Lib/urlparse.py

   1 """Parse (absolute and relative) URLs.
   2
   3 See RFC 1808: "Relative Uniform Resource Locators", by R. Fielding,
   4 UC Irvine, June 1995.
   5 """
   6
   7 __all__ = ["urlparse", "urlunparse", "urljoin", "urldefrag",
   8            "urlsplit", "urlunsplit"]
   9
  10 # A classification of schemes ('' means apply by default)
  11 uses_relative = ['ftp', 'http', 'gopher', 'nntp', 'imap',
  12                                'wais', 'file', 'https', 'shttp', 'mms',
  13                                'prospero', 'rtsp', 'rtspu', '']
  14 uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet',
  15                              'imap', 'wais', 'file', 'mms', 'https', 'shttp',
  16                              'snews', 'prospero', 'rtsp', 'rtspu', 'rsync', '']
  17 non_hierarchical = ['gopher', 'hdl', 'mailto', 'news',
  18                                   'telnet', 'wais', 'imap', 'snews', 'sip']
  19 uses_params = ['ftp', 'hdl', 'prospero', 'http', 'imap',
  20                              'https', 'shttp', 'rtsp', 'rtspu', 'sip',
  21                              'mms', '']
  22 uses_query = ['http', 'wais', 'imap', 'https', 'shttp', 'mms',
  23                             'gopher', 'rtsp', 'rtspu', 'sip', '']
  24 uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news',
  25                                'nntp', 'wais', 'https', 'shttp', 'snews',
  26                                'file', 'prospero', '']
  27
  28 # Characters valid in scheme names
  29 scheme_chars = ('abcdefghijklmnopqrstuvwxyz'
  30                 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
  31                 '0123456789'
  32                 '+-.')
  33
  34 MAX_CACHE_SIZE = 20
  35 _parse_cache = {}
  36
  37 def clear_cache():
  38     """Clear the parse cache."""
  39     global _parse_cache
  40     _parse_cache = {}
  41
  42
  43 def urlparse(url, scheme='', allow_fragments=1):
  44     """Parse a URL into 6 components:
  45     <scheme>://<netloc>/<path>;<params>?<query>#<fragment>
  46     Return a 6-tuple: (scheme, netloc, path, params, query, fragment).
  47     Note that we don't break the components up in smaller bits
  48     (e.g. netloc is a single string) and we don't expand % escapes."""
  49     tuple = urlsplit(url, scheme, allow_fragments)
  50     scheme, netloc, url, query, fragment = tuple
  51     if scheme in uses_params and ';' in url:
  52         url, params = _splitparams(url)
  53     else:
  54         params = ''
  55     return scheme, netloc, url, params, query, fragment
  56
  57 def _splitparams(url):
  58     if '/'  in url:
  59         i = url.find(';', url.rfind('/'))
  60         if i < 0:
  61             return url, ''
  62     else:
  63         i = url.find(';')
  64     return url[:i], url[i+1:]
  65
  66 def _splitnetloc(url, start=0):
  67     for c in '/?#': # the order is important!
  68         delim = url.find(c, start)
  69         if delim >= 0:
  70             break
  71     else:
  72         delim = len(url)
  73     return url[start:delim], url[delim:]
  74
  75 def urlsplit(url, scheme='', allow_fragments=1):
  76     """Parse a URL into 5 components:
  77     <scheme>://<netloc>/<path>?<query>#<fragment>
  78     Return a 5-tuple: (scheme, netloc, path, query, fragment).
  79     Note that we don't break the components up in smaller bits
  80     (e.g. netloc is a single string) and we don't expand % escapes."""
  81     key = url, scheme, allow_fragments
  82     cached = _parse_cache.get(key, None)
  83     if cached:
  84         return cached
  85     if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth
  86         clear_cache()
  87     netloc = query = fragment = ''
  88     i = url.find(':')
  89     if i > 0:
  90         if url[:i] == 'http': # optimize the common case
  91             scheme = url[:i].lower()
  92             url = url[i+1:]
  93             if url[:2] == '//':
  94                 netloc, url = _splitnetloc(url, 2)
  95             if allow_fragments and '#' in url:
  96                 url, fragment = url.split('#', 1)
  97             if '?' in url:
  98                 url, query = url.split('?', 1)
  99             tuple = scheme, netloc, url, query, fragment
 100             _parse_cache[key] = tuple
 101             return tuple
 102         for c in url[:i]:
 103             if c not in scheme_chars:
 104                 break
 105         else:
 106             scheme, url = url[:i].lower(), url[i+1:]
 107     if scheme in uses_netloc and url[:2] == '//':
 108         netloc, url = _splitnetloc(url, 2)
 109     if allow_fragments and scheme in uses_fragment and '#' in url:
 110         url, fragment = url.split('#', 1)
 111     if scheme in uses_query and '?' in url:
 112         url, query = url.split('?', 1)
 113     tuple = scheme, netloc, url, query, fragment
 114     _parse_cache[key] = tuple
 115     return tuple
 116
 117 def urlunparse((scheme, netloc, url, params, query, fragment)):
 118     """Put a parsed URL back together again.  This may result in a
 119     slightly different, but equivalent URL, if the URL that was parsed
 120     originally had redundant delimiters, e.g. a ? with an empty query
 121     (the draft states that these are equivalent)."""
 122     if params:
 123         url = "%s;%s" % (url, params)
 124     return urlunsplit((scheme, netloc, url, query, fragment))
 125
 126 def urlunsplit((scheme, netloc, url, query, fragment)):
 127     if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'):
 128         if url and url[:1] != '/': url = '/' + url
 129         url = '//' + (netloc or '') + url
 130     if scheme:
 131         url = scheme + ':' + url
 132     if query:
 133         url = url + '?' + query
 134     if fragment:
 135         url = url + '#' + fragment
 136     return url
 137
 138 def urljoin(base, url, allow_fragments = 1):
 139     """Join a base URL and a possibly relative URL to form an absolute
 140     interpretation of the latter."""
 141     if not base:
 142         return url
 143     if not url:
 144         return base
 145     bscheme, bnetloc, bpath, bparams, bquery, bfragment = \
 146             urlparse(base, '', allow_fragments)
 147     scheme, netloc, path, params, query, fragment = \
 148             urlparse(url, bscheme, allow_fragments)
 149     if scheme != bscheme or scheme not in uses_relative:
 150         return url
 151     if scheme in uses_netloc:
 152         if netloc:
 153             return urlunparse((scheme, netloc, path,
 154                                params, query, fragment))
 155         netloc = bnetloc
 156     if path[:1] == '/':
 157         return urlunparse((scheme, netloc, path,
 158                            params, query, fragment))
 159     if not (path or params or query):
 160         return urlunparse((scheme, netloc, bpath,
 161                            bparams, bquery, fragment))
 162     segments = bpath.split('/')[:-1] + path.split('/')
 163     # XXX The stuff below is bogus in various ways...
 164     if segments[-1] == '.':
 165         segments[-1] = ''
 166     while '.' in segments:
 167         segments.remove('.')
 168     while 1:
 169         i = 1
 170         n = len(segments) - 1
 171         while i < n:
 172             if (segments[i] == '..'
 173                 and segments[i-1] not in ('', '..')):
 174                 del segments[i-1:i+1]
 175                 break
 176             i = i+1
 177         else:
 178             break
 179     if segments == ['', '..']:
 180         segments[-1] = ''
 181     elif len(segments) >= 2 and segments[-1] == '..':
 182         segments[-2:] = ['']
 183     return urlunparse((scheme, netloc, '/'.join(segments),
 184                        params, query, fragment))
 185
 186 def urldefrag(url):
 187     """Removes any existing fragment from URL.
 188
 189     Returns a tuple of the defragmented URL and the fragment.  If
 190     the URL contained no fragments, the second element is the
 191     empty string.
 192     """
 193     if '#' in url:
 194         s, n, p, a, q, frag = urlparse(url)
 195         defrag = urlunparse((s, n, p, a, q, ''))
 196         return defrag, frag
 197     else:
 198         return url, ''
 199
 200
 201 test_input = """
 202       http://a/b/c/d
 203
 204       g:h        = <URL:g:h>
 205       http:g     = <URL:http://a/b/c/g>
 206       http:      = <URL:http://a/b/c/d>
 207       g          = <URL:http://a/b/c/g>
 208       ./g        = <URL:http://a/b/c/g>
 209       g/         = <URL:http://a/b/c/g/>
 210       /g         = <URL:http://a/g>
 211       //g        = <URL:http://g>
 212       ?y         = <URL:http://a/b/c/d?y>
 213       g?y        = <URL:http://a/b/c/g?y>
 214       g?y/./x    = <URL:http://a/b/c/g?y/./x>
 215       .          = <URL:http://a/b/c/>
 216       ./         = <URL:http://a/b/c/>
 217       ..         = <URL:http://a/b/>
 218       ../        = <URL:http://a/b/>
 219       ../g       = <URL:http://a/b/g>
 220       ../..      = <URL:http://a/>
 221       ../../g    = <URL:http://a/g>
 222       ../../../g = <URL:http://a/../g>
 223       ./../g     = <URL:http://a/b/g>
 224       ./g/.      = <URL:http://a/b/c/g/>
 225       /./g       = <URL:http://a/./g>
 226       g/./h      = <URL:http://a/b/c/g/h>
 227       g/../h     = <URL:http://a/b/c/h>
 228       http:g     = <URL:http://a/b/c/g>
 229       http:      = <URL:http://a/b/c/d>
 230       http:?y         = <URL:http://a/b/c/d?y>
 231       http:g?y        = <URL:http://a/b/c/g?y>
 232       http:g?y/./x    = <URL:http://a/b/c/g?y/./x>
 233 """
 234
 235 def test():
 236     import sys
 237     base = ''
 238     if sys.argv[1:]:
 239         fn = sys.argv[1]
 240         if fn == '-':
 241             fp = sys.stdin
 242         else:
 243             fp = open(fn)
 244     else:
 245         import StringIO
 246         fp = StringIO.StringIO(test_input)
 247     while 1:
 248         line = fp.readline()
 249         if not line: break
 250         words = line.split()
 251         if not words:
 252             continue
 253         url = words[0]
 254         parts = urlparse(url)
 255         print '%-10s : %s' % (url, parts)
 256         abs = urljoin(base, url)
 257         if not base:
 258             base = abs
 259         wrapped = '<URL:%s>' % abs
 260         print '%-10s = %s' % (url, wrapped)
 261         if len(words) == 3 and words[1] == '=':
 262             if wrapped != words[2]:
 263                 print 'EXPECTED', words[2], '!!!!!!!!!!'
 264
 265 if __name__ == '__main__':
 266     test()