11", "k")
205 >>> add("h", "i.example.com:80", "12", "l")
206 >>> add("i", "j.example.com", "13", "m")
207 >>> mgr.find_user_password("f", "g.example.com:100")
208 (None, None)
209 >>> mgr.find_user_password("f", "g.example.com:80")
210 ('10', 'j')
211 >>> mgr.find_user_password("f", "g.example.com")
212 (None, None)
213 >>> mgr.find_user_password("f", "http://g.example.com:100")
214 (None, None)
215 >>> mgr.find_user_password("f", "http://g.example.com:80")
216 ('10', 'j')
217 >>> mgr.find_user_password("f", "http://g.example.com")
218 ('10', 'j')
219 >>> mgr.find_user_password("g", "h.example.com")
220 ('11', 'k')
221 >>> mgr.find_user_password("g", "h.example.com:80")
222 ('11', 'k')
223 >>> mgr.find_user_password("g", "http://h.example.com:80")
224 ('11', 'k')
225 >>> mgr.find_user_password("h", "i.example.com")
226 (None, None)
227 >>> mgr.find_user_password("h", "i.example.com:80")
228 ('12', 'l')
229 >>> mgr.find_user_password("h", "http://i.example.com:80")
230 ('12', 'l')
231 >>> mgr.find_user_password("i", "j.example.com")
232 ('13', 'm')
233 >>> mgr.find_user_password("i", "j.example.com:80")
234 (None, None)
235 >>> mgr.find_user_password("i", "http://j.example.com")
236 ('13', 'm')
237 >>> mgr.find_user_password("i", "http://j.example.com:80")
238 (None, None)
239
240 """
241
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000242class MockOpener:
243 addheaders = []
Senthil Kumaranfb8cc2f2009-07-19 02:44:19 +0000244 def open(self, req, data=None, timeout=socket._GLOBAL_DEFAULT_TIMEOUT):
245 self.req, self.data, self.timeout = req, data, timeout
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000246 def error(self, proto, *args):
247 self.proto, self.args = proto, args
248
249class MockFile:
250 def read(self, count=None): pass
251 def readline(self, count=None): pass
252 def close(self): pass
253
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000254class MockHeaders(dict):
255 def getheaders(self, name):
Guido van Rossumcc2b0162007-02-11 06:12:03 +0000256 return list(self.values())
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000257
Guido van Rossum34d19282007-08-09 01:03:29 +0000258class MockResponse(io.StringIO):
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000259 def __init__(self, code, msg, headers, data, url=None):
Guido van Rossum34d19282007-08-09 01:03:29 +0000260 io.StringIO.__init__(self, data)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000261 self.code, self.msg, self.headers, self.url = code, msg, headers, url
262 def info(self):
263 return self.headers
264 def geturl(self):
265 return self.url
266
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000267class MockCookieJar:
268 def add_cookie_header(self, request):
269 self.ach_req = request
270 def extract_cookies(self, response, request):
271 self.ec_req, self.ec_r = request, response
272
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000273class FakeMethod:
274 def __init__(self, meth_name, action, handle):
275 self.meth_name = meth_name
276 self.handle = handle
277 self.action = action
278 def __call__(self, *args):
279 return self.handle(self.meth_name, self.action, *args)
280
Senthil Kumaran47fff872009-12-20 07:10:31 +0000281class MockHTTPResponse(io.IOBase):
282 def __init__(self, fp, msg, status, reason):
283 self.fp = fp
284 self.msg = msg
285 self.status = status
286 self.reason = reason
287 self.code = 200
288
289 def read(self):
290 return ''
291
292 def info(self):
293 return {}
294
295 def geturl(self):
296 return self.url
297
298
299class MockHTTPClass:
300 def __init__(self):
301 self.level = 0
302 self.req_headers = []
303 self.data = None
304 self.raise_on_endheaders = False
305 self._tunnel_headers = {}
306
307 def __call__(self, host, timeout=socket._GLOBAL_DEFAULT_TIMEOUT):
308 self.host = host
309 self.timeout = timeout
310 return self
311
312 def set_debuglevel(self, level):
313 self.level = level
314
315 def set_tunnel(self, host, port=None, headers=None):
316 self._tunnel_host = host
317 self._tunnel_port = port
318 if headers:
319 self._tunnel_headers = headers
320 else:
321 self._tunnel_headers.clear()
322
Benjamin Peterson3d5b8db2009-12-24 01:14:05 +0000323 def request(self, method, url, body=None, headers=None):
Senthil Kumaran47fff872009-12-20 07:10:31 +0000324 self.method = method
325 self.selector = url
Benjamin Peterson3d5b8db2009-12-24 01:14:05 +0000326 if headers is not None:
327 self.req_headers += headers.items()
Senthil Kumaran47fff872009-12-20 07:10:31 +0000328 self.req_headers.sort()
329 if body:
330 self.data = body
331 if self.raise_on_endheaders:
332 import socket
333 raise socket.error()
334 def getresponse(self):
335 return MockHTTPResponse(MockFile(), {}, 200, "OK")
336
Victor Stinnera4c45d72011-06-17 14:01:18 +0200337 def close(self):
338 pass
339
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000340class MockHandler:
Thomas Wouters477c8d52006-05-27 19:21:47 +0000341 # useful for testing handler machinery
342 # see add_ordered_mock_handlers() docstring
Thomas Wouters49fd7fa2006-04-21 10:40:58 +0000343 handler_order = 500
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000344 def __init__(self, methods):
345 self._define_methods(methods)
346 def _define_methods(self, methods):
347 for spec in methods:
348 if len(spec) == 2: name, action = spec
349 else: name, action = spec, None
350 meth = FakeMethod(name, action, self.handle)
351 setattr(self.__class__, name, meth)
352 def handle(self, fn_name, action, *args, **kwds):
353 self.parent.calls.append((self, fn_name, args, kwds))
354 if action is None:
355 return None
356 elif action == "return self":
357 return self
358 elif action == "return response":
359 res = MockResponse(200, "OK", {}, "")
360 return res
361 elif action == "return request":
362 return Request("http://blah/")
363 elif action.startswith("error"):
364 code = action[action.rfind(" ")+1:]
365 try:
366 code = int(code)
367 except ValueError:
368 pass
369 res = MockResponse(200, "OK", {}, "")
370 return self.parent.error("http", args[0], res, code, "", {})
371 elif action == "raise":
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000372 raise urllib.error.URLError("blah")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000373 assert False
374 def close(self): pass
375 def add_parent(self, parent):
376 self.parent = parent
377 self.parent.calls = []
378 def __lt__(self, other):
379 if not hasattr(other, "handler_order"):
380 # No handler_order, leave in original order. Yuck.
381 return True
382 return self.handler_order < other.handler_order
383
384def add_ordered_mock_handlers(opener, meth_spec):
385 """Create MockHandlers and add them to an OpenerDirector.
386
387 meth_spec: list of lists of tuples and strings defining methods to define
388 on handlers. eg:
389
390 [["http_error", "ftp_open"], ["http_open"]]
391
392 defines methods .http_error() and .ftp_open() on one handler, and
393 .http_open() on another. These methods just record their arguments and
394 return None. Using a tuple instead of a string causes the method to
395 perform some action (see MockHandler.handle()), eg:
396
397 [["http_error"], [("http_open", "return request")]]
398
399 defines .http_error() on one handler (which simply returns None), and
400 .http_open() on another handler, which returns a Request object.
401
402 """
403 handlers = []
404 count = 0
405 for meths in meth_spec:
406 class MockHandlerSubclass(MockHandler): pass
407 h = MockHandlerSubclass(meths)
Thomas Wouters49fd7fa2006-04-21 10:40:58 +0000408 h.handler_order += count
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000409 h.add_parent(opener)
410 count = count + 1
411 handlers.append(h)
412 opener.add_handler(h)
413 return handlers
414
Thomas Wouters477c8d52006-05-27 19:21:47 +0000415def build_test_opener(*handler_instances):
416 opener = OpenerDirector()
417 for h in handler_instances:
418 opener.add_handler(h)
419 return opener
420
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000421class MockHTTPHandler(urllib.request.BaseHandler):
Thomas Wouters477c8d52006-05-27 19:21:47 +0000422 # useful for testing redirections and auth
423 # sends supplied headers and code as first response
424 # sends 200 OK as second response
425 def __init__(self, code, headers):
426 self.code = code
427 self.headers = headers
428 self.reset()
429 def reset(self):
430 self._count = 0
431 self.requests = []
432 def http_open(self, req):
Barry Warsaw820c1202008-06-12 04:06:45 +0000433 import email, http.client, copy
Guido van Rossum34d19282007-08-09 01:03:29 +0000434 from io import StringIO
Thomas Wouters477c8d52006-05-27 19:21:47 +0000435 self.requests.append(copy.deepcopy(req))
436 if self._count == 0:
437 self._count = self._count + 1
Georg Brandl24420152008-05-26 16:32:26 +0000438 name = http.client.responses[self.code]
Barry Warsaw820c1202008-06-12 04:06:45 +0000439 msg = email.message_from_string(self.headers)
Thomas Wouters477c8d52006-05-27 19:21:47 +0000440 return self.parent.error(
441 "http", req, MockFile(), self.code, name, msg)
442 else:
443 self.req = req
Barry Warsaw820c1202008-06-12 04:06:45 +0000444 msg = email.message_from_string("\r\n\r\n")
Thomas Wouters477c8d52006-05-27 19:21:47 +0000445 return MockResponse(200, "OK", msg, "", req.get_full_url())
446
Senthil Kumaran47fff872009-12-20 07:10:31 +0000447class MockHTTPSHandler(urllib.request.AbstractHTTPHandler):
448 # Useful for testing the Proxy-Authorization request by verifying the
449 # properties of httpcon
Benjamin Peterson3d5b8db2009-12-24 01:14:05 +0000450
451 def __init__(self):
452 urllib.request.AbstractHTTPHandler.__init__(self)
453 self.httpconn = MockHTTPClass()
454
Senthil Kumaran47fff872009-12-20 07:10:31 +0000455 def https_open(self, req):
456 return self.do_open(self.httpconn, req)
457
Thomas Wouters477c8d52006-05-27 19:21:47 +0000458class MockPasswordManager:
459 def add_password(self, realm, uri, user, password):
460 self.realm = realm
461 self.url = uri
462 self.user = user
463 self.password = password
464 def find_user_password(self, realm, authuri):
465 self.target_realm = realm
466 self.target_url = authuri
467 return self.user, self.password
468
469
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000470class OpenerDirectorTests(unittest.TestCase):
471
Guido van Rossumb5a755e2007-07-18 18:15:48 +0000472 def test_add_non_handler(self):
473 class NonHandler(object):
474 pass
475 self.assertRaises(TypeError,
476 OpenerDirector().add_handler, NonHandler())
477
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000478 def test_badly_named_methods(self):
479 # test work-around for three methods that accidentally follow the
480 # naming conventions for handler methods
481 # (*_open() / *_request() / *_response())
482
483 # These used to call the accidentally-named methods, causing a
484 # TypeError in real code; here, returning self from these mock
485 # methods would either cause no exception, or AttributeError.
486
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000487 from urllib.error import URLError
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000488
489 o = OpenerDirector()
490 meth_spec = [
491 [("do_open", "return self"), ("proxy_open", "return self")],
492 [("redirect_request", "return self")],
493 ]
494 handlers = add_ordered_mock_handlers(o, meth_spec)
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000495 o.add_handler(urllib.request.UnknownHandler())
Thomas Wouters4d70c3d2006-06-08 14:42:34 +0000496 for scheme in "do", "proxy", "redirect":
497 self.assertRaises(URLError, o.open, scheme+"://example.com/")
498
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000499 def test_handled(self):
500 # handler returning non-None means no more handlers will be called
501 o = OpenerDirector()
502 meth_spec = [
503 ["http_open", "ftp_open", "http_error_302"],
504 ["ftp_open"],
505 [("http_open", "return self")],
506 [("http_open", "return self")],
507 ]
508 handlers = add_ordered_mock_handlers(o, meth_spec)
509
510 req = Request("http://example.com/")
511 r = o.open(req)
512 # Second .http_open() gets called, third doesn't, since second returned
513 # non-None. Handlers without .http_open() never get any methods called
514 # on them.
515 # In fact, second mock handler defining .http_open() returns self
516 # (instead of response), which becomes the OpenerDirector's return
517 # value.
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000518 self.assertEqual(r, handlers[2])
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000519 calls = [(handlers[0], "http_open"), (handlers[2], "http_open")]
520 for expected, got in zip(calls, o.calls):
521 handler, name, args, kwds = got
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000522 self.assertEqual((handler, name), expected)
523 self.assertEqual(args, (req,))
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000524
525 def test_handler_order(self):
526 o = OpenerDirector()
527 handlers = []
528 for meths, handler_order in [
529 ([("http_open", "return self")], 500),
530 (["http_open"], 0),
531 ]:
532 class MockHandlerSubclass(MockHandler): pass
533 h = MockHandlerSubclass(meths)
534 h.handler_order = handler_order
535 handlers.append(h)
536 o.add_handler(h)
537
538 r = o.open("http://example.com/")
539 # handlers called in reverse order, thanks to their sort order
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000540 self.assertEqual(o.calls[0][0], handlers[1])
541 self.assertEqual(o.calls[1][0], handlers[0])
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000542
543 def test_raise(self):
544 # raising URLError stops processing of request
545 o = OpenerDirector()
546 meth_spec = [
547 [("http_open", "raise")],
548 [("http_open", "return self")],
549 ]
550 handlers = add_ordered_mock_handlers(o, meth_spec)
551
552 req = Request("http://example.com/")
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000553 self.assertRaises(urllib.error.URLError, o.open, req)
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000554 self.assertEqual(o.calls, [(handlers[0], "http_open", (req,), {})])
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000555
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000556 def test_http_error(self):
557 # XXX http_error_default
558 # http errors are a special case
559 o = OpenerDirector()
560 meth_spec = [
561 [("http_open", "error 302")],
562 [("http_error_400", "raise"), "http_open"],
563 [("http_error_302", "return response"), "http_error_303",
564 "http_error"],
565 [("http_error_302")],
566 ]
567 handlers = add_ordered_mock_handlers(o, meth_spec)
568
569 class Unknown:
570 def __eq__(self, other): return True
571
572 req = Request("http://example.com/")
573 r = o.open(req)
574 assert len(o.calls) == 2
575 calls = [(handlers[0], "http_open", (req,)),
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000576 (handlers[2], "http_error_302",
577 (req, Unknown(), 302, "", {}))]
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000578 for expected, got in zip(calls, o.calls):
579 handler, method_name, args = expected
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000580 self.assertEqual((handler, method_name), got[:2])
581 self.assertEqual(args, got[2])
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000582
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700583
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000584 def test_processors(self):
585 # *_request / *_response methods get called appropriately
586 o = OpenerDirector()
587 meth_spec = [
588 [("http_request", "return request"),
589 ("http_response", "return response")],
590 [("http_request", "return request"),
591 ("http_response", "return response")],
592 ]
593 handlers = add_ordered_mock_handlers(o, meth_spec)
594
595 req = Request("http://example.com/")
596 r = o.open(req)
597 # processor methods are called on *all* handlers that define them,
598 # not just the first handler that handles the request
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000599 calls = [
600 (handlers[0], "http_request"), (handlers[1], "http_request"),
601 (handlers[0], "http_response"), (handlers[1], "http_response")]
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000602
603 for i, (handler, name, args, kwds) in enumerate(o.calls):
604 if i < 2:
605 # *_request
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000606 self.assertEqual((handler, name), calls[i])
607 self.assertEqual(len(args), 1)
Ezio Melottie9615932010-01-24 19:26:24 +0000608 self.assertIsInstance(args[0], Request)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000609 else:
610 # *_response
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000611 self.assertEqual((handler, name), calls[i])
612 self.assertEqual(len(args), 2)
Ezio Melottie9615932010-01-24 19:26:24 +0000613 self.assertIsInstance(args[0], Request)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000614 # response from opener.open is None, because there's no
615 # handler that defines http_open to handle it
Benjamin Petersonc9c0f202009-06-30 23:06:06 +0000616 self.assertTrue(args[1] is None or
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000617 isinstance(args[1], MockResponse))
618
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700619 def test_method_deprecations(self):
620 req = Request("http://www.example.com")
Senthil Kumaran08bd4aa2012-04-11 23:05:49 +0800621
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800622 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700623 req.add_data("data")
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800624 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700625 req.has_data()
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800626 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700627 req.get_data()
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800628 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700629 req.get_host()
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800630 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700631 req.get_selector()
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800632 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700633 req.is_unverifiable()
Senthil Kumaran80a133b2012-04-12 19:28:07 +0800634 with self.assertWarns(DeprecationWarning):
Senthil Kumaran38b968b2012-03-14 13:43:53 -0700635 req.get_origin_req_host()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000636
Senthil Kumaran08bd4aa2012-04-11 23:05:49 +0800637
Tim Peters58eb11c2004-01-18 20:29:55 +0000638def sanepathname2url(path):
Victor Stinner6c6f8512010-08-07 10:09:35 +0000639 try:
Marc-André Lemburg8f36af72011-02-25 15:42:01 +0000640 path.encode("utf-8")
Victor Stinner6c6f8512010-08-07 10:09:35 +0000641 except UnicodeEncodeError:
642 raise unittest.SkipTest("path is not encodable to utf8")
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000643 urlpath = urllib.request.pathname2url(path)
Tim Peters58eb11c2004-01-18 20:29:55 +0000644 if os.name == "nt" and urlpath.startswith("///"):
645 urlpath = urlpath[2:]
646 # XXX don't ask me about the mac...
647 return urlpath
648
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000649class HandlerTests(unittest.TestCase):
650
651 def test_ftp(self):
652 class MockFTPWrapper:
653 def __init__(self, data): self.data = data
654 def retrfile(self, filename, filetype):
655 self.filename, self.filetype = filename, filetype
Guido van Rossum34d19282007-08-09 01:03:29 +0000656 return io.StringIO(self.data), len(self.data)
Nadeem Vawda08f5f7a2011-07-23 14:03:00 +0200657 def close(self): pass
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000658
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000659 class NullFTPHandler(urllib.request.FTPHandler):
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000660 def __init__(self, data): self.data = data
Georg Brandlf78e02b2008-06-10 17:40:04 +0000661 def connect_ftp(self, user, passwd, host, port, dirs,
662 timeout=socket._GLOBAL_DEFAULT_TIMEOUT):
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000663 self.user, self.passwd = user, passwd
664 self.host, self.port = host, port
665 self.dirs = dirs
666 self.ftpwrapper = MockFTPWrapper(self.data)
667 return self.ftpwrapper
668
Georg Brandlf78e02b2008-06-10 17:40:04 +0000669 import ftplib
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000670 data = "rheum rhaponicum"
671 h = NullFTPHandler(data)
672 o = h.parent = MockOpener()
673
Senthil Kumarandaa29d02010-11-18 15:36:41 +0000674 for url, host, port, user, passwd, type_, dirs, filename, mimetype in [
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000675 ("ftp://localhost/foo/bar/baz.html",
Senthil Kumarandaa29d02010-11-18 15:36:41 +0000676 "localhost", ftplib.FTP_PORT, "", "", "I",
677 ["foo", "bar"], "baz.html", "text/html"),
678 ("ftp://parrot@localhost/foo/bar/baz.html",
679 "localhost", ftplib.FTP_PORT, "parrot", "", "I",
680 ["foo", "bar"], "baz.html", "text/html"),
681 ("ftp://%25parrot@localhost/foo/bar/baz.html",
682 "localhost", ftplib.FTP_PORT, "%parrot", "", "I",
683 ["foo", "bar"], "baz.html", "text/html"),
684 ("ftp://%2542parrot@localhost/foo/bar/baz.html",
685 "localhost", ftplib.FTP_PORT, "%42parrot", "", "I",
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000686 ["foo", "bar"], "baz.html", "text/html"),
Kurt B. Kaiser3f7cb5d2004-07-11 17:14:13 +0000687 ("ftp://localhost:80/foo/bar/",
Senthil Kumarandaa29d02010-11-18 15:36:41 +0000688 "localhost", 80, "", "", "D",
Kurt B. Kaiser3f7cb5d2004-07-11 17:14:13 +0000689 ["foo", "bar"], "", None),
690 ("ftp://localhost/baz.gif;type=a",
Senthil Kumarandaa29d02010-11-18 15:36:41 +0000691 "localhost", ftplib.FTP_PORT, "", "", "A",
Kurt B. Kaiser3f7cb5d2004-07-11 17:14:13 +0000692 [], "baz.gif", None), # XXX really this should guess image/gif
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000693 ]:
Guido van Rossumcd16bf62007-06-13 18:07:49 +0000694 req = Request(url)
695 req.timeout = None
696 r = h.ftp_open(req)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000697 # ftp authentication not yet implemented by FTPHandler
Senthil Kumarandaa29d02010-11-18 15:36:41 +0000698 self.assertEqual(h.user, user)
699 self.assertEqual(h.passwd, passwd)
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000700 self.assertEqual(h.host, socket.gethostbyname(host))
701 self.assertEqual(h.port, port)
702 self.assertEqual(h.dirs, dirs)
703 self.assertEqual(h.ftpwrapper.filename, filename)
704 self.assertEqual(h.ftpwrapper.filetype, type_)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000705 headers = r.info()
Kurt B. Kaiser3f7cb5d2004-07-11 17:14:13 +0000706 self.assertEqual(headers.get("Content-type"), mimetype)
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000707 self.assertEqual(int(headers["Content-length"]), len(data))
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000708
709 def test_file(self):
Benjamin Petersona0c0a4a2008-06-12 22:15:50 +0000710 import email.utils, socket
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000711 h = urllib.request.FileHandler()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000712 o = h.parent = MockOpener()
713
Benjamin Petersonee8712c2008-05-20 21:35:26 +0000714 TESTFN = support.TESTFN
Tim Peters58eb11c2004-01-18 20:29:55 +0000715 urlpath = sanepathname2url(os.path.abspath(TESTFN))
Guido van Rossum6a2ccd02007-07-16 20:51:57 +0000716 towrite = b"hello, world\n"
Thomas Wouters49fd7fa2006-04-21 10:40:58 +0000717 urls = [
Tim Peters58eb11c2004-01-18 20:29:55 +0000718 "file://localhost%s" % urlpath,
719 "file://%s" % urlpath,
720 "file://%s%s" % (socket.gethostbyname('localhost'), urlpath),
Thomas Wouters49fd7fa2006-04-21 10:40:58 +0000721 ]
722 try:
723 localaddr = socket.gethostbyname(socket.gethostname())
724 except socket.gaierror:
725 localaddr = ''
726 if localaddr:
727 urls.append("file://%s%s" % (localaddr, urlpath))
728
729 for url in urls:
Tim Peters58eb11c2004-01-18 20:29:55 +0000730 f = open(TESTFN, "wb")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000731 try:
732 try:
733 f.write(towrite)
734 finally:
735 f.close()
736
737 r = h.file_open(Request(url))
738 try:
739 data = r.read()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000740 headers = r.info()
Senthil Kumaran4fbed102010-05-08 03:29:09 +0000741 respurl = r.geturl()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000742 finally:
743 r.close()
Tim Peters58eb11c2004-01-18 20:29:55 +0000744 stats = os.stat(TESTFN)
Benjamin Petersona0c0a4a2008-06-12 22:15:50 +0000745 modified = email.utils.formatdate(stats.st_mtime, usegmt=True)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000746 finally:
747 os.remove(TESTFN)
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000748 self.assertEqual(data, towrite)
749 self.assertEqual(headers["Content-type"], "text/plain")
750 self.assertEqual(headers["Content-length"], "13")
Tim Peters58eb11c2004-01-18 20:29:55 +0000751 self.assertEqual(headers["Last-modified"], modified)
Senthil Kumaran4fbed102010-05-08 03:29:09 +0000752 self.assertEqual(respurl, url)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000753
754 for url in [
Tim Peters58eb11c2004-01-18 20:29:55 +0000755 "file://localhost:80%s" % urlpath,
Guido van Rossumd8faa362007-04-27 19:54:29 +0000756 "file:///file_does_not_exist.txt",
757 "file://%s:80%s/%s" % (socket.gethostbyname('localhost'),
758 os.getcwd(), TESTFN),
759 "file://somerandomhost.ontheinternet.com%s/%s" %
760 (os.getcwd(), TESTFN),
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000761 ]:
762 try:
Tim Peters58eb11c2004-01-18 20:29:55 +0000763 f = open(TESTFN, "wb")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000764 try:
765 f.write(towrite)
766 finally:
767 f.close()
768
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000769 self.assertRaises(urllib.error.URLError,
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000770 h.file_open, Request(url))
771 finally:
772 os.remove(TESTFN)
773
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000774 h = urllib.request.FileHandler()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000775 o = h.parent = MockOpener()
776 # XXXX why does // mean ftp (and /// mean not ftp!), and where
777 # is file: scheme specified? I think this is really a bug, and
778 # what was intended was to distinguish between URLs like:
779 # file:/blah.txt (a file)
780 # file://localhost/blah.txt (a file)
781 # file:///blah.txt (a file)
782 # file://ftp.example.com/blah.txt (an ftp URL)
783 for url, ftp in [
Senthil Kumaran383c32d2010-10-14 11:57:35 +0000784 ("file://ftp.example.com//foo.txt", False),
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000785 ("file://ftp.example.com///foo.txt", False),
786# XXXX bug: fails with OSError, should be URLError
787 ("file://ftp.example.com/foo.txt", False),
Senthil Kumaran383c32d2010-10-14 11:57:35 +0000788 ("file://somehost//foo/something.txt", False),
Senthil Kumaran2ef16322010-07-11 03:12:43 +0000789 ("file://localhost//foo/something.txt", False),
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000790 ]:
791 req = Request(url)
792 try:
793 h.file_open(req)
794 # XXXX remove OSError when bug fixed
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000795 except (urllib.error.URLError, OSError):
Florent Xicluna419e3842010-08-08 16:16:07 +0000796 self.assertFalse(ftp)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000797 else:
Florent Xicluna419e3842010-08-08 16:16:07 +0000798 self.assertIs(o.req, req)
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000799 self.assertEqual(req.type, "ftp")
Łukasz Langad7e81cc2011-01-09 18:18:53 +0000800 self.assertEqual(req.type == "ftp", ftp)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000801
802 def test_http(self):
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000803
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000804 h = urllib.request.AbstractHTTPHandler()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000805 o = h.parent = MockOpener()
806
807 url = "http://example.com/"
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000808 for method, data in [("GET", None), ("POST", b"blah")]:
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000809 req = Request(url, data, {"Foo": "bar"})
Guido van Rossumcd16bf62007-06-13 18:07:49 +0000810 req.timeout = None
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000811 req.add_unredirected_header("Spam", "eggs")
812 http = MockHTTPClass()
813 r = h.do_open(http, req)
814
815 # result attributes
816 r.read; r.readline # wrapped MockFile methods
817 r.info; r.geturl # addinfourl methods
818 r.code, r.msg == 200, "OK" # added from MockHTTPClass.getreply()
819 hdrs = r.info()
Guido van Rossume2b70bc2006-08-18 22:13:04 +0000820 hdrs.get; hdrs.__contains__ # r.info() gives dict from .getreply()
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000821 self.assertEqual(r.geturl(), url)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000822
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000823 self.assertEqual(http.host, "example.com")
824 self.assertEqual(http.level, 0)
825 self.assertEqual(http.method, method)
826 self.assertEqual(http.selector, "/")
827 self.assertEqual(http.req_headers,
Jeremy Hyltonb3ee6f92004-02-24 19:40:35 +0000828 [("Connection", "close"),
829 ("Foo", "bar"), ("Spam", "eggs")])
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000830 self.assertEqual(http.data, data)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000831
832 # check socket.error converted to URLError
833 http.raise_on_endheaders = True
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000834 self.assertRaises(urllib.error.URLError, h.do_open, http, req)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000835
Senthil Kumaran29333122011-02-11 11:25:47 +0000836 # Check for TypeError on POST data which is str.
837 req = Request("http://example.com/","badpost")
838 self.assertRaises(TypeError, h.do_request_, req)
839
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000840 # check adding of standard headers
841 o.addheaders = [("Spam", "eggs")]
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000842 for data in b"", None: # POST, GET
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000843 req = Request("http://example.com/", data)
844 r = MockResponse(200, "OK", {}, "")
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000845 newreq = h.do_request_(req)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000846 if data is None: # GET
Benjamin Peterson577473f2010-01-19 00:09:57 +0000847 self.assertNotIn("Content-length", req.unredirected_hdrs)
848 self.assertNotIn("Content-type", req.unredirected_hdrs)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000849 else: # POST
Thomas Wouters00ee7ba2006-08-21 19:07:27 +0000850 self.assertEqual(req.unredirected_hdrs["Content-length"], "0")
851 self.assertEqual(req.unredirected_hdrs["Content-type"],
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000852 "application/x-www-form-urlencoded")
853 # XXX the details of Host could be better tested
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000854 self.assertEqual(req.unredirected_hdrs["Host"], "example.com")
855 self.assertEqual(req.unredirected_hdrs["Spam"], "eggs")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000856
857 # don't clobber existing headers
858 req.add_unredirected_header("Content-length", "foo")
859 req.add_unredirected_header("Content-type", "bar")
860 req.add_unredirected_header("Host", "baz")
861 req.add_unredirected_header("Spam", "foo")
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000862 newreq = h.do_request_(req)
Thomas Wouters00ee7ba2006-08-21 19:07:27 +0000863 self.assertEqual(req.unredirected_hdrs["Content-length"], "foo")
864 self.assertEqual(req.unredirected_hdrs["Content-type"], "bar")
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000865 self.assertEqual(req.unredirected_hdrs["Host"], "baz")
866 self.assertEqual(req.unredirected_hdrs["Spam"], "foo")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000867
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000868 # Check iterable body support
869 def iterable_body():
870 yield b"one"
871 yield b"two"
872 yield b"three"
873
874 for headers in {}, {"Content-Length": 11}:
875 req = Request("http://example.com/", iterable_body(), headers)
876 if not headers:
877 # Having an iterable body without a Content-Length should
878 # raise an exception
879 self.assertRaises(ValueError, h.do_request_, req)
880 else:
881 newreq = h.do_request_(req)
882
Senthil Kumaran29333122011-02-11 11:25:47 +0000883 # A file object.
884 # Test only Content-Length attribute of request.
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000885
Senthil Kumaran29333122011-02-11 11:25:47 +0000886 file_obj = io.BytesIO()
887 file_obj.write(b"Something\nSomething\nSomething\n")
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000888
889 for headers in {}, {"Content-Length": 30}:
890 req = Request("http://example.com/", file_obj, headers)
891 if not headers:
892 # Having an iterable body without a Content-Length should
893 # raise an exception
894 self.assertRaises(ValueError, h.do_request_, req)
895 else:
896 newreq = h.do_request_(req)
897 self.assertEqual(int(newreq.get_header('Content-length')),30)
898
899 file_obj.close()
900
901 # array.array Iterable - Content Length is calculated
902
903 iterable_array = array.array("I",[1,2,3,4])
904
905 for headers in {}, {"Content-Length": 16}:
906 req = Request("http://example.com/", iterable_array, headers)
907 newreq = h.do_request_(req)
908 self.assertEqual(int(newreq.get_header('Content-length')),16)
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000909
Facundo Batista72dc1ea2008-08-16 14:44:32 +0000910 def test_http_doubleslash(self):
911 # Checks the presence of any unnecessary double slash in url does not
912 # break anything. Previously, a double slash directly after the host
Ezio Melottie130a522011-10-19 10:58:56 +0300913 # could cause incorrect parsing.
Facundo Batista72dc1ea2008-08-16 14:44:32 +0000914 h = urllib.request.AbstractHTTPHandler()
915 o = h.parent = MockOpener()
916
Senthil Kumaran7bc0d872010-12-19 10:49:52 +0000917 data = b""
Facundo Batista72dc1ea2008-08-16 14:44:32 +0000918 ds_urls = [
919 "http://example.com/foo/bar/baz.html",
920 "http://example.com//foo/bar/baz.html",
921 "http://example.com/foo//bar/baz.html",
922 "http://example.com/foo/bar//baz.html"
923 ]
924
925 for ds_url in ds_urls:
926 ds_req = Request(ds_url, data)
927
928 # Check whether host is determined correctly if there is no proxy
929 np_ds_req = h.do_request_(ds_req)
930 self.assertEqual(np_ds_req.unredirected_hdrs["Host"],"example.com")
931
932 # Check whether host is determined correctly if there is a proxy
933 ds_req.set_proxy("someproxy:3128",None)
934 p_ds_req = h.do_request_(ds_req)
935 self.assertEqual(p_ds_req.unredirected_hdrs["Host"],"example.com")
936
Senthil Kumaranc2958622010-11-22 04:48:26 +0000937 def test_fixpath_in_weirdurls(self):
938 # Issue4493: urllib2 to supply '/' when to urls where path does not
939 # start with'/'
940
941 h = urllib.request.AbstractHTTPHandler()
942 o = h.parent = MockOpener()
943
944 weird_url = 'http://www.python.org?getspam'
945 req = Request(weird_url)
946 newreq = h.do_request_(req)
947 self.assertEqual(newreq.host,'www.python.org')
948 self.assertEqual(newreq.selector,'/?getspam')
949
950 url_without_path = 'http://www.python.org'
951 req = Request(url_without_path)
952 newreq = h.do_request_(req)
953 self.assertEqual(newreq.host,'www.python.org')
954 self.assertEqual(newreq.selector,'')
955
Facundo Batista72dc1ea2008-08-16 14:44:32 +0000956
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000957 def test_errors(self):
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000958 h = urllib.request.HTTPErrorProcessor()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000959 o = h.parent = MockOpener()
960
961 url = "http://example.com/"
962 req = Request(url)
Guido van Rossumd8faa362007-04-27 19:54:29 +0000963 # all 2xx are passed through
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000964 r = MockResponse(200, "OK", {}, "", url)
965 newr = h.http_response(req, r)
Florent Xicluna419e3842010-08-08 16:16:07 +0000966 self.assertIs(r, newr)
967 self.assertFalse(hasattr(o, "proto")) # o.error not called
Guido van Rossumd8faa362007-04-27 19:54:29 +0000968 r = MockResponse(202, "Accepted", {}, "", url)
969 newr = h.http_response(req, r)
Florent Xicluna419e3842010-08-08 16:16:07 +0000970 self.assertIs(r, newr)
971 self.assertFalse(hasattr(o, "proto")) # o.error not called
Guido van Rossumd8faa362007-04-27 19:54:29 +0000972 r = MockResponse(206, "Partial content", {}, "", url)
973 newr = h.http_response(req, r)
Florent Xicluna419e3842010-08-08 16:16:07 +0000974 self.assertIs(r, newr)
975 self.assertFalse(hasattr(o, "proto")) # o.error not called
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000976 # anything else calls o.error (and MockOpener returns None, here)
Guido van Rossumd8faa362007-04-27 19:54:29 +0000977 r = MockResponse(502, "Bad gateway", {}, "", url)
Florent Xicluna419e3842010-08-08 16:16:07 +0000978 self.assertIsNone(h.http_response(req, r))
Jeremy Hyltondf38ea92003-12-17 20:42:38 +0000979 self.assertEqual(o.proto, "http") # o.error called
Guido van Rossumd8faa362007-04-27 19:54:29 +0000980 self.assertEqual(o.args, (req, r, 502, "Bad gateway", {}))
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000981
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000982 def test_cookies(self):
983 cj = MockCookieJar()
Jeremy Hylton1afc1692008-06-18 20:49:58 +0000984 h = urllib.request.HTTPCookieProcessor(cj)
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000985 o = h.parent = MockOpener()
986
987 req = Request("http://example.com/")
988 r = MockResponse(200, "OK", {}, "")
989 newreq = h.http_request(req)
Florent Xicluna419e3842010-08-08 16:16:07 +0000990 self.assertIs(cj.ach_req, req)
991 self.assertIs(cj.ach_req, newreq)
992 self.assertEqual(req.get_origin_req_host(), "example.com")
993 self.assertFalse(req.is_unverifiable())
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000994 newr = h.http_response(req, r)
Florent Xicluna419e3842010-08-08 16:16:07 +0000995 self.assertIs(cj.ec_req, req)
996 self.assertIs(cj.ec_r, r)
997 self.assertIs(r, newr)
Martin v. Löwis2a6ba902004-05-31 18:22:40 +0000998
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +0000999 def test_redirect(self):
1000 from_url = "http://example.com/a.html"
1001 to_url = "http://example.com/b.html"
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001002 h = urllib.request.HTTPRedirectHandler()
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001003 o = h.parent = MockOpener()
1004
1005 # ordinary redirect behaviour
1006 for code in 301, 302, 303, 307:
1007 for data in None, "blah\nblah\n":
1008 method = getattr(h, "http_error_%s" % code)
1009 req = Request(from_url, data)
Senthil Kumaranfb8cc2f2009-07-19 02:44:19 +00001010 req.timeout = socket._GLOBAL_DEFAULT_TIMEOUT
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001011 req.add_header("Nonsense", "viking=withhold")
Christian Heimes77c02eb2008-02-09 02:18:51 +00001012 if data is not None:
1013 req.add_header("Content-Length", str(len(data)))
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001014 req.add_unredirected_header("Spam", "spam")
1015 try:
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001016 method(req, MockFile(), code, "Blah",
1017 MockHeaders({"location": to_url}))
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001018 except urllib.error.HTTPError:
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001019 # 307 in response to POST requires user OK
Benjamin Petersonc9c0f202009-06-30 23:06:06 +00001020 self.assertTrue(code == 307 and data is not None)
Jeremy Hyltondf38ea92003-12-17 20:42:38 +00001021 self.assertEqual(o.req.get_full_url(), to_url)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001022 try:
Jeremy Hyltondf38ea92003-12-17 20:42:38 +00001023 self.assertEqual(o.req.get_method(), "GET")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001024 except AttributeError:
Florent Xicluna419e3842010-08-08 16:16:07 +00001025 self.assertFalse(o.req.has_data())
Christian Heimes77c02eb2008-02-09 02:18:51 +00001026
1027 # now it's a GET, there should not be headers regarding content
1028 # (possibly dragged from before being a POST)
1029 headers = [x.lower() for x in o.req.headers]
Benjamin Peterson577473f2010-01-19 00:09:57 +00001030 self.assertNotIn("content-length", headers)
1031 self.assertNotIn("content-type", headers)
Christian Heimes77c02eb2008-02-09 02:18:51 +00001032
Jeremy Hyltondf38ea92003-12-17 20:42:38 +00001033 self.assertEqual(o.req.headers["Nonsense"],
1034 "viking=withhold")
Benjamin Peterson577473f2010-01-19 00:09:57 +00001035 self.assertNotIn("Spam", o.req.headers)
1036 self.assertNotIn("Spam", o.req.unredirected_hdrs)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001037
1038 # loop detection
1039 req = Request(from_url)
Senthil Kumaranfb8cc2f2009-07-19 02:44:19 +00001040 req.timeout = socket._GLOBAL_DEFAULT_TIMEOUT
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001041 def redirect(h, req, url=to_url):
1042 h.http_error_302(req, MockFile(), 302, "Blah",
1043 MockHeaders({"location": url}))
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001044 # Note that the *original* request shares the same record of
1045 # redirections with the sub-requests caused by the redirections.
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001046
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001047 # detect infinite loop redirect of a URL to itself
1048 req = Request(from_url, origin_req_host="example.com")
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001049 count = 0
Senthil Kumaranfb8cc2f2009-07-19 02:44:19 +00001050 req.timeout = socket._GLOBAL_DEFAULT_TIMEOUT
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001051 try:
1052 while 1:
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001053 redirect(h, req, "http://example.com/")
1054 count = count + 1
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001055 except urllib.error.HTTPError:
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001056 # don't stop until max_repeats, because cookies may introduce state
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001057 self.assertEqual(count, urllib.request.HTTPRedirectHandler.max_repeats)
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001058
1059 # detect endless non-repeating chain of redirects
1060 req = Request(from_url, origin_req_host="example.com")
1061 count = 0
Senthil Kumaranfb8cc2f2009-07-19 02:44:19 +00001062 req.timeout = socket._GLOBAL_DEFAULT_TIMEOUT
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001063 try:
1064 while 1:
1065 redirect(h, req, "http://example.com/%d" % count)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001066 count = count + 1
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001067 except urllib.error.HTTPError:
Jeremy Hyltondf38ea92003-12-17 20:42:38 +00001068 self.assertEqual(count,
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001069 urllib.request.HTTPRedirectHandler.max_redirections)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001070
guido@google.coma119df92011-03-29 11:41:02 -07001071
1072 def test_invalid_redirect(self):
1073 from_url = "http://example.com/a.html"
1074 valid_schemes = ['http','https','ftp']
1075 invalid_schemes = ['file','imap','ldap']
1076 schemeless_url = "example.com/b.html"
1077 h = urllib.request.HTTPRedirectHandler()
1078 o = h.parent = MockOpener()
1079 req = Request(from_url)
1080 req.timeout = socket._GLOBAL_DEFAULT_TIMEOUT
1081
1082 for scheme in invalid_schemes:
1083 invalid_url = scheme + '://' + schemeless_url
1084 self.assertRaises(urllib.error.HTTPError, h.http_error_302,
1085 req, MockFile(), 302, "Security Loophole",
1086 MockHeaders({"location": invalid_url}))
1087
1088 for scheme in valid_schemes:
1089 valid_url = scheme + '://' + schemeless_url
1090 h.http_error_302(req, MockFile(), 302, "That's fine",
1091 MockHeaders({"location": valid_url}))
1092 self.assertEqual(o.req.get_full_url(), valid_url)
1093
Senthil Kumaran6497aa32012-01-04 13:46:59 +08001094 def test_relative_redirect(self):
1095 from_url = "http://example.com/a.html"
1096 relative_url = "/b.html"
1097 h = urllib.request.HTTPRedirectHandler()
1098 o = h.parent = MockOpener()
1099 req = Request(from_url)
1100 req.timeout = socket._GLOBAL_DEFAULT_TIMEOUT
1101
1102 valid_url = urllib.parse.urljoin(from_url,relative_url)
1103 h.http_error_302(req, MockFile(), 302, "That's fine",
1104 MockHeaders({"location": valid_url}))
1105 self.assertEqual(o.req.get_full_url(), valid_url)
1106
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001107 def test_cookie_redirect(self):
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001108 # cookies shouldn't leak into redirected requests
Georg Brandl24420152008-05-26 16:32:26 +00001109 from http.cookiejar import CookieJar
1110 from test.test_http_cookiejar import interact_netscape
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001111
1112 cj = CookieJar()
1113 interact_netscape(cj, "http://www.example.com/", "spam=eggs")
Thomas Wouters477c8d52006-05-27 19:21:47 +00001114 hh = MockHTTPHandler(302, "Location: http://www.cracker.com/\r\n\r\n")
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001115 hdeh = urllib.request.HTTPDefaultErrorHandler()
1116 hrh = urllib.request.HTTPRedirectHandler()
1117 cp = urllib.request.HTTPCookieProcessor(cj)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001118 o = build_test_opener(hh, hdeh, hrh, cp)
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001119 o.open("http://www.example.com/")
Florent Xicluna419e3842010-08-08 16:16:07 +00001120 self.assertFalse(hh.req.has_header("Cookie"))
Martin v. Löwis2a6ba902004-05-31 18:22:40 +00001121
Senthil Kumaran26430412011-04-13 07:01:19 +08001122 def test_redirect_fragment(self):
1123 redirected_url = 'http://www.example.com/index.html#OK\r\n\r\n'
1124 hh = MockHTTPHandler(302, 'Location: ' + redirected_url)
1125 hdeh = urllib.request.HTTPDefaultErrorHandler()
1126 hrh = urllib.request.HTTPRedirectHandler()
1127 o = build_test_opener(hh, hdeh, hrh)
1128 fp = o.open('http://www.example.com')
1129 self.assertEqual(fp.geturl(), redirected_url.strip())
1130
Thomas Wouters49fd7fa2006-04-21 10:40:58 +00001131 def test_proxy(self):
1132 o = OpenerDirector()
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001133 ph = urllib.request.ProxyHandler(dict(http="proxy.example.com:3128"))
Thomas Wouters49fd7fa2006-04-21 10:40:58 +00001134 o.add_handler(ph)
1135 meth_spec = [
1136 [("http_open", "return response")]
1137 ]
1138 handlers = add_ordered_mock_handlers(o, meth_spec)
1139
1140 req = Request("http://acme.example.com/")
1141 self.assertEqual(req.get_host(), "acme.example.com")
1142 r = o.open(req)
1143 self.assertEqual(req.get_host(), "proxy.example.com:3128")
1144
1145 self.assertEqual([(handlers[0], "http_open")],
1146 [tup[0:2] for tup in o.calls])
1147
Senthil Kumaran7bb04972009-10-11 04:58:55 +00001148 def test_proxy_no_proxy(self):
1149 os.environ['no_proxy'] = 'python.org'
1150 o = OpenerDirector()
1151 ph = urllib.request.ProxyHandler(dict(http="proxy.example.com"))
1152 o.add_handler(ph)
1153 req = Request("http://www.perl.org/")
1154 self.assertEqual(req.get_host(), "www.perl.org")
1155 r = o.open(req)
1156 self.assertEqual(req.get_host(), "proxy.example.com")
1157 req = Request("http://www.python.org")
1158 self.assertEqual(req.get_host(), "www.python.org")
1159 r = o.open(req)
1160 self.assertEqual(req.get_host(), "www.python.org")
1161 del os.environ['no_proxy']
1162
Ronald Oussorene72e1612011-03-14 18:15:25 -04001163 def test_proxy_no_proxy_all(self):
1164 os.environ['no_proxy'] = '*'
1165 o = OpenerDirector()
1166 ph = urllib.request.ProxyHandler(dict(http="proxy.example.com"))
1167 o.add_handler(ph)
1168 req = Request("http://www.python.org")
1169 self.assertEqual(req.get_host(), "www.python.org")
1170 r = o.open(req)
1171 self.assertEqual(req.get_host(), "www.python.org")
1172 del os.environ['no_proxy']
1173
Senthil Kumaran7bb04972009-10-11 04:58:55 +00001174
Senthil Kumaran97f0c6b2009-07-25 04:24:38 +00001175 def test_proxy_https(self):
1176 o = OpenerDirector()
1177 ph = urllib.request.ProxyHandler(dict(https="proxy.example.com:3128"))
1178 o.add_handler(ph)
1179 meth_spec = [
1180 [("https_open", "return response")]
1181 ]
1182 handlers = add_ordered_mock_handlers(o, meth_spec)
1183
1184 req = Request("https://www.example.com/")
1185 self.assertEqual(req.get_host(), "www.example.com")
1186 r = o.open(req)
1187 self.assertEqual(req.get_host(), "proxy.example.com:3128")
1188 self.assertEqual([(handlers[0], "https_open")],
1189 [tup[0:2] for tup in o.calls])
1190
Senthil Kumaran47fff872009-12-20 07:10:31 +00001191 def test_proxy_https_proxy_authorization(self):
1192 o = OpenerDirector()
1193 ph = urllib.request.ProxyHandler(dict(https='proxy.example.com:3128'))
1194 o.add_handler(ph)
1195 https_handler = MockHTTPSHandler()
1196 o.add_handler(https_handler)
1197 req = Request("https://www.example.com/")
1198 req.add_header("Proxy-Authorization","FooBar")
1199 req.add_header("User-Agent","Grail")
1200 self.assertEqual(req.get_host(), "www.example.com")
1201 self.assertIsNone(req._tunnel_host)
1202 r = o.open(req)
1203 # Verify Proxy-Authorization gets tunneled to request.
1204 # httpsconn req_headers do not have the Proxy-Authorization header but
1205 # the req will have.
Ezio Melottib58e0bd2010-01-23 15:40:09 +00001206 self.assertNotIn(("Proxy-Authorization","FooBar"),
Senthil Kumaran47fff872009-12-20 07:10:31 +00001207 https_handler.httpconn.req_headers)
Ezio Melottib58e0bd2010-01-23 15:40:09 +00001208 self.assertIn(("User-Agent","Grail"),
1209 https_handler.httpconn.req_headers)
Senthil Kumaran47fff872009-12-20 07:10:31 +00001210 self.assertIsNotNone(req._tunnel_host)
1211 self.assertEqual(req.get_host(), "proxy.example.com:3128")
1212 self.assertEqual(req.get_header("Proxy-authorization"),"FooBar")
Senthil Kumaran97f0c6b2009-07-25 04:24:38 +00001213
Senthil Kumaran4de00a22011-05-11 21:17:57 +08001214 # TODO: This should be only for OSX
1215 @unittest.skipUnless(sys.platform == 'darwin', "only relevant for OSX")
Ronald Oussorene72e1612011-03-14 18:15:25 -04001216 def test_osx_proxy_bypass(self):
1217 bypass = {
1218 'exclude_simple': False,
1219 'exceptions': ['foo.bar', '*.bar.com', '127.0.0.1', '10.10',
1220 '10.0/16']
1221 }
1222 # Check hosts that should trigger the proxy bypass
1223 for host in ('foo.bar', 'www.bar.com', '127.0.0.1', '10.10.0.1',
1224 '10.0.0.1'):
1225 self.assertTrue(_proxy_bypass_macosx_sysconf(host, bypass),
1226 'expected bypass of %s to be True' % host)
1227 # Check hosts that should not trigger the proxy bypass
1228 for host in ('abc.foo.bar', 'bar.com', '127.0.0.2', '10.11.0.1', 'test'):
1229 self.assertFalse(_proxy_bypass_macosx_sysconf(host, bypass),
1230 'expected bypass of %s to be False' % host)
1231
1232 # Check the exclude_simple flag
1233 bypass = {'exclude_simple': True, 'exceptions': []}
1234 self.assertTrue(_proxy_bypass_macosx_sysconf('test', bypass))
1235
Christian Heimes4fbc72b2008-03-22 00:47:35 +00001236 def test_basic_auth(self, quote_char='"'):
Thomas Wouters477c8d52006-05-27 19:21:47 +00001237 opener = OpenerDirector()
1238 password_manager = MockPasswordManager()
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001239 auth_handler = urllib.request.HTTPBasicAuthHandler(password_manager)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001240 realm = "ACME Widget Store"
1241 http_handler = MockHTTPHandler(
Christian Heimes4fbc72b2008-03-22 00:47:35 +00001242 401, 'WWW-Authenticate: Basic realm=%s%s%s\r\n\r\n' %
1243 (quote_char, realm, quote_char) )
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001244 opener.add_handler(auth_handler)
1245 opener.add_handler(http_handler)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001246 self._test_basic_auth(opener, auth_handler, "Authorization",
1247 realm, http_handler, password_manager,
1248 "http://acme.example.com/protected",
1249 "http://acme.example.com/protected",
1250 )
1251
Christian Heimes4fbc72b2008-03-22 00:47:35 +00001252 def test_basic_auth_with_single_quoted_realm(self):
1253 self.test_basic_auth(quote_char="'")
1254
Senthil Kumaran34f3fcc2012-05-15 22:30:25 +08001255 def test_basic_auth_with_unquoted_realm(self):
1256 opener = OpenerDirector()
1257 password_manager = MockPasswordManager()
1258 auth_handler = urllib.request.HTTPBasicAuthHandler(password_manager)
1259 realm = "ACME Widget Store"
1260 http_handler = MockHTTPHandler(
1261 401, 'WWW-Authenticate: Basic realm=%s\r\n\r\n' % realm)
1262 opener.add_handler(auth_handler)
1263 opener.add_handler(http_handler)
Senthil Kumaran0ea91cb2012-05-15 23:59:42 +08001264 with self.assertWarns(UserWarning):
1265 self._test_basic_auth(opener, auth_handler, "Authorization",
1266 realm, http_handler, password_manager,
1267 "http://acme.example.com/protected",
1268 "http://acme.example.com/protected",
1269 )
Senthil Kumaran34f3fcc2012-05-15 22:30:25 +08001270
Thomas Wouters477c8d52006-05-27 19:21:47 +00001271 def test_proxy_basic_auth(self):
1272 opener = OpenerDirector()
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001273 ph = urllib.request.ProxyHandler(dict(http="proxy.example.com:3128"))
Thomas Wouters477c8d52006-05-27 19:21:47 +00001274 opener.add_handler(ph)
1275 password_manager = MockPasswordManager()
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001276 auth_handler = urllib.request.ProxyBasicAuthHandler(password_manager)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001277 realm = "ACME Networks"
1278 http_handler = MockHTTPHandler(
1279 407, 'Proxy-Authenticate: Basic realm="%s"\r\n\r\n' % realm)
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001280 opener.add_handler(auth_handler)
1281 opener.add_handler(http_handler)
Thomas Wouters00ee7ba2006-08-21 19:07:27 +00001282 self._test_basic_auth(opener, auth_handler, "Proxy-authorization",
Thomas Wouters477c8d52006-05-27 19:21:47 +00001283 realm, http_handler, password_manager,
1284 "http://acme.example.com:3128/protected",
1285 "proxy.example.com:3128",
1286 )
1287
1288 def test_basic_and_digest_auth_handlers(self):
1289 # HTTPDigestAuthHandler threw an exception if it couldn't handle a 40*
1290 # response (http://python.org/sf/1479302), where it should instead
1291 # return None to allow another handler (especially
1292 # HTTPBasicAuthHandler) to handle the response.
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001293
1294 # Also (http://python.org/sf/14797027, RFC 2617 section 1.2), we must
1295 # try digest first (since it's the strongest auth scheme), so we record
1296 # order of calls here to check digest comes first:
1297 class RecordingOpenerDirector(OpenerDirector):
1298 def __init__(self):
1299 OpenerDirector.__init__(self)
1300 self.recorded = []
1301 def record(self, info):
1302 self.recorded.append(info)
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001303 class TestDigestAuthHandler(urllib.request.HTTPDigestAuthHandler):
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001304 def http_error_401(self, *args, **kwds):
1305 self.parent.record("digest")
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001306 urllib.request.HTTPDigestAuthHandler.http_error_401(self,
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001307 *args, **kwds)
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001308 class TestBasicAuthHandler(urllib.request.HTTPBasicAuthHandler):
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001309 def http_error_401(self, *args, **kwds):
1310 self.parent.record("basic")
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001311 urllib.request.HTTPBasicAuthHandler.http_error_401(self,
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001312 *args, **kwds)
1313
1314 opener = RecordingOpenerDirector()
Thomas Wouters477c8d52006-05-27 19:21:47 +00001315 password_manager = MockPasswordManager()
1316 digest_handler = TestDigestAuthHandler(password_manager)
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001317 basic_handler = TestBasicAuthHandler(password_manager)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001318 realm = "ACME Networks"
1319 http_handler = MockHTTPHandler(
1320 401, 'WWW-Authenticate: Basic realm="%s"\r\n\r\n' % realm)
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001321 opener.add_handler(basic_handler)
1322 opener.add_handler(digest_handler)
1323 opener.add_handler(http_handler)
1324
1325 # check basic auth isn't blocked by digest handler failing
Thomas Wouters477c8d52006-05-27 19:21:47 +00001326 self._test_basic_auth(opener, basic_handler, "Authorization",
1327 realm, http_handler, password_manager,
1328 "http://acme.example.com/protected",
1329 "http://acme.example.com/protected",
1330 )
Thomas Wouters4d70c3d2006-06-08 14:42:34 +00001331 # check digest was tried before basic (twice, because
1332 # _test_basic_auth called .open() twice)
1333 self.assertEqual(opener.recorded, ["digest", "basic"]*2)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001334
Senthil Kumaran4de00a22011-05-11 21:17:57 +08001335 def test_unsupported_auth_digest_handler(self):
1336 opener = OpenerDirector()
1337 # While using DigestAuthHandler
1338 digest_auth_handler = urllib.request.HTTPDigestAuthHandler(None)
1339 http_handler = MockHTTPHandler(
1340 401, 'WWW-Authenticate: Kerberos\r\n\r\n')
1341 opener.add_handler(digest_auth_handler)
1342 opener.add_handler(http_handler)
1343 self.assertRaises(ValueError,opener.open,"http://www.example.com")
1344
1345 def test_unsupported_auth_basic_handler(self):
1346 # While using BasicAuthHandler
1347 opener = OpenerDirector()
1348 basic_auth_handler = urllib.request.HTTPBasicAuthHandler(None)
1349 http_handler = MockHTTPHandler(
1350 401, 'WWW-Authenticate: NTLM\r\n\r\n')
1351 opener.add_handler(basic_auth_handler)
1352 opener.add_handler(http_handler)
1353 self.assertRaises(ValueError,opener.open,"http://www.example.com")
1354
Thomas Wouters477c8d52006-05-27 19:21:47 +00001355 def _test_basic_auth(self, opener, auth_handler, auth_header,
1356 realm, http_handler, password_manager,
1357 request_url, protected_url):
Christian Heimes05e8be12008-02-23 18:30:17 +00001358 import base64
Thomas Wouters477c8d52006-05-27 19:21:47 +00001359 user, password = "wile", "coyote"
Thomas Wouters477c8d52006-05-27 19:21:47 +00001360
1361 # .add_password() fed through to password manager
1362 auth_handler.add_password(realm, request_url, user, password)
1363 self.assertEqual(realm, password_manager.realm)
1364 self.assertEqual(request_url, password_manager.url)
1365 self.assertEqual(user, password_manager.user)
1366 self.assertEqual(password, password_manager.password)
1367
1368 r = opener.open(request_url)
1369
1370 # should have asked the password manager for the username/password
1371 self.assertEqual(password_manager.target_realm, realm)
1372 self.assertEqual(password_manager.target_url, protected_url)
1373
1374 # expect one request without authorization, then one with
1375 self.assertEqual(len(http_handler.requests), 2)
1376 self.assertFalse(http_handler.requests[0].has_header(auth_header))
Guido van Rossum98b349f2007-08-27 21:47:52 +00001377 userpass = bytes('%s:%s' % (user, password), "ascii")
Guido van Rossum98297ee2007-11-06 21:34:58 +00001378 auth_hdr_value = ('Basic ' +
Georg Brandl706824f2009-06-04 09:42:55 +00001379 base64.encodebytes(userpass).strip().decode())
Thomas Wouters477c8d52006-05-27 19:21:47 +00001380 self.assertEqual(http_handler.requests[1].get_header(auth_header),
1381 auth_hdr_value)
Senthil Kumaranca2fc9e2010-02-24 16:53:16 +00001382 self.assertEqual(http_handler.requests[1].unredirected_hdrs[auth_header],
1383 auth_hdr_value)
Thomas Wouters477c8d52006-05-27 19:21:47 +00001384 # if the password manager can't find a password, the handler won't
1385 # handle the HTTP auth error
1386 password_manager.user = password_manager.password = None
1387 http_handler.reset()
1388 r = opener.open(request_url)
1389 self.assertEqual(len(http_handler.requests), 1)
1390 self.assertFalse(http_handler.requests[0].has_header(auth_header))
1391
Senthil Kumaran4de00a22011-05-11 21:17:57 +08001392
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001393class MiscTests(unittest.TestCase):
1394
1395 def test_build_opener(self):
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001396 class MyHTTPHandler(urllib.request.HTTPHandler): pass
1397 class FooHandler(urllib.request.BaseHandler):
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001398 def foo_open(self): pass
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001399 class BarHandler(urllib.request.BaseHandler):
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001400 def bar_open(self): pass
1401
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001402 build_opener = urllib.request.build_opener
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001403
1404 o = build_opener(FooHandler, BarHandler)
1405 self.opener_has_handler(o, FooHandler)
1406 self.opener_has_handler(o, BarHandler)
1407
1408 # can take a mix of classes and instances
1409 o = build_opener(FooHandler, BarHandler())
1410 self.opener_has_handler(o, FooHandler)
1411 self.opener_has_handler(o, BarHandler)
1412
1413 # subclasses of default handlers override default handlers
1414 o = build_opener(MyHTTPHandler)
1415 self.opener_has_handler(o, MyHTTPHandler)
1416
1417 # a particular case of overriding: default handlers can be passed
1418 # in explicitly
1419 o = build_opener()
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001420 self.opener_has_handler(o, urllib.request.HTTPHandler)
1421 o = build_opener(urllib.request.HTTPHandler)
1422 self.opener_has_handler(o, urllib.request.HTTPHandler)
1423 o = build_opener(urllib.request.HTTPHandler())
1424 self.opener_has_handler(o, urllib.request.HTTPHandler)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001425
Christian Heimes81ee3ef2008-05-04 22:42:01 +00001426 # Issue2670: multiple handlers sharing the same base class
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001427 class MyOtherHTTPHandler(urllib.request.HTTPHandler): pass
Christian Heimes81ee3ef2008-05-04 22:42:01 +00001428 o = build_opener(MyHTTPHandler, MyOtherHTTPHandler)
1429 self.opener_has_handler(o, MyHTTPHandler)
1430 self.opener_has_handler(o, MyOtherHTTPHandler)
1431
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001432 def opener_has_handler(self, opener, handler_class):
Florent Xicluna419e3842010-08-08 16:16:07 +00001433 self.assertTrue(any(h.__class__ == handler_class
1434 for h in opener.handlers))
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001435
Benjamin Peterson6ebe78f2008-12-21 00:06:59 +00001436class RequestTests(unittest.TestCase):
1437
1438 def setUp(self):
1439 self.get = Request("http://www.python.org/~jeremy/")
1440 self.post = Request("http://www.python.org/~jeremy/",
1441 "data",
1442 headers={"X-Test": "test"})
1443
1444 def test_method(self):
1445 self.assertEqual("POST", self.post.get_method())
1446 self.assertEqual("GET", self.get.get_method())
1447
1448 def test_add_data(self):
Florent Xicluna419e3842010-08-08 16:16:07 +00001449 self.assertFalse(self.get.has_data())
Benjamin Peterson6ebe78f2008-12-21 00:06:59 +00001450 self.assertEqual("GET", self.get.get_method())
1451 self.get.add_data("spam")
Benjamin Petersonc9c0f202009-06-30 23:06:06 +00001452 self.assertTrue(self.get.has_data())
Benjamin Peterson6ebe78f2008-12-21 00:06:59 +00001453 self.assertEqual("POST", self.get.get_method())
1454
1455 def test_get_full_url(self):
1456 self.assertEqual("http://www.python.org/~jeremy/",
1457 self.get.get_full_url())
1458
1459 def test_selector(self):
1460 self.assertEqual("/~jeremy/", self.get.get_selector())
1461 req = Request("http://www.python.org/")
1462 self.assertEqual("/", req.get_selector())
1463
1464 def test_get_type(self):
1465 self.assertEqual("http", self.get.get_type())
1466
1467 def test_get_host(self):
1468 self.assertEqual("www.python.org", self.get.get_host())
1469
1470 def test_get_host_unquote(self):
1471 req = Request("http://www.%70ython.org/")
1472 self.assertEqual("www.python.org", req.get_host())
1473
1474 def test_proxy(self):
Florent Xicluna419e3842010-08-08 16:16:07 +00001475 self.assertFalse(self.get.has_proxy())
Benjamin Peterson6ebe78f2008-12-21 00:06:59 +00001476 self.get.set_proxy("www.perl.org", "http")
Benjamin Petersonc9c0f202009-06-30 23:06:06 +00001477 self.assertTrue(self.get.has_proxy())
Benjamin Peterson6ebe78f2008-12-21 00:06:59 +00001478 self.assertEqual("www.python.org", self.get.get_origin_req_host())
1479 self.assertEqual("www.perl.org", self.get.get_host())
1480
Senthil Kumarand95cc752010-08-08 11:27:53 +00001481 def test_wrapped_url(self):
1482 req = Request("<URL:http://www.python.org>")
1483 self.assertEqual("www.python.org", req.get_host())
1484
Senthil Kumaran26430412011-04-13 07:01:19 +08001485 def test_url_fragment(self):
Senthil Kumarand95cc752010-08-08 11:27:53 +00001486 req = Request("http://www.python.org/?qs=query#fragment=true")
1487 self.assertEqual("/?qs=query", req.get_selector())
1488 req = Request("http://www.python.org/#fun=true")
1489 self.assertEqual("/", req.get_selector())
1490
Senthil Kumaran26430412011-04-13 07:01:19 +08001491 # Issue 11703: geturl() omits fragment in the original URL.
1492 url = 'http://docs.python.org/library/urllib2.html#OK'
1493 req = Request(url)
1494 self.assertEqual(req.get_full_url(), url)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001495
Jason R. Coombsaa204db2011-11-07 10:50:32 -05001496def test_HTTPError_interface():
1497 """
1498 Issue 13211 reveals that HTTPError didn't implement the URLError
1499 interface even though HTTPError is a subclass of URLError.
1500
Jason R. Coombs7ff21d72011-12-03 23:18:11 -05001501 >>> msg = 'something bad happened'
1502 >>> url = code = hdrs = fp = None
Jason R. Coombs9c3895f2011-12-04 08:14:18 -05001503 >>> err = urllib.error.HTTPError(url, code, msg, hdrs, fp)
Jason R. Coombsaa204db2011-11-07 10:50:32 -05001504 >>> assert hasattr(err, 'reason')
1505 >>> err.reason
1506 'something bad happened'
1507 """
1508
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001509def test_main(verbose=None):
Thomas Wouters477c8d52006-05-27 19:21:47 +00001510 from test import test_urllib2
Benjamin Petersonee8712c2008-05-20 21:35:26 +00001511 support.run_doctest(test_urllib2, verbose)
Jeremy Hylton1afc1692008-06-18 20:49:58 +00001512 support.run_doctest(urllib.request, verbose)
Andrew M. Kuchlingbd3200f2004-06-29 13:15:46 +00001513 tests = (TrivialTests,
1514 OpenerDirectorTests,
1515 HandlerTests,
Benjamin Peterson6ebe78f2008-12-21 00:06:59 +00001516 MiscTests,
1517 RequestTests)
Benjamin Petersonee8712c2008-05-20 21:35:26 +00001518 support.run_unittest(*tests)
Jeremy Hyltonc1be59f2003-12-14 05:27:34 +00001519
1520if __name__ == "__main__":
1521 test_main(verbose=True)