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', '']
  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 urlsplit(url, scheme='', allow_fragments=1):
  67     """Parse a URL into 5 components:
  68     <scheme>://<netloc>/<path>?<query>#<fragment>
  69     Return a 5-tuple: (scheme, netloc, path, query, fragment).
  70     Note that we don't break the components up in smaller bits
  71     (e.g. netloc is a single string) and we don't expand % escapes."""
  72     key = url, scheme, allow_fragments
  73     cached = _parse_cache.get(key, None)
  74     if cached:
  75         return cached
  76     if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth
  77         clear_cache()
  78     netloc = query = fragment = ''
  79     i = url.find(':')
  80     if i > 0:
  81         if url[:i] == 'http': # optimize the common case
  82             scheme = url[:i].lower()
  83             url = url[i+1:]
  84             if url[:2] == '//':
  85                 i = url.find('/', 2)
  86                 if i < 0:
  87                     i = url.find('#')
  88                     if i < 0:
  89                         i = len(url)
  90                 netloc = url[2:i]
  91                 url = url[i:]
  92             if allow_fragments and '#' in url:
  93                 url, fragment = url.split('#', 1)
  94             if '?' in url:
  95                 url, query = url.split('?', 1)
  96             tuple = scheme, netloc, url, query, fragment
  97             _parse_cache[key] = tuple
  98             return tuple
  99         for c in url[:i]:
 100             if c not in scheme_chars:
 101                 break
 102         else:
 103             scheme, url = url[:i].lower(), url[i+1:]
 104     if scheme in uses_netloc:
 105         if url[:2] == '//':
 106             i = url.find('/', 2)
 107             if i < 0:
 108                 i = len(url)
 109             netloc, url = url[2:i], url[i:]
 110     if allow_fragments and scheme in uses_fragment and '#' in url:
 111         url, fragment = url.split('#', 1)
 112     if scheme in uses_query and '?' in url:
 113         url, query = url.split('?', 1)
 114     tuple = scheme, netloc, url, query, fragment
 115     _parse_cache[key] = tuple
 116     return tuple
 117
 118 def urlunparse((scheme, netloc, url, params, query, fragment)):
 119     """Put a parsed URL back together again.  This may result in a
 120     slightly different, but equivalent URL, if the URL that was parsed
 121     originally had redundant delimiters, e.g. a ? with an empty query
 122     (the draft states that these are equivalent)."""
 123     if params:
 124         url = "%s;%s" % (url, params)
 125     return urlunsplit((scheme, netloc, url, query, fragment))
 126
 127 def urlunsplit((scheme, netloc, url, query, fragment)):
 128     if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'):
 129         if url and url[:1] != '/': url = '/' + url
 130         url = '//' + (netloc or '') + url
 131     if scheme:
 132         url = scheme + ':' + url
 133     if query:
 134         url = url + '?' + query
 135     if fragment:
 136         url = url + '#' + fragment
 137     return url
 138
 139 def urljoin(base, url, allow_fragments = 1):
 140     """Join a base URL and a possibly relative URL to form an absolute
 141     interpretation of the latter."""
 142     if not base:
 143         return url
 144     if not url:
 145         return base
 146     bscheme, bnetloc, bpath, bparams, bquery, bfragment = \
 147             urlparse(base, '', allow_fragments)
 148     scheme, netloc, path, params, query, fragment = \
 149             urlparse(url, bscheme, allow_fragments)
 150     if scheme != bscheme or scheme not in uses_relative:
 151         return url
 152     if scheme in uses_netloc:
 153         if netloc:
 154             return urlunparse((scheme, netloc, path,
 155                                params, query, fragment))
 156         netloc = bnetloc
 157     if path[:1] == '/':
 158         return urlunparse((scheme, netloc, path,
 159                            params, query, fragment))
 160     if not path:
 161         if not params:
 162             params = bparams
 163             if not query:
 164                 query = bquery
 165         return urlunparse((scheme, netloc, bpath,
 166                            params, query, fragment))
 167     segments = bpath.split('/')[:-1] + path.split('/')
 168     # XXX The stuff below is bogus in various ways...
 169     if segments[-1] == '.':
 170         segments[-1] = ''
 171     while '.' in segments:
 172         segments.remove('.')
 173     while 1:
 174         i = 1
 175         n = len(segments) - 1
 176         while i < n:
 177             if (segments[i] == '..'
 178                 and segments[i-1] not in ('', '..')):
 179                 del segments[i-1:i+1]
 180                 break
 181             i = i+1
 182         else:
 183             break
 184     if segments == ['', '..']:
 185         segments[-1] = ''
 186     elif len(segments) >= 2 and segments[-1] == '..':
 187         segments[-2:] = ['']
 188     return urlunparse((scheme, netloc, '/'.join(segments),
 189                        params, query, fragment))
 190
 191 def urldefrag(url):
 192     """Removes any existing fragment from URL.
 193
 194     Returns a tuple of the defragmented URL and the fragment.  If
 195     the URL contained no fragments, the second element is the
 196     empty string.
 197     """
 198     if '#' in url:
 199         s, n, p, a, q, frag = urlparse(url)
 200         defrag = urlunparse((s, n, p, a, q, ''))
 201         return defrag, frag
 202     else:
 203         return url, ''
 204
 205
 206 test_input = """
 207       http://a/b/c/d
 208
 209       g:h        = <URL:g:h>
 210       http:g     = <URL:http://a/b/c/g>
 211       http:      = <URL:http://a/b/c/d>
 212       g          = <URL:http://a/b/c/g>
 213       ./g        = <URL:http://a/b/c/g>
 214       g/         = <URL:http://a/b/c/g/>
 215       /g         = <URL:http://a/g>
 216       //g        = <URL:http://g>
 217       ?y         = <URL:http://a/b/c/d?y>
 218       g?y        = <URL:http://a/b/c/g?y>
 219       g?y/./x    = <URL:http://a/b/c/g?y/./x>
 220       .          = <URL:http://a/b/c/>
 221       ./         = <URL:http://a/b/c/>
 222       ..         = <URL:http://a/b/>
 223       ../        = <URL:http://a/b/>
 224       ../g       = <URL:http://a/b/g>
 225       ../..      = <URL:http://a/>
 226       ../../g    = <URL:http://a/g>
 227       ../../../g = <URL:http://a/../g>
 228       ./../g     = <URL:http://a/b/g>
 229       ./g/.      = <URL:http://a/b/c/g/>
 230       /./g       = <URL:http://a/./g>
 231       g/./h      = <URL:http://a/b/c/g/h>
 232       g/../h     = <URL:http://a/b/c/h>
 233       http:g     = <URL:http://a/b/c/g>
 234       http:      = <URL:http://a/b/c/d>
 235       http:?y         = <URL:http://a/b/c/d?y>
 236       http:g?y        = <URL:http://a/b/c/g?y>
 237       http:g?y/./x    = <URL:http://a/b/c/g?y/./x>
 238 """
 239
 240 def test():
 241     import sys
 242     base = ''
 243     if sys.argv[1:]:
 244         fn = sys.argv[1]
 245         if fn == '-':
 246             fp = sys.stdin
 247         else:
 248             fp = open(fn)
 249     else:
 250         import StringIO
 251         fp = StringIO.StringIO(test_input)
 252     while 1:
 253         line = fp.readline()
 254         if not line: break
 255         words = line.split()
 256         if not words:
 257             continue
 258         url = words[0]
 259         parts = urlparse(url)
 260         print '%-10s : %s' % (url, parts)
 261         abs = urljoin(base, url)
 262         if not base:
 263             base = abs
 264         wrapped = '<URL:%s>' % abs
 265         print '%-10s = %s' % (url, wrapped)
 266         if len(words) == 3 and words[1] == '=':
 267             if wrapped != words[2]:
 268                 print 'EXPECTED', words[2], '!!!!!!!!!!'
 269
 270 if __name__ == '__main__':
 271     test()