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', 'wais', 'file',
  12                  'https', 'shttp',
  13                  'prospero', 'rtsp', 'rtspu', '']
  14 uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet', 'wais',
  15                'file',
  16                'https', 'shttp', 'snews',
  17                'prospero', 'rtsp', 'rtspu', '']
  18 non_hierarchical = ['gopher', 'hdl', 'mailto', 'news', 'telnet', 'wais',
  19                     'snews', 'sip',
  20                     ]
  21 uses_params = ['ftp', 'hdl', 'prospero', 'http',
  22                'https', 'shttp', 'rtsp', 'rtspu', 'sip',
  23                '']
  24 uses_query = ['http', 'wais',
  25               'https', 'shttp',
  26               'gopher', 'rtsp', 'rtspu', 'sip',
  27               '']
  28 uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news', 'nntp', 'wais',
  29                  'https', 'shttp', 'snews',
  30                  'file', 'prospero', '']
  31
  32 # Characters valid in scheme names
  33 scheme_chars = ('abcdefghijklmnopqrstuvwxyz'
  34                 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
  35                 '0123456789'
  36                 '+-.')
  37
  38 MAX_CACHE_SIZE = 20
  39 _parse_cache = {}
  40
  41 def clear_cache():
  42     """Clear the parse cache."""
  43     global _parse_cache
  44     _parse_cache = {}
  45
  46
  47 def urlparse(url, scheme='', allow_fragments=1):
  48     """Parse a URL into 6 components:
  49     <scheme>://<netloc>/<path>;<params>?<query>#<fragment>
  50     Return a 6-tuple: (scheme, netloc, path, params, query, fragment).
  51     Note that we don't break the components up in smaller bits
  52     (e.g. netloc is a single string) and we don't expand % escapes."""
  53     tuple = urlsplit(url, scheme, allow_fragments)
  54     scheme, netloc, url, query, fragment = tuple
  55     if scheme in uses_params and ';' in url:
  56         url, params = _splitparams(url)
  57     else:
  58         params = ''
  59     return scheme, netloc, url, params, query, fragment
  60
  61 def _splitparams(url):
  62     if '/'  in url:
  63         i = url.find(';', url.rfind('/'))
  64         if i < 0:
  65             return url, ''
  66     else:
  67         i = url.find(';')
  68     return url[:i], url[i+1:]
  69
  70 def urlsplit(url, scheme='', allow_fragments=1):
  71     """Parse a URL into 5 components:
  72     <scheme>://<netloc>/<path>?<query>#<fragment>
  73     Return a 5-tuple: (scheme, netloc, path, query, fragment).
  74     Note that we don't break the components up in smaller bits
  75     (e.g. netloc is a single string) and we don't expand % escapes."""
  76     key = url, scheme, allow_fragments
  77     cached = _parse_cache.get(key, None)
  78     if cached:
  79         return cached
  80     if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth
  81         clear_cache()
  82     netloc = query = fragment = ''
  83     i = url.find(':')
  84     if i > 0:
  85         if url[:i] == 'http': # optimize the common case
  86             scheme = url[:i].lower()
  87             url = url[i+1:]
  88             if url[:2] == '//':
  89                 i = url.find('/', 2)
  90                 if i < 0:
  91                     i = url.find('#')
  92                     if i < 0:
  93                         i = len(url)
  94                 netloc = url[2:i]
  95                 url = url[i:]
  96             if allow_fragments and '#' in url:
  97                 url, fragment = url.split('#', 1)
  98             if '?' in url:
  99                 url, query = url.split('?', 1)
 100             tuple = scheme, netloc, url, query, fragment
 101             _parse_cache[key] = tuple
 102             return tuple
 103         for c in url[:i]:
 104             if c not in scheme_chars:
 105                 break
 106         else:
 107             scheme, url = url[:i].lower(), url[i+1:]
 108     if scheme in uses_netloc:
 109         if url[:2] == '//':
 110             i = url.find('/', 2)
 111             if i < 0:
 112                 i = len(url)
 113             netloc, url = url[2:i], url[i:]
 114     if allow_fragments and scheme in uses_fragment and '#' in url:
 115         url, fragment = url.split('#', 1)
 116     if scheme in uses_query and '?' in url:
 117         url, query = url.split('?', 1)
 118     tuple = scheme, netloc, url, query, fragment
 119     _parse_cache[key] = tuple
 120     return tuple
 121
 122 def urlunparse((scheme, netloc, url, params, query, fragment)):
 123     """Put a parsed URL back together again.  This may result in a
 124     slightly different, but equivalent URL, if the URL that was parsed
 125     originally had redundant delimiters, e.g. a ? with an empty query
 126     (the draft states that these are equivalent)."""
 127     if params:
 128         url = "%s;%s" % (url, params)
 129     return urlunsplit((scheme, netloc, url, query, fragment))
 130
 131 def urlunsplit((scheme, netloc, url, query, fragment)):
 132     if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'):
 133         if url and url[:1] != '/': url = '/' + url
 134         url = '//' + (netloc or '') + url
 135     if scheme:
 136         url = scheme + ':' + url
 137     if query:
 138         url = url + '?' + query
 139     if fragment:
 140         url = url + '#' + fragment
 141     return url
 142
 143 def urljoin(base, url, allow_fragments = 1):
 144     """Join a base URL and a possibly relative URL to form an absolute
 145     interpretation of the latter."""
 146     if not base:
 147         return url
 148     if not url:
 149         return base
 150     bscheme, bnetloc, bpath, bparams, bquery, bfragment = \
 151             urlparse(base, '', allow_fragments)
 152     scheme, netloc, path, params, query, fragment = \
 153             urlparse(url, bscheme, allow_fragments)
 154     if scheme != bscheme or scheme not in uses_relative:
 155         return url
 156     if scheme in uses_netloc:
 157         if netloc:
 158             return urlunparse((scheme, netloc, path,
 159                                params, query, fragment))
 160         netloc = bnetloc
 161     if path[:1] == '/':
 162         return urlunparse((scheme, netloc, path,
 163                            params, query, fragment))
 164     if not path:
 165         if not params:
 166             params = bparams
 167             if not query:
 168                 query = bquery
 169         return urlunparse((scheme, netloc, bpath,
 170                            params, query, fragment))
 171     segments = bpath.split('/')[:-1] + path.split('/')
 172     # XXX The stuff below is bogus in various ways...
 173     if segments[-1] == '.':
 174         segments[-1] = ''
 175     while '.' in segments:
 176         segments.remove('.')
 177     while 1:
 178         i = 1
 179         n = len(segments) - 1
 180         while i < n:
 181             if (segments[i] == '..'
 182                 and segments[i-1] not in ('', '..')):
 183                 del segments[i-1:i+1]
 184                 break
 185             i = i+1
 186         else:
 187             break
 188     if segments == ['', '..']:
 189         segments[-1] = ''
 190     elif len(segments) >= 2 and segments[-1] == '..':
 191         segments[-2:] = ['']
 192     return urlunparse((scheme, netloc, '/'.join(segments),
 193                        params, query, fragment))
 194
 195 def urldefrag(url):
 196     """Removes any existing fragment from URL.
 197
 198     Returns a tuple of the defragmented URL and the fragment.  If
 199     the URL contained no fragments, the second element is the
 200     empty string.
 201     """
 202     if '#' in url:
 203         s, n, p, a, q, frag = urlparse(url)
 204         defrag = urlunparse((s, n, p, a, q, ''))
 205         return defrag, frag
 206     else:
 207         return url, ''
 208
 209
 210 test_input = """
 211       http://a/b/c/d
 212
 213       g:h        = <URL:g:h>
 214       http:g     = <URL:http://a/b/c/g>
 215       http:      = <URL:http://a/b/c/d>
 216       g          = <URL:http://a/b/c/g>
 217       ./g        = <URL:http://a/b/c/g>
 218       g/         = <URL:http://a/b/c/g/>
 219       /g         = <URL:http://a/g>
 220       //g        = <URL:http://g>
 221       ?y         = <URL:http://a/b/c/d?y>
 222       g?y        = <URL:http://a/b/c/g?y>
 223       g?y/./x    = <URL:http://a/b/c/g?y/./x>
 224       .          = <URL:http://a/b/c/>
 225       ./         = <URL:http://a/b/c/>
 226       ..         = <URL:http://a/b/>
 227       ../        = <URL:http://a/b/>
 228       ../g       = <URL:http://a/b/g>
 229       ../..      = <URL:http://a/>
 230       ../../g    = <URL:http://a/g>
 231       ../../../g = <URL:http://a/../g>
 232       ./../g     = <URL:http://a/b/g>
 233       ./g/.      = <URL:http://a/b/c/g/>
 234       /./g       = <URL:http://a/./g>
 235       g/./h      = <URL:http://a/b/c/g/h>
 236       g/../h     = <URL:http://a/b/c/h>
 237       http:g     = <URL:http://a/b/c/g>
 238       http:      = <URL:http://a/b/c/d>
 239       http:?y         = <URL:http://a/b/c/d?y>
 240       http:g?y        = <URL:http://a/b/c/g?y>
 241       http:g?y/./x    = <URL:http://a/b/c/g?y/./x>
 242 """
 243 # XXX The result for //g is actually http://g/; is this a problem?
 244
 245 def test():
 246     import sys
 247     base = ''
 248     if sys.argv[1:]:
 249         fn = sys.argv[1]
 250         if fn == '-':
 251             fp = sys.stdin
 252         else:
 253             fp = open(fn)
 254     else:
 255         import StringIO
 256         fp = StringIO.StringIO(test_input)
 257     while 1:
 258         line = fp.readline()
 259         if not line: break
 260         words = line.split()
 261         if not words:
 262             continue
 263         url = words[0]
 264         parts = urlparse(url)
 265         print '%-10s : %s' % (url, parts)
 266         abs = urljoin(base, url)
 267         if not base:
 268             base = abs
 269         wrapped = '<URL:%s>' % abs
 270         print '%-10s = %s' % (url, wrapped)
 271         if len(words) == 3 and words[1] == '=':
 272             if wrapped != words[2]:
 273                 print 'EXPECTED', words[2], '!!!!!!!!!!'
 274
 275 if __name__ == '__main__':
 276     test()