urlnorm.py: URL normalisation

urlnorm.py

#!/usr/bin/env python

"""
urlnorm.py - URL normalisation routines

urlnorm normalises a URL by;
  * lowercasing the scheme and hostname
  * taking out default port if present (e.g., http://www.foo.com:80/)
  * collapsing the path (./, ../, etc)
  * removing the last character in the hostname if it is '.'
  * unquoting any %-escaped characters

Available functions:
  norms - given a URL (string), returns a normalised URL
  norm - given a URL tuple, returns a normalised tuple
  test - test suite
  
CHANGES:
0.92 - unknown schemes now pass the port through silently
0.91 - general cleanup
     - changed dictionaries to lists where appropriate
     - more fine-grained authority parsing and normalisation    
"""

__license__ = """
Copyright (c) 1999-2002 Mark Nottingham <mnot@pobox.com>

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
"""

__version__ = "0.93"

from urlparse import urlparse, urlunparse
from urllib import unquote
from string import lower
import re

_collapse = re.compile('([^/]+/\.\./?|/\./|//|/\.$|/\.\.$)')
_server_authority = re.compile('^(?:([^\@]+)\@)?([^\:]+)(?:\:(.+))?$')
_default_port = {   'http': '80',
                    'https': '443',
                    'gopher': '70',
                    'news': '119',
                    'snews': '563',
                    'nntp': '119',
                    'snntp': '563',
                    'ftp': '21',
                    'telnet': '23',
                    'prospero': '191',
                }
_relative_schemes = [   'http',
                        'https',
                        'news',
                        'snews',
                        'nntp',
                        'snntp',
                        'ftp',
                        'file',
                        ''
                    ]
_server_authority_schemes = [   'http',
                                'https',
                                'news',
                                'snews',
                                'ftp',
                            ]


def norms(urlstring):
    """given a string URL, return its normalised form"""
    return urlunparse(norm(urlparse(urlstring)))


def norm(urltuple):
    """given a six-tuple URL, return its normalised form"""
    (scheme, authority, path, parameters, query, fragment) = urltuple
    scheme = lower(scheme)
    if authority:
        userinfo, host, port = _server_authority.match(authority).groups()
        if host[-1] == '.':
            host = host[:-1]
        authority = lower(host)
        if userinfo:
            authority = "%s@%s" % (userinfo, authority)
        if port and port != _default_port.get(scheme, None):
            authority = "%s:%s" % (authority, port)
    if scheme in _relative_schemes:
        last_path = path
        while 1:
            path = _collapse.sub('/', path, 1)
            if last_path == path:
                break
            last_path = path
    path = unquote(path)
    return (scheme, authority, path, parameters, query, fragment)



def test():
    """ test suite; some taken from RFC1808. """
    tests = {    
        '/foo/bar/.':                    '/foo/bar/', 
        '/foo/bar/./':                   '/foo/bar/',
        '/foo/bar/..':                   '/foo/',
        '/foo/bar/../':                  '/foo/',
        '/foo/bar/../baz':               '/foo/baz',
        '/foo/bar/../..':                '/',
        '/foo/bar/../../':               '/',
        '/foo/bar/../../baz':            '/baz',
        '/foo/bar/../../../baz':         '/../baz',
        '/foo/bar/../../../../baz':      '/baz',
        '/./foo':                        '/foo',
        '/../foo':                       '/../foo',
        '/foo.':                         '/foo.',
        '/.foo':                         '/.foo',
        '/foo..':                        '/foo..',
        '/..foo':                        '/..foo',
        '/./../foo':                     '/../foo',
        '/./foo/.':                      '/foo/',
        '/foo/./bar':                    '/foo/bar',
        '/foo/../bar':                   '/bar',
        '/foo//':                        '/foo/',
        '/foo///bar//':                  '/foo/bar/',    
        'http://www.foo.com:80/foo':     'http://www.foo.com/foo',
        'http://www.foo.com:8000/foo':   'http://www.foo.com:8000/foo',
        'http://www.foo.com./foo/bar.html': 'http://www.foo.com/foo/bar.html',
        'http://www.foo.com.:81/foo':    'http://www.foo.com:81/foo',
        'http://www.foo.com/%7ebar':     'http://www.foo.com/~bar',
        'http://www.foo.com/%7Ebar':     'http://www.foo.com/~bar',
        'ftp://user:pass@ftp.foo.net/foo/bar': 'ftp://user:pass@ftp.foo.net/foo/bar',
        'http://USER:pass@www.Example.COM/foo/bar': 'http://USER:pass@www.example.com/foo/bar',
        'http://www.example.com./':      'http://www.example.com/',
        '-':                             '-',
    }

    n_correct, n_fail = 0, 0
    test_keys = tests.keys()
    test_keys.sort()            
    for i in test_keys:
        print 'ORIGINAL:', i
        cleaned = norms(i)
        answer = tests[i]
        print 'CLEANED: ', cleaned
        print 'CORRECT: ', answer
        if cleaned != answer:
            print "*** TEST FAILED"
            n_fail = n_fail + 1
        else:
            n_correct = n_correct + 1
        print        
    print "TOTAL CORRECT:", n_correct
    print "TOTAL FAILURE:", n_fail


if __name__ == '__main__':
    test()

Very usefull piece of code, i am trying to prse some URLs and sometimes is painfull.

cool~

Thank you Mark, I think your norm function should go into the urllib Python standard library.

A few remarks on your code though:

1. The _server_authority_schemes variable is not used.

2. The _server_authority regex does not allow empty port components with their ":" delimiters while RFC 3986 does (and recommends to normalize them by removing the ":" delimiter):

    >>> norms('http://example.com:/')
    Traceback (most recent call last):
    [...]
        userinfo, host, port = _authority.match(authority).groups()
    AttributeError: 'NoneType' object has no attribute 'groups'
   

   To correct this, the line:

    _server_authority = re.compile("^(?:([^\@]+)\@)?([^\:]+)(?:\:(.+))?$")
   

   should be changed to this line (note the change of the last + quantifier to the * quantifier):

    _server_authority = re.compile("^(?:([^\@]+)\@)?([^\:]+)(?:\:(.*))?$")
   

   Also you could add this line to the test function:

    'http://example.com:/':     'http://example.com/',
   

3. The norm function does not quote the path segment, so the norms function returns:

    >>> norms('http://example.com/ foo')
    'http://example.com/ foo',
   

   instead of:

    >>> norms('http://example.com/ foo')
    'http://example.com/%20foo',
   

   To correct this, the line in the norm function:

    path = urllib.parse.unquote(path)
   

   should be appended with this line:

    path = urllib.parse.quote(path)
   

   Also you could add these lines to the test function (the second line to check percent-encoding normalization of unreserved characters):

    'http://example.com/ foo':     'http://example.com/%20foo',
    'http://example.com/fo%6F':    'http://example.com/foo',
   

4. The _collapse regular expression incorrectly removes consecutive '/' delimiters in the path component, so the norms function returns:

    >>> norms('/foo////bar')
    '/foo/bar'
   

   instead of:

    >>> norms('/foo////bar')
    '/foo////bar'
   

   To correct this, a line calling the remove_dot_segments function specified in RFC 3986:

    path = remove_dot_segments(path)
   

   should replace these lines:

    last_path = path
    while 1:
        path = _collapse.sub('/', path, 1)
        if last_path == path:
            break
        last_path = path
   

   and this line should be removed:

    _collapse = re.compile('([^/]+/\.\./?|/\./|//|/\.$|/\.\.$)')
   

5. The norm function doesn't apply scheme-based normalization, as the norms function returns:

    >>> norms('http://example.com')
    'http://example.com'
   

   instead of (note the trailing slash):

    >>> norms('http://example.com')
    'http://example.com/'
   

   To correct this, the line in the norm function:

    path = unquote(path)
   

   should be prepended with these lines:

    if authority and not path:
        path = "/"
   

   Also you could add this line to the test function:

    'http://www.foo.com':     'http://www.foo.com/',
   

6. A Python 3 version of the code would be great.

I noticed another issue, but related to the urllib.parse module (and your norms function cannot do anything about it). The Python library documentation of the urllib.parse.urlunparse and urllib.parse.urlunsplit functions states:

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).

So with the http://example.com/? URI:

>>> import urllib.parse
>>> urllib.parse.urlunparse(urllib.parse.urlparse("http://example.com/?"))
'http://example.com/'
>>> urllib.parse.urlunsplit(urllib.parse.urlsplit("http://example.com/?"))
'http://example.com/'

But RFC 3986 states the exact opposite:

Normalization should not remove delimiters when their associated component is empty unless licensed to do so by the scheme specification. For example, the URI "http://example.com/?" cannot be assumed to be equivalent to any of the examples above. Likewise, the presence or absence of delimiters within a userinfo subcomponent is usually significant to its interpretation. The fragment component is not subject to any scheme-based normalization; thus, two URIs that differ only by the suffix "#" are considered different regardless of the scheme.

Consequently, both urllib.parse.urlparse or urllib.parse.urlsplit lose the "delimiter + empty component" information of the URI string, so they report false equivalent URIs:

>>> import urllib.parse
>>> urllib.parse.urlparse("http://example.com/?") == urllib.parse.urlparse("http://example.com/")
True
>>> urllib.parse.urlsplit("http://example.com/?") == urllib.parse.urlsplit("http://example.com/")
True

So you could add these lines to your tests (the last two left commented for the moment as urllib is broken):

'http://@example.com/':       'http://@example.com/',
# 'http://example.com/?':     'http://example.com/?',
# 'http://example.com/#':     'http://example.com/#',

For those interested, I have forked and improved this Gist (Python 3, RFC 3986 compliance, Unittest framework, all the above corrections) here:
https://gist.github.com/maggyero/9bc1382b74b0eaf67bb020669c01b234