Lib/test/test_urlparse.py

   1 #! /usr/bin/env python
   2
   3 from test import test_support
   4 import unittest
   5 import urlparse
   6
   7 RFC1808_BASE = "http://a/b/c/d;p?q#f"
   8 RFC2396_BASE = "http://a/b/c/d;p?q"
   9
  10 class UrlParseTestCase(unittest.TestCase):
  11     def test_frags(self):
  12         for url, expected in [('http://www.python.org',
  13                                ('http', 'www.python.org', '', '', '', '')),
  14                               ('http://www.python.org#abc',
  15                                ('http', 'www.python.org', '', '', '', 'abc')),
  16                               ('http://www.python.org/#abc',
  17                                ('http', 'www.python.org', '/', '', '', 'abc')),
  18                               (RFC1808_BASE,
  19                                ('http', 'a', '/b/c/d', 'p', 'q', 'f')),
  20                               ('file:///tmp/junk.txt',
  21                                ('file', '', '/tmp/junk.txt', '', '', '')),
  22                               ]:
  23             result = urlparse.urlparse(url)
  24             self.assertEqual(result, expected)
  25             # put it back together and it should be the same
  26             result2 = urlparse.urlunparse(result)
  27             self.assertEqual(result2, url)
  28
  29     def checkJoin(self, base, relurl, expected):
  30         self.assertEqual(urlparse.urljoin(base, relurl), expected)
  31
  32     def test_RFC1808(self):
  33         # "normal" cases from RFC 1808:
  34         self.checkJoin(RFC1808_BASE, 'g:h', 'g:h')
  35         self.checkJoin(RFC1808_BASE, 'g', 'http://a/b/c/g')
  36         self.checkJoin(RFC1808_BASE, './g', 'http://a/b/c/g')
  37         self.checkJoin(RFC1808_BASE, 'g/', 'http://a/b/c/g/')
  38         self.checkJoin(RFC1808_BASE, '/g', 'http://a/g')
  39         self.checkJoin(RFC1808_BASE, '//g', 'http://g')
  40         self.checkJoin(RFC1808_BASE, '?y', 'http://a/b/c/d;p?y')
  41         self.checkJoin(RFC1808_BASE, 'g?y', 'http://a/b/c/g?y')
  42         self.checkJoin(RFC1808_BASE, 'g?y/./x', 'http://a/b/c/g?y/./x')
  43         self.checkJoin(RFC1808_BASE, '#s', 'http://a/b/c/d;p?q#s')
  44         self.checkJoin(RFC1808_BASE, 'g#s', 'http://a/b/c/g#s')
  45         self.checkJoin(RFC1808_BASE, 'g#s/./x', 'http://a/b/c/g#s/./x')
  46         self.checkJoin(RFC1808_BASE, 'g?y#s', 'http://a/b/c/g?y#s')
  47         self.checkJoin(RFC1808_BASE, ';x', 'http://a/b/c/d;x')
  48         self.checkJoin(RFC1808_BASE, 'g;x', 'http://a/b/c/g;x')
  49         self.checkJoin(RFC1808_BASE, 'g;x?y#s', 'http://a/b/c/g;x?y#s')
  50         self.checkJoin(RFC1808_BASE, '.', 'http://a/b/c/')
  51         self.checkJoin(RFC1808_BASE, './', 'http://a/b/c/')
  52         self.checkJoin(RFC1808_BASE, '..', 'http://a/b/')
  53         self.checkJoin(RFC1808_BASE, '../', 'http://a/b/')
  54         self.checkJoin(RFC1808_BASE, '../g', 'http://a/b/g')
  55         self.checkJoin(RFC1808_BASE, '../..', 'http://a/')
  56         self.checkJoin(RFC1808_BASE, '../../', 'http://a/')
  57         self.checkJoin(RFC1808_BASE, '../../g', 'http://a/g')
  58
  59         # "abnormal" cases from RFC 1808:
  60         self.checkJoin(RFC1808_BASE, '', 'http://a/b/c/d;p?q#f')
  61         self.checkJoin(RFC1808_BASE, '../../../g', 'http://a/../g')
  62         self.checkJoin(RFC1808_BASE, '../../../../g', 'http://a/../../g')
  63         self.checkJoin(RFC1808_BASE, '/./g', 'http://a/./g')
  64         self.checkJoin(RFC1808_BASE, '/../g', 'http://a/../g')
  65         self.checkJoin(RFC1808_BASE, 'g.', 'http://a/b/c/g.')
  66         self.checkJoin(RFC1808_BASE, '.g', 'http://a/b/c/.g')
  67         self.checkJoin(RFC1808_BASE, 'g..', 'http://a/b/c/g..')
  68         self.checkJoin(RFC1808_BASE, '..g', 'http://a/b/c/..g')
  69         self.checkJoin(RFC1808_BASE, './../g', 'http://a/b/g')
  70         self.checkJoin(RFC1808_BASE, './g/.', 'http://a/b/c/g/')
  71         self.checkJoin(RFC1808_BASE, 'g/./h', 'http://a/b/c/g/h')
  72         self.checkJoin(RFC1808_BASE, 'g/../h', 'http://a/b/c/h')
  73
  74         # RFC 1808 and RFC 1630 disagree on these (according to RFC 1808),
  75         # so we'll not actually run these tests (which expect 1808 behavior).
  76         #self.checkJoin(RFC1808_BASE, 'http:g', 'http:g')
  77         #self.checkJoin(RFC1808_BASE, 'http:', 'http:')
  78
  79     def test_RFC2396(self):
  80         # cases from RFC 2396
  81
  82         ### urlparse.py as of v 1.32 fails on these two
  83         #self.checkJoin(RFC2396_BASE, '?y', 'http://a/b/c/?y')
  84         #self.checkJoin(RFC2396_BASE, ';x', 'http://a/b/c/;x')
  85
  86         self.checkJoin(RFC2396_BASE, 'g:h', 'g:h')
  87         self.checkJoin(RFC2396_BASE, 'g', 'http://a/b/c/g')
  88         self.checkJoin(RFC2396_BASE, './g', 'http://a/b/c/g')
  89         self.checkJoin(RFC2396_BASE, 'g/', 'http://a/b/c/g/')
  90         self.checkJoin(RFC2396_BASE, '/g', 'http://a/g')
  91         self.checkJoin(RFC2396_BASE, '//g', 'http://g')
  92         self.checkJoin(RFC2396_BASE, 'g?y', 'http://a/b/c/g?y')
  93         self.checkJoin(RFC2396_BASE, '#s', 'http://a/b/c/d;p?q#s')
  94         self.checkJoin(RFC2396_BASE, 'g#s', 'http://a/b/c/g#s')
  95         self.checkJoin(RFC2396_BASE, 'g?y#s', 'http://a/b/c/g?y#s')
  96         self.checkJoin(RFC2396_BASE, 'g;x', 'http://a/b/c/g;x')
  97         self.checkJoin(RFC2396_BASE, 'g;x?y#s', 'http://a/b/c/g;x?y#s')
  98         self.checkJoin(RFC2396_BASE, '.', 'http://a/b/c/')
  99         self.checkJoin(RFC2396_BASE, './', 'http://a/b/c/')
 100         self.checkJoin(RFC2396_BASE, '..', 'http://a/b/')
 101         self.checkJoin(RFC2396_BASE, '../', 'http://a/b/')
 102         self.checkJoin(RFC2396_BASE, '../g', 'http://a/b/g')
 103         self.checkJoin(RFC2396_BASE, '../..', 'http://a/')
 104         self.checkJoin(RFC2396_BASE, '../../', 'http://a/')
 105         self.checkJoin(RFC2396_BASE, '../../g', 'http://a/g')
 106         self.checkJoin(RFC2396_BASE, '', RFC2396_BASE)
 107         self.checkJoin(RFC2396_BASE, '../../../g', 'http://a/../g')
 108         self.checkJoin(RFC2396_BASE, '../../../../g', 'http://a/../../g')
 109         self.checkJoin(RFC2396_BASE, '/./g', 'http://a/./g')
 110         self.checkJoin(RFC2396_BASE, '/../g', 'http://a/../g')
 111         self.checkJoin(RFC2396_BASE, 'g.', 'http://a/b/c/g.')
 112         self.checkJoin(RFC2396_BASE, '.g', 'http://a/b/c/.g')
 113         self.checkJoin(RFC2396_BASE, 'g..', 'http://a/b/c/g..')
 114         self.checkJoin(RFC2396_BASE, '..g', 'http://a/b/c/..g')
 115         self.checkJoin(RFC2396_BASE, './../g', 'http://a/b/g')
 116         self.checkJoin(RFC2396_BASE, './g/.', 'http://a/b/c/g/')
 117         self.checkJoin(RFC2396_BASE, 'g/./h', 'http://a/b/c/g/h')
 118         self.checkJoin(RFC2396_BASE, 'g/../h', 'http://a/b/c/h')
 119         self.checkJoin(RFC2396_BASE, 'g;x=1/./y', 'http://a/b/c/g;x=1/y')
 120         self.checkJoin(RFC2396_BASE, 'g;x=1/../y', 'http://a/b/c/y')
 121         self.checkJoin(RFC2396_BASE, 'g?y/./x', 'http://a/b/c/g?y/./x')
 122         self.checkJoin(RFC2396_BASE, 'g?y/../x', 'http://a/b/c/g?y/../x')
 123         self.checkJoin(RFC2396_BASE, 'g#s/./x', 'http://a/b/c/g#s/./x')
 124         self.checkJoin(RFC2396_BASE, 'g#s/../x', 'http://a/b/c/g#s/../x')
 125
 126 def test_main():
 127     test_support.run_unittest(UrlParseTestCase)
 128
 129 if __name__ == "__main__":
 130     test_main()