mirror of https://github.com/python/cpython
192 lines
7.5 KiB
Python
192 lines
7.5 KiB
Python
#!/usr/bin/env python3
|
|
|
|
import unittest
|
|
from test import support
|
|
|
|
import contextlib
|
|
import socket
|
|
import urllib.request
|
|
import sys
|
|
import os
|
|
import email.message
|
|
import time
|
|
|
|
|
|
class URLTimeoutTest(unittest.TestCase):
|
|
# XXX this test doesn't seem to test anything useful.
|
|
|
|
TIMEOUT = 30.0
|
|
|
|
def setUp(self):
|
|
socket.setdefaulttimeout(self.TIMEOUT)
|
|
|
|
def tearDown(self):
|
|
socket.setdefaulttimeout(None)
|
|
|
|
def testURLread(self):
|
|
with support.transient_internet("www.python.org"):
|
|
f = urllib.request.urlopen("http://www.python.org/")
|
|
x = f.read()
|
|
|
|
|
|
class urlopenNetworkTests(unittest.TestCase):
|
|
"""Tests urllib.reqest.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.
|
|
|
|
setUp is not used for always constructing a connection to
|
|
http://www.python.org/ since there a few tests that don't use that address
|
|
and making a connection is expensive enough to warrant minimizing unneeded
|
|
connections.
|
|
|
|
"""
|
|
|
|
@contextlib.contextmanager
|
|
def urlopen(self, *args, **kwargs):
|
|
resource = args[0]
|
|
with support.transient_internet(resource):
|
|
r = urllib.request.urlopen(*args, **kwargs)
|
|
try:
|
|
yield r
|
|
finally:
|
|
r.close()
|
|
|
|
def test_basic(self):
|
|
# Simple test expected to pass.
|
|
with self.urlopen("http://www.python.org/") as open_url:
|
|
for attr in ("read", "readline", "readlines", "fileno", "close",
|
|
"info", "geturl"):
|
|
self.assertTrue(hasattr(open_url, attr), "object returned from "
|
|
"urlopen lacks the %s attribute" % attr)
|
|
self.assertTrue(open_url.read(), "calling 'read' failed")
|
|
|
|
def test_readlines(self):
|
|
# Test both readline and readlines.
|
|
with self.urlopen("http://www.python.org/") as open_url:
|
|
self.assertIsInstance(open_url.readline(), bytes,
|
|
"readline did not return a string")
|
|
self.assertIsInstance(open_url.readlines(), list,
|
|
"readlines did not return a list")
|
|
|
|
def test_info(self):
|
|
# Test 'info'.
|
|
with self.urlopen("http://www.python.org/") as open_url:
|
|
info_obj = open_url.info()
|
|
self.assertIsInstance(info_obj, email.message.Message,
|
|
"object returned by 'info' is not an "
|
|
"instance of email.message.Message")
|
|
self.assertEqual(info_obj.get_content_subtype(), "html")
|
|
|
|
def test_geturl(self):
|
|
# Make sure same URL as opened is returned by geturl.
|
|
URL = "http://www.python.org/"
|
|
with self.urlopen(URL) as open_url:
|
|
gotten_url = open_url.geturl()
|
|
self.assertEqual(gotten_url, URL)
|
|
|
|
def test_getcode(self):
|
|
# test getcode() with the fancy opener to get 404 error codes
|
|
URL = "http://www.python.org/XXXinvalidXXX"
|
|
with support.transient_internet(URL):
|
|
open_url = urllib.request.FancyURLopener().open(URL)
|
|
try:
|
|
code = open_url.getcode()
|
|
finally:
|
|
open_url.close()
|
|
self.assertEqual(code, 404)
|
|
|
|
def test_fileno(self):
|
|
if sys.platform in ('win32',):
|
|
# On Windows, socket handles are not file descriptors; this
|
|
# test can't pass on Windows.
|
|
return
|
|
# Make sure fd returned by fileno is valid.
|
|
with self.urlopen("http://www.python.org/", timeout=None) as open_url:
|
|
fd = open_url.fileno()
|
|
with os.fdopen(fd, encoding='utf-8') as f:
|
|
self.assertTrue(f.read(), "reading from file created using fd "
|
|
"returned by fileno failed")
|
|
|
|
def test_bad_address(self):
|
|
# Make sure proper exception is raised when connecting to a bogus
|
|
# address.
|
|
bogus_domain = "sadflkjsasf.i.nvali.d"
|
|
try:
|
|
socket.gethostbyname(bogus_domain)
|
|
except socket.gaierror:
|
|
pass
|
|
else:
|
|
# This happens with some overzealous DNS providers such as OpenDNS
|
|
self.skipTest("%r should not resolve for test to work" % bogus_domain)
|
|
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.urlopen, "http://www.sadflkjsasadf.com/")
|
|
urllib.request.urlopen,
|
|
"http://sadflkjsasf.i.nvali.d/")
|
|
|
|
|
|
class urlretrieveNetworkTests(unittest.TestCase):
|
|
"""Tests urllib.request.urlretrieve using the network."""
|
|
|
|
@contextlib.contextmanager
|
|
def urlretrieve(self, *args):
|
|
resource = args[0]
|
|
with support.transient_internet(resource):
|
|
file_location, info = urllib.request.urlretrieve(*args)
|
|
try:
|
|
yield file_location, info
|
|
finally:
|
|
support.unlink(file_location)
|
|
|
|
def test_basic(self):
|
|
# Test basic functionality.
|
|
with self.urlretrieve("http://www.python.org/") as (file_location, info):
|
|
self.assertTrue(os.path.exists(file_location), "file location returned by"
|
|
" urlretrieve is not a valid path")
|
|
with open(file_location, encoding='utf-8') as f:
|
|
self.assertTrue(f.read(), "reading from the file location returned"
|
|
" by urlretrieve failed")
|
|
|
|
def test_specified_path(self):
|
|
# Make sure that specifying the location of the file to write to works.
|
|
with self.urlretrieve("http://www.python.org/",
|
|
support.TESTFN) as (file_location, info):
|
|
self.assertEqual(file_location, support.TESTFN)
|
|
self.assertTrue(os.path.exists(file_location))
|
|
with open(file_location, encoding='utf-8') as f:
|
|
self.assertTrue(f.read(), "reading from temporary file failed")
|
|
|
|
def test_header(self):
|
|
# Make sure header returned as 2nd value from urlretrieve is good.
|
|
with self.urlretrieve("http://www.python.org/") as (file_location, info):
|
|
self.assertIsInstance(info, email.message.Message,
|
|
"info is not an instance of email.message.Message")
|
|
|
|
def test_data_header(self):
|
|
logo = "http://www.python.org/community/logos/python-logo-master-v3-TM.png"
|
|
with self.urlretrieve(logo) as (file_location, fileheaders):
|
|
datevalue = fileheaders.get('Date')
|
|
dateformat = '%a, %d %b %Y %H:%M:%S GMT'
|
|
try:
|
|
time.strptime(datevalue, dateformat)
|
|
except ValueError:
|
|
self.fail('Date value not in %r format', dateformat)
|
|
|
|
|
|
def test_main():
|
|
support.requires('network')
|
|
support.run_unittest(URLTimeoutTest,
|
|
urlopenNetworkTests,
|
|
urlretrieveNetworkTests)
|
|
|
|
if __name__ == "__main__":
|
|
test_main()
|