| 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', '', 'sftp']
|
|---|
| 14 | uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet',
|
|---|
| 15 | 'imap', 'wais', 'file', 'mms', 'https', 'shttp',
|
|---|
| 16 | 'snews', 'prospero', 'rtsp', 'rtspu', 'rsync', '',
|
|---|
| 17 | 'svn', 'svn+ssh', 'sftp']
|
|---|
| 18 | non_hierarchical = ['gopher', 'hdl', 'mailto', 'news',
|
|---|
| 19 | 'telnet', 'wais', 'imap', 'snews', 'sip', 'sips']
|
|---|
| 20 | uses_params = ['ftp', 'hdl', 'prospero', 'http', 'imap',
|
|---|
| 21 | 'https', 'shttp', 'rtsp', 'rtspu', 'sip', 'sips',
|
|---|
| 22 | 'mms', '', 'sftp']
|
|---|
| 23 | uses_query = ['http', 'wais', 'imap', 'https', 'shttp', 'mms',
|
|---|
| 24 | 'gopher', 'rtsp', 'rtspu', 'sip', 'sips', '']
|
|---|
| 25 | uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news',
|
|---|
| 26 | 'nntp', 'wais', 'https', 'shttp', 'snews',
|
|---|
| 27 | 'file', 'prospero', '']
|
|---|
| 28 |
|
|---|
| 29 | # Characters valid in scheme names
|
|---|
| 30 | scheme_chars = ('abcdefghijklmnopqrstuvwxyz'
|
|---|
| 31 | 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'
|
|---|
| 32 | '0123456789'
|
|---|
| 33 | '+-.')
|
|---|
| 34 |
|
|---|
| 35 | MAX_CACHE_SIZE = 20
|
|---|
| 36 | _parse_cache = {}
|
|---|
| 37 |
|
|---|
| 38 | def clear_cache():
|
|---|
| 39 | """Clear the parse cache."""
|
|---|
| 40 | global _parse_cache
|
|---|
| 41 | _parse_cache = {}
|
|---|
| 42 |
|
|---|
| 43 |
|
|---|
| 44 | class BaseResult(tuple):
|
|---|
| 45 | """Base class for the parsed result objects.
|
|---|
| 46 |
|
|---|
| 47 | This provides the attributes shared by the two derived result
|
|---|
| 48 | objects as read-only properties. The derived classes are
|
|---|
| 49 | responsible for checking the right number of arguments were
|
|---|
| 50 | supplied to the constructor.
|
|---|
| 51 |
|
|---|
| 52 | """
|
|---|
| 53 |
|
|---|
| 54 | __slots__ = ()
|
|---|
| 55 |
|
|---|
| 56 | # Attributes that access the basic components of the URL:
|
|---|
| 57 |
|
|---|
| 58 | @property
|
|---|
| 59 | def scheme(self):
|
|---|
| 60 | return self[0]
|
|---|
| 61 |
|
|---|
| 62 | @property
|
|---|
| 63 | def netloc(self):
|
|---|
| 64 | return self[1]
|
|---|
| 65 |
|
|---|
| 66 | @property
|
|---|
| 67 | def path(self):
|
|---|
| 68 | return self[2]
|
|---|
| 69 |
|
|---|
| 70 | @property
|
|---|
| 71 | def query(self):
|
|---|
| 72 | return self[-2]
|
|---|
| 73 |
|
|---|
| 74 | @property
|
|---|
| 75 | def fragment(self):
|
|---|
| 76 | return self[-1]
|
|---|
| 77 |
|
|---|
| 78 | # Additional attributes that provide access to parsed-out portions
|
|---|
| 79 | # of the netloc:
|
|---|
| 80 |
|
|---|
| 81 | @property
|
|---|
| 82 | def username(self):
|
|---|
| 83 | netloc = self.netloc
|
|---|
| 84 | if "@" in netloc:
|
|---|
| 85 | userinfo = netloc.split("@", 1)[0]
|
|---|
| 86 | if ":" in userinfo:
|
|---|
| 87 | userinfo = userinfo.split(":", 1)[0]
|
|---|
| 88 | return userinfo
|
|---|
| 89 | return None
|
|---|
| 90 |
|
|---|
| 91 | @property
|
|---|
| 92 | def password(self):
|
|---|
| 93 | netloc = self.netloc
|
|---|
| 94 | if "@" in netloc:
|
|---|
| 95 | userinfo = netloc.split("@", 1)[0]
|
|---|
| 96 | if ":" in userinfo:
|
|---|
| 97 | return userinfo.split(":", 1)[1]
|
|---|
| 98 | return None
|
|---|
| 99 |
|
|---|
| 100 | @property
|
|---|
| 101 | def hostname(self):
|
|---|
| 102 | netloc = self.netloc
|
|---|
| 103 | if "@" in netloc:
|
|---|
| 104 | netloc = netloc.split("@", 1)[1]
|
|---|
| 105 | if ":" in netloc:
|
|---|
| 106 | netloc = netloc.split(":", 1)[0]
|
|---|
| 107 | return netloc.lower() or None
|
|---|
| 108 |
|
|---|
| 109 | @property
|
|---|
| 110 | def port(self):
|
|---|
| 111 | netloc = self.netloc
|
|---|
| 112 | if "@" in netloc:
|
|---|
| 113 | netloc = netloc.split("@", 1)[1]
|
|---|
| 114 | if ":" in netloc:
|
|---|
| 115 | port = netloc.split(":", 1)[1]
|
|---|
| 116 | return int(port, 10)
|
|---|
| 117 | return None
|
|---|
| 118 |
|
|---|
| 119 |
|
|---|
| 120 | class SplitResult(BaseResult):
|
|---|
| 121 |
|
|---|
| 122 | __slots__ = ()
|
|---|
| 123 |
|
|---|
| 124 | def __new__(cls, scheme, netloc, path, query, fragment):
|
|---|
| 125 | return BaseResult.__new__(
|
|---|
| 126 | cls, (scheme, netloc, path, query, fragment))
|
|---|
| 127 |
|
|---|
| 128 | def geturl(self):
|
|---|
| 129 | return urlunsplit(self)
|
|---|
| 130 |
|
|---|
| 131 |
|
|---|
| 132 | class ParseResult(BaseResult):
|
|---|
| 133 |
|
|---|
| 134 | __slots__ = ()
|
|---|
| 135 |
|
|---|
| 136 | def __new__(cls, scheme, netloc, path, params, query, fragment):
|
|---|
| 137 | return BaseResult.__new__(
|
|---|
| 138 | cls, (scheme, netloc, path, params, query, fragment))
|
|---|
| 139 |
|
|---|
| 140 | @property
|
|---|
| 141 | def params(self):
|
|---|
| 142 | return self[3]
|
|---|
| 143 |
|
|---|
| 144 | def geturl(self):
|
|---|
| 145 | return urlunparse(self)
|
|---|
| 146 |
|
|---|
| 147 |
|
|---|
| 148 | def urlparse(url, scheme='', allow_fragments=True):
|
|---|
| 149 | """Parse a URL into 6 components:
|
|---|
| 150 | <scheme>://<netloc>/<path>;<params>?<query>#<fragment>
|
|---|
| 151 | Return a 6-tuple: (scheme, netloc, path, params, query, fragment).
|
|---|
| 152 | Note that we don't break the components up in smaller bits
|
|---|
| 153 | (e.g. netloc is a single string) and we don't expand % escapes."""
|
|---|
| 154 | tuple = urlsplit(url, scheme, allow_fragments)
|
|---|
| 155 | scheme, netloc, url, query, fragment = tuple
|
|---|
| 156 | if scheme in uses_params and ';' in url:
|
|---|
| 157 | url, params = _splitparams(url)
|
|---|
| 158 | else:
|
|---|
| 159 | params = ''
|
|---|
| 160 | return ParseResult(scheme, netloc, url, params, query, fragment)
|
|---|
| 161 |
|
|---|
| 162 | def _splitparams(url):
|
|---|
| 163 | if '/' in url:
|
|---|
| 164 | i = url.find(';', url.rfind('/'))
|
|---|
| 165 | if i < 0:
|
|---|
| 166 | return url, ''
|
|---|
| 167 | else:
|
|---|
| 168 | i = url.find(';')
|
|---|
| 169 | return url[:i], url[i+1:]
|
|---|
| 170 |
|
|---|
| 171 | def _splitnetloc(url, start=0):
|
|---|
| 172 | for c in '/?#': # the order is important!
|
|---|
| 173 | delim = url.find(c, start)
|
|---|
| 174 | if delim >= 0:
|
|---|
| 175 | break
|
|---|
| 176 | else:
|
|---|
| 177 | delim = len(url)
|
|---|
| 178 | return url[start:delim], url[delim:]
|
|---|
| 179 |
|
|---|
| 180 | def urlsplit(url, scheme='', allow_fragments=True):
|
|---|
| 181 | """Parse a URL into 5 components:
|
|---|
| 182 | <scheme>://<netloc>/<path>?<query>#<fragment>
|
|---|
| 183 | Return a 5-tuple: (scheme, netloc, path, query, fragment).
|
|---|
| 184 | Note that we don't break the components up in smaller bits
|
|---|
| 185 | (e.g. netloc is a single string) and we don't expand % escapes."""
|
|---|
| 186 | allow_fragments = bool(allow_fragments)
|
|---|
| 187 | key = url, scheme, allow_fragments
|
|---|
| 188 | cached = _parse_cache.get(key, None)
|
|---|
| 189 | if cached:
|
|---|
| 190 | return cached
|
|---|
| 191 | if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth
|
|---|
| 192 | clear_cache()
|
|---|
| 193 | netloc = query = fragment = ''
|
|---|
| 194 | i = url.find(':')
|
|---|
| 195 | if i > 0:
|
|---|
| 196 | if url[:i] == 'http': # optimize the common case
|
|---|
| 197 | scheme = url[:i].lower()
|
|---|
| 198 | url = url[i+1:]
|
|---|
| 199 | if url[:2] == '//':
|
|---|
| 200 | netloc, url = _splitnetloc(url, 2)
|
|---|
| 201 | if allow_fragments and '#' in url:
|
|---|
| 202 | url, fragment = url.split('#', 1)
|
|---|
| 203 | if '?' in url:
|
|---|
| 204 | url, query = url.split('?', 1)
|
|---|
| 205 | v = SplitResult(scheme, netloc, url, query, fragment)
|
|---|
| 206 | _parse_cache[key] = v
|
|---|
| 207 | return v
|
|---|
| 208 | for c in url[:i]:
|
|---|
| 209 | if c not in scheme_chars:
|
|---|
| 210 | break
|
|---|
| 211 | else:
|
|---|
| 212 | scheme, url = url[:i].lower(), url[i+1:]
|
|---|
| 213 | if scheme in uses_netloc and url[:2] == '//':
|
|---|
| 214 | netloc, url = _splitnetloc(url, 2)
|
|---|
| 215 | if allow_fragments and scheme in uses_fragment and '#' in url:
|
|---|
| 216 | url, fragment = url.split('#', 1)
|
|---|
| 217 | if scheme in uses_query and '?' in url:
|
|---|
| 218 | url, query = url.split('?', 1)
|
|---|
| 219 | v = SplitResult(scheme, netloc, url, query, fragment)
|
|---|
| 220 | _parse_cache[key] = v
|
|---|
| 221 | return v
|
|---|
| 222 |
|
|---|
| 223 | def urlunparse((scheme, netloc, url, params, query, fragment)):
|
|---|
| 224 | """Put a parsed URL back together again. This may result in a
|
|---|
| 225 | slightly different, but equivalent URL, if the URL that was parsed
|
|---|
| 226 | originally had redundant delimiters, e.g. a ? with an empty query
|
|---|
| 227 | (the draft states that these are equivalent)."""
|
|---|
| 228 | if params:
|
|---|
| 229 | url = "%s;%s" % (url, params)
|
|---|
| 230 | return urlunsplit((scheme, netloc, url, query, fragment))
|
|---|
| 231 |
|
|---|
| 232 | def urlunsplit((scheme, netloc, url, query, fragment)):
|
|---|
| 233 | if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'):
|
|---|
| 234 | if url and url[:1] != '/': url = '/' + url
|
|---|
| 235 | url = '//' + (netloc or '') + url
|
|---|
| 236 | if scheme:
|
|---|
| 237 | url = scheme + ':' + url
|
|---|
| 238 | if query:
|
|---|
| 239 | url = url + '?' + query
|
|---|
| 240 | if fragment:
|
|---|
| 241 | url = url + '#' + fragment
|
|---|
| 242 | return url
|
|---|
| 243 |
|
|---|
| 244 | def urljoin(base, url, allow_fragments=True):
|
|---|
| 245 | """Join a base URL and a possibly relative URL to form an absolute
|
|---|
| 246 | interpretation of the latter."""
|
|---|
| 247 | if not base:
|
|---|
| 248 | return url
|
|---|
| 249 | if not url:
|
|---|
| 250 | return base
|
|---|
| 251 | bscheme, bnetloc, bpath, bparams, bquery, bfragment = \
|
|---|
| 252 | urlparse(base, '', allow_fragments)
|
|---|
| 253 | scheme, netloc, path, params, query, fragment = \
|
|---|
| 254 | urlparse(url, bscheme, allow_fragments)
|
|---|
| 255 | if scheme != bscheme or scheme not in uses_relative:
|
|---|
| 256 | return url
|
|---|
| 257 | if scheme in uses_netloc:
|
|---|
| 258 | if netloc:
|
|---|
| 259 | return urlunparse((scheme, netloc, path,
|
|---|
| 260 | params, query, fragment))
|
|---|
| 261 | netloc = bnetloc
|
|---|
| 262 | if path[:1] == '/':
|
|---|
| 263 | return urlunparse((scheme, netloc, path,
|
|---|
| 264 | params, query, fragment))
|
|---|
| 265 | if not (path or params or query):
|
|---|
| 266 | return urlunparse((scheme, netloc, bpath,
|
|---|
| 267 | bparams, bquery, fragment))
|
|---|
| 268 | segments = bpath.split('/')[:-1] + path.split('/')
|
|---|
| 269 | # XXX The stuff below is bogus in various ways...
|
|---|
| 270 | if segments[-1] == '.':
|
|---|
| 271 | segments[-1] = ''
|
|---|
| 272 | while '.' in segments:
|
|---|
| 273 | segments.remove('.')
|
|---|
| 274 | while 1:
|
|---|
| 275 | i = 1
|
|---|
| 276 | n = len(segments) - 1
|
|---|
| 277 | while i < n:
|
|---|
| 278 | if (segments[i] == '..'
|
|---|
| 279 | and segments[i-1] not in ('', '..')):
|
|---|
| 280 | del segments[i-1:i+1]
|
|---|
| 281 | break
|
|---|
| 282 | i = i+1
|
|---|
| 283 | else:
|
|---|
| 284 | break
|
|---|
| 285 | if segments == ['', '..']:
|
|---|
| 286 | segments[-1] = ''
|
|---|
| 287 | elif len(segments) >= 2 and segments[-1] == '..':
|
|---|
| 288 | segments[-2:] = ['']
|
|---|
| 289 | return urlunparse((scheme, netloc, '/'.join(segments),
|
|---|
| 290 | params, query, fragment))
|
|---|
| 291 |
|
|---|
| 292 | def urldefrag(url):
|
|---|
| 293 | """Removes any existing fragment from URL.
|
|---|
| 294 |
|
|---|
| 295 | Returns a tuple of the defragmented URL and the fragment. If
|
|---|
| 296 | the URL contained no fragments, the second element is the
|
|---|
| 297 | empty string.
|
|---|
| 298 | """
|
|---|
| 299 | if '#' in url:
|
|---|
| 300 | s, n, p, a, q, frag = urlparse(url)
|
|---|
| 301 | defrag = urlunparse((s, n, p, a, q, ''))
|
|---|
| 302 | return defrag, frag
|
|---|
| 303 | else:
|
|---|
| 304 | return url, ''
|
|---|
| 305 |
|
|---|
| 306 |
|
|---|
| 307 | test_input = """
|
|---|
| 308 | http://a/b/c/d
|
|---|
| 309 |
|
|---|
| 310 | g:h = <URL:g:h>
|
|---|
| 311 | http:g = <URL:http://a/b/c/g>
|
|---|
| 312 | http: = <URL:http://a/b/c/d>
|
|---|
| 313 | g = <URL:http://a/b/c/g>
|
|---|
| 314 | ./g = <URL:http://a/b/c/g>
|
|---|
| 315 | g/ = <URL:http://a/b/c/g/>
|
|---|
| 316 | /g = <URL:http://a/g>
|
|---|
| 317 | //g = <URL:http://g>
|
|---|
| 318 | ?y = <URL:http://a/b/c/d?y>
|
|---|
| 319 | g?y = <URL:http://a/b/c/g?y>
|
|---|
| 320 | g?y/./x = <URL:http://a/b/c/g?y/./x>
|
|---|
| 321 | . = <URL:http://a/b/c/>
|
|---|
| 322 | ./ = <URL:http://a/b/c/>
|
|---|
| 323 | .. = <URL:http://a/b/>
|
|---|
| 324 | ../ = <URL:http://a/b/>
|
|---|
| 325 | ../g = <URL:http://a/b/g>
|
|---|
| 326 | ../.. = <URL:http://a/>
|
|---|
| 327 | ../../g = <URL:http://a/g>
|
|---|
| 328 | ../../../g = <URL:http://a/../g>
|
|---|
| 329 | ./../g = <URL:http://a/b/g>
|
|---|
| 330 | ./g/. = <URL:http://a/b/c/g/>
|
|---|
| 331 | /./g = <URL:http://a/./g>
|
|---|
| 332 | g/./h = <URL:http://a/b/c/g/h>
|
|---|
| 333 | g/../h = <URL:http://a/b/c/h>
|
|---|
| 334 | http:g = <URL:http://a/b/c/g>
|
|---|
| 335 | http: = <URL:http://a/b/c/d>
|
|---|
| 336 | http:?y = <URL:http://a/b/c/d?y>
|
|---|
| 337 | http:g?y = <URL:http://a/b/c/g?y>
|
|---|
| 338 | http:g?y/./x = <URL:http://a/b/c/g?y/./x>
|
|---|
| 339 | """
|
|---|
| 340 |
|
|---|
| 341 | def test():
|
|---|
| 342 | import sys
|
|---|
| 343 | base = ''
|
|---|
| 344 | if sys.argv[1:]:
|
|---|
| 345 | fn = sys.argv[1]
|
|---|
| 346 | if fn == '-':
|
|---|
| 347 | fp = sys.stdin
|
|---|
| 348 | else:
|
|---|
| 349 | fp = open(fn)
|
|---|
| 350 | else:
|
|---|
| 351 | try:
|
|---|
| 352 | from cStringIO import StringIO
|
|---|
| 353 | except ImportError:
|
|---|
| 354 | from StringIO import StringIO
|
|---|
| 355 | fp = StringIO(test_input)
|
|---|
| 356 | while 1:
|
|---|
| 357 | line = fp.readline()
|
|---|
| 358 | if not line: break
|
|---|
| 359 | words = line.split()
|
|---|
| 360 | if not words:
|
|---|
| 361 | continue
|
|---|
| 362 | url = words[0]
|
|---|
| 363 | parts = urlparse(url)
|
|---|
| 364 | print '%-10s : %s' % (url, parts)
|
|---|
| 365 | abs = urljoin(base, url)
|
|---|
| 366 | if not base:
|
|---|
| 367 | base = abs
|
|---|
| 368 | wrapped = '<URL:%s>' % abs
|
|---|
| 369 | print '%-10s = %s' % (url, wrapped)
|
|---|
| 370 | if len(words) == 3 and words[1] == '=':
|
|---|
| 371 | if wrapped != words[2]:
|
|---|
| 372 | print 'EXPECTED', words[2], '!!!!!!!!!!'
|
|---|
| 373 |
|
|---|
| 374 | if __name__ == '__main__':
|
|---|
| 375 | test()
|
|---|
