Issue 9873: the URL parsing functions now accept ASCII encoded byte sequences in…

Issue 9873: the URL parsing functions now accept ASCII encoded byte sequences in addition to character strings

Doc/library/urllib.parse.rst

@@ -24,7 +24,15 @@ following URL schemes: ``file``, ``ftp``, ``gopher``, ``hdl``, ``http``,
 ``rsync``, ``rtsp``, ``rtspu``, ``sftp``, ``shttp``, ``sip``, ``sips``,
 ``snews``, ``svn``, ``svn+ssh``, ``telnet``, ``wais``.
 
-The :mod:`urllib.parse` module defines the following functions:
+The :mod:`urllib.parse` module defines functions that fall into two broad
+categories: URL parsing and URL quoting. These are covered in detail in
+the following sections.
+
+URL Parsing
+-----------
+
+The URL parsing functions focus on splitting a URL string into its components,
+or on combining URL components into a URL string.
 
 .. function:: urlparse(urlstring, scheme='', allow_fragments=True)
 
@@ -242,6 +250,161 @@ The :mod:`urllib.parse` module defines the following functions:
    string.  If there is no fragment identifier in *url*, return *url* unmodified
    and an empty string.
 
+   The return value is actually an instance of a subclass of :class:`tuple`.  This
+   class has the following additional read-only convenience attributes:
+
+   +------------------+-------+-------------------------+----------------------+
+   | Attribute        | Index | Value                   | Value if not present |
+   +==================+=======+=========================+======================+
+   | :attr:`url`      | 0     | URL with no fragment    | empty string         |
+   +------------------+-------+-------------------------+----------------------+
+   | :attr:`fragment` | 1     | Fragment identifier     | empty string         |
+   +------------------+-------+-------------------------+----------------------+
+
+   See section :ref:`urlparse-result-object` for more information on the result
+   object.
+
+   .. versionchanged:: 3.2
+      Result is a structured object rather than a simple 2-tuple
+
+
+Parsing ASCII Encoded Bytes
+---------------------------
+
+The URL parsing functions were originally designed to operate on character
+strings only. In practice, it is useful to be able to manipulate properly
+quoted and encoded URLs as sequences of ASCII bytes. Accordingly, the
+URL parsing functions in this module all operate on :class:`bytes` and
+:class:`bytearray` objects in addition to :class:`str` objects.
+
+If :class:`str` data is passed in, the result will also contain only
+:class:`str` data. If :class:`bytes` or :class:`bytearray` data is
+passed in, the result will contain only :class:`bytes` data.
+
+Attempting to mix :class:`str` data with :class:`bytes` or
+:class:`bytearray` in a single function call will result in a
+:exc:`TypeError` being thrown, while attempting to pass in non-ASCII
+byte values will trigger :exc:`UnicodeDecodeError`.
+
+To support easier conversion of result objects between :class:`str` and
+:class:`bytes`, all return values from URL parsing functions provide
+either an :meth:`encode` method (when the result contains :class:`str`
+data) or a :meth:`decode` method (when the result contains :class:`bytes`
+data). The signatures of these methods match those of the corresponding
+:class:`str` and :class:`bytes` methods (except that the default encoding
+is ``'ascii'`` rather than ``'utf-8'``). Each produces a value of a
+corresponding type that contains either :class:`bytes` data (for
+:meth:`encode` methods) or :class:`str` data (for
+:meth:`decode` methods).
+
+Applications that need to operate on potentially improperly quoted URLs
+that may contain non-ASCII data will need to do their own decoding from
+bytes to characters before invoking the URL parsing methods.
+
+The behaviour described in this section applies only to the URL parsing
+functions. The URL quoting functions use their own rules when producing
+or consuming byte sequences as detailed in the documentation of the
+individual URL quoting functions.
+
+.. versionchanged:: 3.2
+   URL parsing functions now accept ASCII encoded byte sequences
+
+
+.. _urlparse-result-object:
+
+Structured Parse Results
+------------------------
+
+The result objects from the :func:`urlparse`, :func:`urlsplit`  and
+:func:`urldefrag`functions are subclasses of the :class:`tuple` type.
+These subclasses add the attributes listed in the documentation for
+those functions, the encoding and decoding support described in the
+previous section, as well as an additional method:
+
+.. method:: urllib.parse.SplitResult.geturl()
+
+   Return the re-combined version of the original URL as a string. This may
+   differ from the original URL in that the scheme may be normalized to lower
+   case and empty components may be dropped. Specifically, empty parameters,
+   queries, and fragment identifiers will be removed.
+
+   For :func:`urldefrag` results, only empty fragment identifiers will be removed.
+   For :func:`urlsplit` and :func:`urlparse` results, all noted changes will be
+   made to the URL returned by this method.
+
+   The result of this method remains unchanged if passed back through the original
+   parsing function:
+
+      >>> from urllib.parse import urlsplit
+      >>> url = 'HTTP://www.Python.org/doc/#'
+      >>> r1 = urlsplit(url)
+      >>> r1.geturl()
+      'http://www.Python.org/doc/'
+      >>> r2 = urlsplit(r1.geturl())
+      >>> r2.geturl()
+      'http://www.Python.org/doc/'
+
+
+The following classes provide the implementations of the structured parse
+results when operating on :class:`str` objects:
+
+.. class:: DefragResult(url, fragment)
+
+   Concrete class for :func:`urldefrag` results containing :class:`str`
+   data. The :meth:`encode` method returns a :class:`DefragResultBytes`
+   instance.
+
+   .. versionadded:: 3.2
+
+.. class:: ParseResult(scheme, netloc, path, params, query, fragment)
+
+   Concrete class for :func:`urlparse` results containing :class:`str`
+   data. The :meth:`encode` method returns a :class:`ParseResultBytes`
+   instance.
+
+.. class:: SplitResult(scheme, netloc, path, query, fragment)
+
+   Concrete class for :func:`urlsplit` results containing :class:`str`
+   data. The :meth:`encode` method returns a :class:`SplitResultBytes`
+   instance.
+
+
+The following classes provide the implementations of the parse results when
+operating on :class:`bytes` or :class:`bytearray` objects:
+
+.. class:: DefragResultBytes(url, fragment)
+
+   Concrete class for :func:`urldefrag` results containing :class:`bytes`
+   data. The :meth:`decode` method returns a :class:`DefragResult`
+   instance.
+
+   .. versionadded:: 3.2
+
+.. class:: ParseResultBytes(scheme, netloc, path, params, query, fragment)
+
+   Concrete class for :func:`urlparse` results containing :class:`bytes`
+   data. The :meth:`decode` method returns a :class:`ParseResult`
+   instance.
+
+   .. versionadded:: 3.2
+
+.. class:: SplitResultBytes(scheme, netloc, path, query, fragment)
+
+   Concrete class for :func:`urlsplit` results containing :class:`bytes`
+   data. The :meth:`decode` method returns a :class:`SplitResult`
+   instance.
+
+   .. versionadded:: 3.2
+
+
+URL Quoting
+-----------
+
+The URL quoting functions focus on taking program data and making it safe
+for use as URL components by quoting special characters and appropriately
+encoding non-ASCII text. They also support reversing these operations to
+recreate the original data from the contents of a URL component if that
+task isn't already covered by the URL parsing functions above.
 
 .. function:: quote(string, safe='/', encoding=None, errors=None)
 
@@ -322,8 +485,7 @@ The :mod:`urllib.parse` module defines the following functions:
    If it is a :class:`str`, unescaped non-ASCII characters in *string*
    are encoded into UTF-8 bytes.
 
-   Example: ``unquote_to_bytes('a%26%EF')`` yields
-   ``b'a&\xef'``.
+   Example: ``unquote_to_bytes('a%26%EF')`` yields ``b'a&\xef'``.
 
 
 .. function:: urlencode(query, doseq=False, safe='', encoding=None, errors=None)
@@ -340,12 +502,13 @@ The :mod:`urllib.parse` module defines the following functions:
    the optional parameter *doseq* is evaluates to *True*, individual
    ``key=value`` pairs separated by ``'&'`` are generated for each element of
    the value sequence for the key.  The order of parameters in the encoded
-   string will match the order of parameter tuples in the sequence. This module
-   provides the functions :func:`parse_qs` and :func:`parse_qsl` which are used
-   to parse query strings into Python data structures.
+   string will match the order of parameter tuples in the sequence.
 
    When *query* parameter is a :class:`str`, the *safe*, *encoding* and *error*
-   parameters are sent the :func:`quote_plus` for encoding.
+   parameters are passed down to :func:`quote_plus` for encoding.
+
+   To reverse this encoding process, :func:`parse_qs` and :func:`parse_qsl` are
+   provided in this module to parse query strings into Python data structures.
 
    .. versionchanged:: 3.2
       Query parameter supports bytes and string objects.
@@ -376,57 +539,3 @@ The :mod:`urllib.parse` module defines the following functions:
 
    :rfc:`1738` - Uniform Resource Locators (URL)
       This specifies the formal syntax and semantics of absolute URLs.
-
-
-.. _urlparse-result-object:
-
-Results of :func:`urlparse` and :func:`urlsplit`
-------------------------------------------------
-
-The result objects from the :func:`urlparse` and :func:`urlsplit` functions are
-subclasses of the :class:`tuple` type.  These subclasses add the attributes
-described in those functions, as well as provide an additional method:
-
-.. method:: ParseResult.geturl()
-
-   Return the re-combined version of the original URL as a string. This may differ
-   from the original URL in that the scheme will always be normalized to lower case
-   and empty components may be dropped. Specifically, empty parameters, queries,
-   and fragment identifiers will be removed.
-
-   The result of this method is a fixpoint if passed back through the original
-   parsing function:
-
-      >>> import urllib.parse
-      >>> url = 'HTTP://www.Python.org/doc/#'
-
-      >>> r1 = urllib.parse.urlsplit(url)
-      >>> r1.geturl()
-      'http://www.Python.org/doc/'
-
-      >>> r2 = urllib.parse.urlsplit(r1.geturl())
-      >>> r2.geturl()
-      'http://www.Python.org/doc/'
-
-
-The following classes provide the implementations of the parse results:
-
-.. class:: BaseResult
-
-   Base class for the concrete result classes.  This provides most of the
-   attribute definitions.  It does not provide a :meth:`geturl` method.  It is
-   derived from :class:`tuple`, but does not override the :meth:`__init__` or
-   :meth:`__new__` methods.
-
-
-.. class:: ParseResult(scheme, netloc, path, params, query, fragment)
-
-   Concrete class for :func:`urlparse` results.  The :meth:`__new__` method is
-   overridden to support checking that the right number of arguments are passed.
-
-
-.. class:: SplitResult(scheme, netloc, path, query, fragment)
-
-   Concrete class for :func:`urlsplit` results.  The :meth:`__new__` method is
-   overridden to support checking that the right number of arguments are passed.
-

Doc/whatsnew/3.2.rst

@@ -573,6 +573,14 @@ New, Improved, and Deprecated Modules
   (Contributed by Rodolpho Eckhardt and Nick Coghlan, :issue:`10220`.)
 
 .. XXX: Mention inspect.getattr_static (Michael Foord)
+.. XXX: Mention urllib.parse changes
+          Issue 9873 (Nick Coghlan):
+            - ASCII byte sequence support in URL parsing
+            - named tuple for urldefrag return value
+          Issue 5468 (Dan Mahn) for urlencode:
+            - bytes input support
+            - non-UTF8 percent encoding of non-ASCII characters
+          Issue 2987 for IPv6 (RFC2732) support in urlparse
 
 Multi-threading
 ===============

Lib/test/test_urlparse.py

@@ -24,6 +24,17 @@ parse_qsl_test_cases = [
     ("&a=b", [('a', 'b')]),
     ("a=a+b&b=b+c", [('a', 'a b'), ('b', 'b c')]),
     ("a=1&a=2", [('a', '1'), ('a', '2')]),
+    (b"", []),
+    (b"&", []),
+    (b"&&", []),
+    (b"=", [(b'', b'')]),
+    (b"=a", [(b'', b'a')]),
+    (b"a", [(b'a', b'')]),
+    (b"a=", [(b'a', b'')]),
+    (b"a=", [(b'a', b'')]),
+    (b"&a=b", [(b'a', b'b')]),
+    (b"a=a+b&b=b+c", [(b'a', b'a b'), (b'b', b'b c')]),
+    (b"a=1&a=2", [(b'a', b'1'), (b'a', b'2')]),
 ]
 
 class UrlParseTestCase(unittest.TestCase):
@@ -86,7 +97,7 @@ class UrlParseTestCase(unittest.TestCase):
 
 
     def test_roundtrips(self):
-        testcases = [
+        str_cases = [
             ('file:///tmp/junk.txt',
              ('file', '', '/tmp/junk.txt', '', '', ''),
              ('file', '', '/tmp/junk.txt', '', '')),
@@ -110,16 +121,21 @@ class UrlParseTestCase(unittest.TestCase):
             ('git+ssh', 'git@github.com','/user/project.git',
              '','',''),
             ('git+ssh', 'git@github.com','/user/project.git',
-             '', ''))
+             '', '')),
             ]
-        for url, parsed, split in testcases:
+        def _encode(t):
+            return (t[0].encode('ascii'),
+                    tuple(x.encode('ascii') for x in t[1]),
+                    tuple(x.encode('ascii') for x in t[2]))
+        bytes_cases = [_encode(x) for x in str_cases]
+        for url, parsed, split in str_cases + bytes_cases:
             self.checkRoundtrips(url, parsed, split)
 
     def test_http_roundtrips(self):
         # urllib.parse.urlsplit treats 'http:' as an optimized special case,
         # so we test both 'http:' and 'https:' in all the following.
         # Three cheers for white box knowledge!
-        testcases = [
+        str_cases = [
             ('://www.python.org',
              ('www.python.org', '', '', '', ''),
              ('www.python.org', '', '', '')),
@@ -136,19 +152,34 @@ class UrlParseTestCase(unittest.TestCase):
              ('a', '/b/c/d', 'p', 'q', 'f'),
              ('a', '/b/c/d;p', 'q', 'f')),
             ]
-        for scheme in ('http', 'https'):
-            for url, parsed, split in testcases:
-                url = scheme + url
-                parsed = (scheme,) + parsed
-                split = (scheme,) + split
-                self.checkRoundtrips(url, parsed, split)
+        def _encode(t):
+            return (t[0].encode('ascii'),
+                    tuple(x.encode('ascii') for x in t[1]),
+                    tuple(x.encode('ascii') for x in t[2]))
+        bytes_cases = [_encode(x) for x in str_cases]
+        str_schemes = ('http', 'https')
+        bytes_schemes = (b'http', b'https')
+        str_tests = str_schemes, str_cases
+        bytes_tests = bytes_schemes, bytes_cases
+        for schemes, test_cases in (str_tests, bytes_tests):
+            for scheme in schemes:
+                for url, parsed, split in test_cases:
+                    url = scheme + url
+                    parsed = (scheme,) + parsed
+                    split = (scheme,) + split
+                    self.checkRoundtrips(url, parsed, split)
 
     def checkJoin(self, base, relurl, expected):
-        self.assertEqual(urllib.parse.urljoin(base, relurl), expected,
-                         (base, relurl, expected))
+        str_components = (base, relurl, expected)
+        self.assertEqual(urllib.parse.urljoin(base, relurl), expected)
+        bytes_components = baseb, relurlb, expectedb = [
+                            x.encode('ascii') for x in str_components]
+        self.assertEqual(urllib.parse.urljoin(baseb, relurlb), expectedb)
 
     def test_unparse_parse(self):
-        for u in ['Python', './Python','x-newscheme://foo.com/stuff','x://y','x:/y','x:/','/',]:
+        str_cases = ['Python', './Python','x-newscheme://foo.com/stuff','x://y','x:/y','x:/','/',]
+        bytes_cases = [x.encode('ascii') for x in str_cases]
+        for u in str_cases + bytes_cases:
             self.assertEqual(urllib.parse.urlunsplit(urllib.parse.urlsplit(u)), u)
             self.assertEqual(urllib.parse.urlunparse(urllib.parse.urlparse(u)), u)
 
@@ -328,7 +359,7 @@ class UrlParseTestCase(unittest.TestCase):
         self.checkJoin(SIMPLE_BASE, 'http:g?y/./x','http://a/b/c/g?y/./x')
 
     def test_RFC2732(self):
-        for url, hostname, port in [
+        str_cases = [
             ('http://Test.python.org:5432/foo/', 'test.python.org', 5432),
             ('http://12.34.56.78:5432/foo/', '12.34.56.78', 5432),
             ('http://[::1]:5432/foo/', '::1', 5432),
@@ -349,20 +380,26 @@ class UrlParseTestCase(unittest.TestCase):
             ('http://[::12.34.56.78]/foo/', '::12.34.56.78', None),
             ('http://[::ffff:12.34.56.78]/foo/',
              '::ffff:12.34.56.78', None),
-            ]:
+            ]
+        def _encode(t):
+            return t[0].encode('ascii'), t[1].encode('ascii'), t[2]
+        bytes_cases = [_encode(x) for x in str_cases]
+        for url, hostname, port in str_cases + bytes_cases:
             urlparsed = urllib.parse.urlparse(url)
             self.assertEqual((urlparsed.hostname, urlparsed.port) , (hostname, port))
 
-        for invalid_url in [
+        str_cases = [
                 'http://::12.34.56.78]/',
                 'http://[::1/foo/',
                 'ftp://[::1/foo/bad]/bad',
                 'http://[::1/foo/bad]/bad',
-                'http://[::ffff:12.34.56.78']:
+                'http://[::ffff:12.34.56.78']
+        bytes_cases = [x.encode('ascii') for x in str_cases]
+        for invalid_url in str_cases + bytes_cases:
             self.assertRaises(ValueError, urllib.parse.urlparse, invalid_url)
 
     def test_urldefrag(self):
-        for url, defrag, frag in [
+        str_cases = [
             ('http://python.org#frag', 'http://python.org', 'frag'),
             ('http://python.org', 'http://python.org', ''),
             ('http://python.org/#frag', 'http://python.org/', 'frag'),
@@ -373,8 +410,16 @@ class UrlParseTestCase(unittest.TestCase):
             ('http://python.org/p?q', 'http://python.org/p?q', ''),
             (RFC1808_BASE, 'http://a/b/c/d;p?q', 'f'),
             (RFC2396_BASE, 'http://a/b/c/d;p?q', ''),
-            ]:
-            self.assertEqual(urllib.parse.urldefrag(url), (defrag, frag))
+        ]
+        def _encode(t):
+            return type(t)(x.encode('ascii') for x in t)
+        bytes_cases = [_encode(x) for x in str_cases]
+        for url, defrag, frag in str_cases + bytes_cases:
+            result = urllib.parse.urldefrag(url)
+            self.assertEqual(result.geturl(), url)
+            self.assertEqual(result, (defrag, frag))
+            self.assertEqual(result.url, defrag)
+            self.assertEqual(result.fragment, frag)
 
     def test_urlsplit_attributes(self):
         url = "HTTP://WWW.PYTHON.ORG/doc/#frag"
@@ -390,7 +435,8 @@ class UrlParseTestCase(unittest.TestCase):
         self.assertEqual(p.port, None)
         # geturl() won't return exactly the original URL in this case
         # since the scheme is always case-normalized
-        #self.assertEqual(p.geturl(), url)
+        # We handle this by ignoring the first 4 characters of the URL
+        self.assertEqual(p.geturl()[4:], url[4:])
 
         url = "http://User:Pass@www.python.org:080/doc/?query=yes#frag"
         p = urllib.parse.urlsplit(url)
@@ -422,6 +468,45 @@ class UrlParseTestCase(unittest.TestCase):
         self.assertEqual(p.port, 80)
         self.assertEqual(p.geturl(), url)
 
+        # And check them all again, only with bytes this time
+        url = b"HTTP://WWW.PYTHON.ORG/doc/#frag"
+        p = urllib.parse.urlsplit(url)
+        self.assertEqual(p.scheme, b"http")
+        self.assertEqual(p.netloc, b"WWW.PYTHON.ORG")
+        self.assertEqual(p.path, b"/doc/")
+        self.assertEqual(p.query, b"")
+        self.assertEqual(p.fragment, b"frag")
+        self.assertEqual(p.username, None)
+        self.assertEqual(p.password, None)
+        self.assertEqual(p.hostname, b"www.python.org")
+        self.assertEqual(p.port, None)
+        self.assertEqual(p.geturl()[4:], url[4:])
+
+        url = b"http://User:Pass@www.python.org:080/doc/?query=yes#frag"
+        p = urllib.parse.urlsplit(url)
+        self.assertEqual(p.scheme, b"http")
+        self.assertEqual(p.netloc, b"User:Pass@www.python.org:080")
+        self.assertEqual(p.path, b"/doc/")
+        self.assertEqual(p.query, b"query=yes")
+        self.assertEqual(p.fragment, b"frag")
+        self.assertEqual(p.username, b"User")
+        self.assertEqual(p.password, b"Pass")
+        self.assertEqual(p.hostname, b"www.python.org")
+        self.assertEqual(p.port, 80)
+        self.assertEqual(p.geturl(), url)
+
+        url = b"http://User@example.com:Pass@www.python.org:080/doc/?query=yes#frag"
+        p = urllib.parse.urlsplit(url)
+        self.assertEqual(p.scheme, b"http")
+        self.assertEqual(p.netloc, b"User@example.com:Pass@www.python.org:080")
+        self.assertEqual(p.path, b"/doc/")
+        self.assertEqual(p.query, b"query=yes")
+        self.assertEqual(p.fragment, b"frag")
+        self.assertEqual(p.username, b"User@example.com")
+        self.assertEqual(p.password, b"Pass")
+        self.assertEqual(p.hostname, b"www.python.org")
+        self.assertEqual(p.port, 80)
+        self.assertEqual(p.geturl(), url)
 
     def test_attributes_bad_port(self):
         """Check handling of non-integer ports."""
@@ -433,6 +518,15 @@ class UrlParseTestCase(unittest.TestCase):
         self.assertEqual(p.netloc, "www.example.net:foo")
         self.assertRaises(ValueError, lambda: p.port)
 
+        # Once again, repeat ourselves to test bytes
+        p = urllib.parse.urlsplit(b"http://www.example.net:foo")
+        self.assertEqual(p.netloc, b"www.example.net:foo")
+        self.assertRaises(ValueError, lambda: p.port)
+
+        p = urllib.parse.urlparse(b"http://www.example.net:foo")
+        self.assertEqual(p.netloc, b"www.example.net:foo")
+        self.assertRaises(ValueError, lambda: p.port)
+
     def test_attributes_without_netloc(self):
         # This example is straight from RFC 3261.  It looks like it
         # should allow the username, hostname, and port to be filled
@@ -456,10 +550,30 @@ class UrlParseTestCase(unittest.TestCase):
         self.assertEqual(p.port, None)
         self.assertEqual(p.geturl(), uri)
 
+        # You guessed it, repeating the test with bytes input
+        uri = b"sip:alice@atlanta.com;maddr=239.255.255.1;ttl=15"
+        p = urllib.parse.urlsplit(uri)
+        self.assertEqual(p.netloc, b"")
+        self.assertEqual(p.username, None)
+        self.assertEqual(p.password, None)
+        self.assertEqual(p.hostname, None)
+        self.assertEqual(p.port, None)
+        self.assertEqual(p.geturl(), uri)
+
+        p = urllib.parse.urlparse(uri)
+        self.assertEqual(p.netloc, b"")
+        self.assertEqual(p.username, None)
+        self.assertEqual(p.password, None)
+        self.assertEqual(p.hostname, None)
+        self.assertEqual(p.port, None)
+        self.assertEqual(p.geturl(), uri)
+
     def test_noslash(self):
         # Issue 1637: http://foo.com?query is legal
         self.assertEqual(urllib.parse.urlparse("http://example.com?blahblah=/foo"),
                          ('http', 'example.com', '', '', 'blahblah=/foo', ''))
+        self.assertEqual(urllib.parse.urlparse(b"http://example.com?blahblah=/foo"),
+                         (b'http', b'example.com', b'', b'', b'blahblah=/foo', b''))
 
     def test_withoutscheme(self):
         # Test urlparse without scheme
@@ -472,6 +586,13 @@ class UrlParseTestCase(unittest.TestCase):
                 ('','www.python.org:80','','','',''))
         self.assertEqual(urllib.parse.urlparse("http://www.python.org:80"),
                 ('http','www.python.org:80','','','',''))
+        # Repeat for bytes input
+        self.assertEqual(urllib.parse.urlparse(b"path"),
+                (b'',b'',b'path',b'',b'',b''))
+        self.assertEqual(urllib.parse.urlparse(b"//www.python.org:80"),
+                (b'',b'www.python.org:80',b'',b'',b'',b''))
+        self.assertEqual(urllib.parse.urlparse(b"http://www.python.org:80"),
+                (b'http',b'www.python.org:80',b'',b'',b'',b''))
 
     def test_portseparator(self):
         # Issue 754016 makes changes for port separator ':' from scheme separator
@@ -481,6 +602,13 @@ class UrlParseTestCase(unittest.TestCase):
         self.assertEqual(urllib.parse.urlparse("https:"),('https','','','','',''))
         self.assertEqual(urllib.parse.urlparse("http://www.python.org:80"),
                 ('http','www.python.org:80','','','',''))
+        # As usual, need to check bytes input as well
+        self.assertEqual(urllib.parse.urlparse(b"path:80"),
+                (b'',b'',b'path:80',b'',b'',b''))
+        self.assertEqual(urllib.parse.urlparse(b"http:"),(b'http',b'',b'',b'',b'',b''))
+        self.assertEqual(urllib.parse.urlparse(b"https:"),(b'https',b'',b'',b'',b'',b''))
+        self.assertEqual(urllib.parse.urlparse(b"http://www.python.org:80"),
+                (b'http',b'www.python.org:80',b'',b'',b'',b''))
 
     def test_usingsys(self):
         # Issue 3314: sys module is used in the error
@@ -492,6 +620,71 @@ class UrlParseTestCase(unittest.TestCase):
                          ('s3', 'foo.com', '/stuff', '', '', ''))
         self.assertEqual(urllib.parse.urlparse("x-newscheme://foo.com/stuff"),
                          ('x-newscheme', 'foo.com', '/stuff', '', '', ''))
+        # And for bytes...
+        self.assertEqual(urllib.parse.urlparse(b"s3://foo.com/stuff"),
+                         (b's3', b'foo.com', b'/stuff', b'', b'', b''))
+        self.assertEqual(urllib.parse.urlparse(b"x-newscheme://foo.com/stuff"),
+                         (b'x-newscheme', b'foo.com', b'/stuff', b'', b'', b''))
+
+    def test_mixed_types_rejected(self):
+        # Several functions that process either strings or ASCII encoded bytes
+        # accept multiple arguments. Check they reject mixed type input
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlparse("www.python.org", b"http")
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlparse(b"www.python.org", "http")
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlsplit("www.python.org", b"http")
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlsplit(b"www.python.org", "http")
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlunparse(( b"http", "www.python.org","","","",""))
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlunparse(("http", b"www.python.org","","","",""))
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlunsplit((b"http", "www.python.org","","",""))
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urlunsplit(("http", b"www.python.org","","",""))
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urljoin("http://python.org", b"http://python.org")
+        with self.assertRaisesRegexp(TypeError, "Cannot mix str"):
+            urllib.parse.urljoin(b"http://python.org", "http://python.org")
+
+    def _check_result_type(self, str_type):
+        num_args = len(str_type._fields)
+        bytes_type = str_type._encoded_counterpart
+        self.assertIs(bytes_type._decoded_counterpart, str_type)
+        str_args = ('',) * num_args
+        bytes_args = (b'',) * num_args
+        str_result = str_type(*str_args)
+        bytes_result = bytes_type(*bytes_args)
+        encoding = 'ascii'
+        errors = 'strict'
+        self.assertEqual(str_result, str_args)
+        self.assertEqual(bytes_result.decode(), str_args)
+        self.assertEqual(bytes_result.decode(), str_result)
+        self.assertEqual(bytes_result.decode(encoding), str_args)
+        self.assertEqual(bytes_result.decode(encoding), str_result)
+        self.assertEqual(bytes_result.decode(encoding, errors), str_args)
+        self.assertEqual(bytes_result.decode(encoding, errors), str_result)
+        self.assertEqual(bytes_result, bytes_args)
+        self.assertEqual(str_result.encode(), bytes_args)
+        self.assertEqual(str_result.encode(), bytes_result)
+        self.assertEqual(str_result.encode(encoding), bytes_args)
+        self.assertEqual(str_result.encode(encoding), bytes_result)
+        self.assertEqual(str_result.encode(encoding, errors), bytes_args)
+        self.assertEqual(str_result.encode(encoding, errors), bytes_result)
+
+    def test_result_pairs(self):
+        # Check encoding and decoding between result pairs
+        result_types = [
+          urllib.parse.DefragResult,
+          urllib.parse.SplitResult,
+          urllib.parse.ParseResult,
+        ]
+        for result_type in result_types:
+            self._check_result_type(result_type)
+
 
 def test_main():
     support.run_unittest(UrlParseTestCase)

Lib/urllib/parse.py

@@ -60,6 +60,7 @@ scheme_chars = ('abcdefghijklmnopqrstuvwxyz'
                 '0123456789'
                 '+-.')
 
+# XXX: Consider replacing with functools.lru_cache
 MAX_CACHE_SIZE = 20
 _parse_cache = {}
 
@@ -69,66 +70,210 @@ def clear_cache():
     _safe_quoters.clear()
 
 
-class ResultMixin(object):
-    """Shared methods for the parsed result objects."""
+# Helpers for bytes handling
+# For 3.2, we deliberately require applications that
+# handle improperly quoted URLs to do their own
+# decoding and encoding. If valid use cases are
+# presented, we may relax this by using latin-1
+# decoding internally for 3.3
+_implicit_encoding = 'ascii'
+_implicit_errors = 'strict'
+
+def _noop(obj):
+    return obj
+
+def _encode_result(obj, encoding=_implicit_encoding,
+                        errors=_implicit_errors):
+    return obj.encode(encoding, errors)
+
+def _decode_args(args, encoding=_implicit_encoding,
+                       errors=_implicit_errors):
+    return tuple(x.decode(encoding, errors) if x else '' for x in args)
+
+def _coerce_args(*args):
+    # Invokes decode if necessary to create str args
+    # and returns the coerced inputs along with
+    # an appropriate result coercion function
+    #   - noop for str inputs
+    #   - encoding function otherwise
+    str_input = isinstance(args[0], str)
+    for arg in args[1:]:
+        # We special-case the empty string to support the
+        # "scheme=''" default argument to some functions
+        if arg and isinstance(arg, str) != str_input:
+            raise TypeError("Cannot mix str and non-str arguments")
+    if str_input:
+        return args + (_noop,)
+    return _decode_args(args) + (_encode_result,)
+
+# Result objects are more helpful than simple tuples
+class _ResultMixinStr(object):
+    """Standard approach to encoding parsed results from str to bytes"""
+    __slots__ = ()
+
+    def encode(self, encoding='ascii', errors='strict'):
+        return self._encoded_counterpart(*(x.encode(encoding, errors) for x in self))
+
+
+class _ResultMixinBytes(object):
+    """Standard approach to decoding parsed results from bytes to str"""
+    __slots__ = ()
+
+    def decode(self, encoding='ascii', errors='strict'):
+        return self._decoded_counterpart(*(x.decode(encoding, errors) for x in self))
+
+
+class _NetlocResultMixinBase(object):
+    """Shared methods for the parsed result objects containing a netloc element"""
+    __slots__ = ()
 
     @property
     def username(self):
-        netloc = self.netloc
-        if "@" in netloc:
-            userinfo = netloc.rsplit("@", 1)[0]
-            if ":" in userinfo:
-                userinfo = userinfo.split(":", 1)[0]
-            return userinfo
-        return None
+        return self._userinfo[0]
 
     @property
     def password(self):
-        netloc = self.netloc
-        if "@" in netloc:
-            userinfo = netloc.rsplit("@", 1)[0]
-            if ":" in userinfo:
-                return userinfo.split(":", 1)[1]
-        return None
+        return self._userinfo[1]
 
     @property
     def hostname(self):
-        netloc = self.netloc.split('@')[-1]
-        if '[' in netloc and ']' in netloc:
-            return netloc.split(']')[0][1:].lower()
-        elif ':' in netloc:
-            return netloc.split(':')[0].lower()
-        elif netloc == '':
-            return None
-        else:
-            return netloc.lower()
+        hostname = self._hostinfo[0]
+        if not hostname:
+            hostname = None
+        elif hostname is not None:
+            hostname = hostname.lower()
+        return hostname
 
     @property
     def port(self):
-        netloc = self.netloc.split('@')[-1].split(']')[-1]
-        if ':' in netloc:
-            port = netloc.split(':')[1]
-            return int(port, 10)
+        port = self._hostinfo[1]
+        if port is not None:
+            port = int(port, 10)
+        return port
+
+
+class _NetlocResultMixinStr(_NetlocResultMixinBase, _ResultMixinStr):
+    __slots__ = ()
+
+    @property
+    def _userinfo(self):
+        netloc = self.netloc
+        userinfo, have_info, hostinfo = netloc.rpartition('@')
+        if have_info:
+            username, have_password, password = userinfo.partition(':')
+            if not have_password:
+                password = None
         else:
-            return None
+            username = password = None
+        return username, password
+
+    @property
+    def _hostinfo(self):
+        netloc = self.netloc
+        _, _, hostinfo = netloc.rpartition('@')
+        _, have_open_br, bracketed = hostinfo.partition('[')
+        if have_open_br:
+            hostname, _, port = bracketed.partition(']')
+            _, have_port, port = port.partition(':')
+        else:
+            hostname, have_port, port = hostinfo.partition(':')
+        if not have_port:
+            port = None
+        return hostname, port
+
+
+class _NetlocResultMixinBytes(_NetlocResultMixinBase, _ResultMixinBytes):
+    __slots__ = ()
+
+    @property
+    def _userinfo(self):
+        netloc = self.netloc
+        userinfo, have_info, hostinfo = netloc.rpartition(b'@')
+        if have_info:
+            username, have_password, password = userinfo.partition(b':')
+            if not have_password:
+                password = None
+        else:
+            username = password = None
+        return username, password
+
+    @property
+    def _hostinfo(self):
+        netloc = self.netloc
+        _, _, hostinfo = netloc.rpartition(b'@')
+        _, have_open_br, bracketed = hostinfo.partition(b'[')
+        if have_open_br:
+            hostname, _, port = bracketed.partition(b']')
+            _, have_port, port = port.partition(b':')
+        else:
+            hostname, have_port, port = hostinfo.partition(b':')
+        if not have_port:
+            port = None
+        return hostname, port
+
 
 from collections import namedtuple
 
-class SplitResult(namedtuple('SplitResult', 'scheme netloc path query fragment'), ResultMixin):
+_DefragResultBase = namedtuple('DefragResult', 'url fragment')
+_SplitResultBase = namedtuple('SplitResult', 'scheme netloc path query fragment')
+_ParseResultBase = namedtuple('ParseResult', 'scheme netloc path params query fragment')
+
+# For backwards compatibility, alias _NetlocResultMixinStr
+# ResultBase is no longer part of the documented API, but it is
+# retained since deprecating it isn't worth the hassle
+ResultBase = _NetlocResultMixinStr
 
+# Structured result objects for string data
+class DefragResult(_DefragResultBase, _ResultMixinStr):
     __slots__ = ()
+    def geturl(self):
+        if self.fragment:
+            return self.url + '#' + self.fragment
+        else:
+            return self.url
 
+class SplitResult(_SplitResultBase, _NetlocResultMixinStr):
+    __slots__ = ()
     def geturl(self):
         return urlunsplit(self)
 
+class ParseResult(_ParseResultBase, _NetlocResultMixinStr):
+    __slots__ = ()
+    def geturl(self):
+        return urlunparse(self)
 
-class ParseResult(namedtuple('ParseResult', 'scheme netloc path params query fragment'), ResultMixin):
+# Structured result objects for bytes data
+class DefragResultBytes(_DefragResultBase, _ResultMixinBytes):
+    __slots__ = ()
+    def geturl(self):
+        if self.fragment:
+            return self.url + b'#' + self.fragment
+        else:
+            return self.url
 
+class SplitResultBytes(_SplitResultBase, _NetlocResultMixinBytes):
     __slots__ = ()
+    def geturl(self):
+        return urlunsplit(self)
 
+class ParseResultBytes(_ParseResultBase, _NetlocResultMixinBytes):
+    __slots__ = ()
     def geturl(self):
         return urlunparse(self)
 
+# Set up the encode/decode result pairs
+def _fix_result_transcoding():
+    _result_pairs = (
+        (DefragResult, DefragResultBytes),
+        (SplitResult, SplitResultBytes),
+        (ParseResult, ParseResultBytes),
+    )
+    for _decoded, _encoded in _result_pairs:
+        _decoded._encoded_counterpart = _encoded
+        _encoded._decoded_counterpart = _decoded
+
+_fix_result_transcoding()
+del _fix_result_transcoding
 
 def urlparse(url, scheme='', allow_fragments=True):
     """Parse a URL into 6 components:
@@ -136,13 +281,15 @@ def urlparse(url, scheme='', allow_fragments=True):
     Return a 6-tuple: (scheme, netloc, path, params, query, fragment).
     Note that we don't break the components up in smaller bits
     (e.g. netloc is a single string) and we don't expand % escapes."""
+    url, scheme, _coerce_result = _coerce_args(url, scheme)
     tuple = urlsplit(url, scheme, allow_fragments)
     scheme, netloc, url, query, fragment = tuple
     if scheme in uses_params and ';' in url:
         url, params = _splitparams(url)
     else:
         params = ''
-    return ParseResult(scheme, netloc, url, params, query, fragment)
+    result = ParseResult(scheme, netloc, url, params, query, fragment)
+    return _coerce_result(result)
 
 def _splitparams(url):
     if '/'  in url:
@@ -167,11 +314,12 @@ def urlsplit(url, scheme='', allow_fragments=True):
     Return a 5-tuple: (scheme, netloc, path, query, fragment).
     Note that we don't break the components up in smaller bits
     (e.g. netloc is a single string) and we don't expand % escapes."""
+    url, scheme, _coerce_result = _coerce_args(url, scheme)
     allow_fragments = bool(allow_fragments)
     key = url, scheme, allow_fragments, type(url), type(scheme)
     cached = _parse_cache.get(key, None)
     if cached:
-        return cached
+        return _coerce_result(cached)
     if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth
         clear_cache()
     netloc = query = fragment = ''
@@ -191,7 +339,7 @@ def urlsplit(url, scheme='', allow_fragments=True):
                 url, query = url.split('?', 1)
             v = SplitResult(scheme, netloc, url, query, fragment)
             _parse_cache[key] = v
-            return v
+            return _coerce_result(v)
         if url.endswith(':') or not url[i+1].isdigit():
             for c in url[:i]:
                 if c not in scheme_chars:
@@ -209,17 +357,18 @@ def urlsplit(url, scheme='', allow_fragments=True):
         url, query = url.split('?', 1)
     v = SplitResult(scheme, netloc, url, query, fragment)
     _parse_cache[key] = v
-    return v
+    return _coerce_result(v)
 
 def urlunparse(components):
     """Put a parsed URL back together again.  This may result in a
     slightly different, but equivalent URL, if the URL that was parsed
     originally had redundant delimiters, e.g. a ? with an empty query
     (the draft states that these are equivalent)."""
-    scheme, netloc, url, params, query, fragment = components
+    scheme, netloc, url, params, query, fragment, _coerce_result = (
+                                                  _coerce_args(*components))
     if params:
         url = "%s;%s" % (url, params)
-    return urlunsplit((scheme, netloc, url, query, fragment))
+    return _coerce_result(urlunsplit((scheme, netloc, url, query, fragment)))
 
 def urlunsplit(components):
     """Combine the elements of a tuple as returned by urlsplit() into a
@@ -227,7 +376,8 @@ def urlunsplit(components):
     This may result in a slightly different, but equivalent URL, if the URL that
     was parsed originally had unnecessary delimiters (for example, a ? with an
     empty query; the RFC states that these are equivalent)."""
-    scheme, netloc, url, query, fragment = components
+    scheme, netloc, url, query, fragment, _coerce_result = (
+                                          _coerce_args(*components))
     if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'):
         if url and url[:1] != '/': url = '/' + url
         url = '//' + (netloc or '') + url
@@ -237,7 +387,7 @@ def urlunsplit(components):
         url = url + '?' + query
     if fragment:
         url = url + '#' + fragment
-    return url
+    return _coerce_result(url)
 
 def urljoin(base, url, allow_fragments=True):
     """Join a base URL and a possibly relative URL to form an absolute
@@ -246,32 +396,33 @@ def urljoin(base, url, allow_fragments=True):
         return url
     if not url:
         return base
+    base, url, _coerce_result = _coerce_args(base, url)
     bscheme, bnetloc, bpath, bparams, bquery, bfragment = \
             urlparse(base, '', allow_fragments)
     scheme, netloc, path, params, query, fragment = \
             urlparse(url, bscheme, allow_fragments)
     if scheme != bscheme or scheme not in uses_relative:
-        return url
+        return _coerce_result(url)
     if scheme in uses_netloc:
         if netloc:
-            return urlunparse((scheme, netloc, path,
-                               params, query, fragment))
+            return _coerce_result(urlunparse((scheme, netloc, path,
+                                              params, query, fragment)))
         netloc = bnetloc
     if path[:1] == '/':
-        return urlunparse((scheme, netloc, path,
-                           params, query, fragment))
+        return _coerce_result(urlunparse((scheme, netloc, path,
+                                          params, query, fragment)))
     if not path:
         path = bpath
         if not params:
             params = bparams
         else:
             path = path[:-1]
-            return urlunparse((scheme, netloc, path,
-                                params, query, fragment))
+            return _coerce_result(urlunparse((scheme, netloc, path,
+                                              params, query, fragment)))
         if not query:
             query = bquery
-        return urlunparse((scheme, netloc, path,
-                           params, query, fragment))
+        return _coerce_result(urlunparse((scheme, netloc, path,
+                                          params, query, fragment)))
     segments = bpath.split('/')[:-1] + path.split('/')
     # XXX The stuff below is bogus in various ways...
     if segments[-1] == '.':
@@ -293,8 +444,8 @@ def urljoin(base, url, allow_fragments=True):
         segments[-1] = ''
     elif len(segments) >= 2 and segments[-1] == '..':
         segments[-2:] = ['']
-    return urlunparse((scheme, netloc, '/'.join(segments),
-                       params, query, fragment))
+    return _coerce_result(urlunparse((scheme, netloc, '/'.join(segments),
+                                      params, query, fragment)))
 
 def urldefrag(url):
     """Removes any existing fragment from URL.
@@ -303,12 +454,14 @@ def urldefrag(url):
     the URL contained no fragments, the second element is the
     empty string.
     """
+    url, _coerce_result = _coerce_args(url)
     if '#' in url:
         s, n, p, a, q, frag = urlparse(url)
         defrag = urlunparse((s, n, p, a, q, ''))
-        return defrag, frag
     else:
-        return url, ''
+        frag = ''
+        defrag = url
+    return _coerce_result(DefragResult(defrag, frag))
 
 def unquote_to_bytes(string):
     """unquote_to_bytes('abc%20def') -> b'abc def'."""
@@ -420,6 +573,7 @@ def parse_qsl(qs, keep_blank_values=False, strict_parsing=False):
 
     Returns a list, as G-d intended.
     """
+    qs, _coerce_result = _coerce_args(qs)
     pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
     r = []
     for name_value in pairs:
@@ -435,10 +589,9 @@ def parse_qsl(qs, keep_blank_values=False, strict_parsing=False):
             else:
                 continue
         if len(nv[1]) or keep_blank_values:
-            name = unquote(nv[0].replace('+', ' '))
-            value = unquote(nv[1].replace('+', ' '))
+            name = _coerce_result(unquote(nv[0].replace('+', ' ')))
+            value = _coerce_result(unquote(nv[1].replace('+', ' ')))
             r.append((name, value))
-
     return r
 
 def unquote_plus(string, encoding='utf-8', errors='replace'):

Misc/NEWS

@@ -43,6 +43,9 @@ Core and Builtins
 Library
 -------
 
+- Issue #9873: The URL parsing functions in urllib.parse now accept
+  ASCII byte sequences as input in addition to character strings.
+
 - Issue #10586: The statistics API for the new functools.lru_cache has
   been changed to a single cache_info() method returning a named tuple.