--- /dev/null
+"""Parse (absolute and relative) URLs.\r
+\r
+urlparse module is based upon the following RFC specifications.\r
+\r
+RFC 3986 (STD66): "Uniform Resource Identifiers" by T. Berners-Lee, R. Fielding\r
+and L. Masinter, January 2005.\r
+\r
+RFC 2732 : "Format for Literal IPv6 Addresses in URL's by R.Hinden, B.Carpenter\r
+and L.Masinter, December 1999.\r
+\r
+RFC 2396: "Uniform Resource Identifiers (URI)": Generic Syntax by T.\r
+Berners-Lee, R. Fielding, and L. Masinter, August 1998.\r
+\r
+RFC 2368: "The mailto URL scheme", by P.Hoffman , L Masinter, J. Zwinski, July 1998.\r
+\r
+RFC 1808: "Relative Uniform Resource Locators", by R. Fielding, UC Irvine, June\r
+1995.\r
+\r
+RFC 1738: "Uniform Resource Locators (URL)" by T. Berners-Lee, L. Masinter, M.\r
+McCahill, December 1994\r
+\r
+RFC 3986 is considered the current standard and any future changes to\r
+urlparse module should conform with it. The urlparse module is\r
+currently not entirely compliant with this RFC due to defacto\r
+scenarios for parsing, and for backward compatibility purposes, some\r
+parsing quirks from older RFCs are retained. The testcases in\r
+test_urlparse.py provides a good indicator of parsing behavior.\r
+\r
+"""\r
+\r
+import re\r
+\r
+__all__ = ["urlparse", "urlunparse", "urljoin", "urldefrag",\r
+ "urlsplit", "urlunsplit", "parse_qs", "parse_qsl"]\r
+\r
+# A classification of schemes ('' means apply by default)\r
+uses_relative = ['ftp', 'http', 'gopher', 'nntp', 'imap',\r
+ 'wais', 'file', 'https', 'shttp', 'mms',\r
+ 'prospero', 'rtsp', 'rtspu', '', 'sftp',\r
+ 'svn', 'svn+ssh']\r
+uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet',\r
+ 'imap', 'wais', 'file', 'mms', 'https', 'shttp',\r
+ 'snews', 'prospero', 'rtsp', 'rtspu', 'rsync', '',\r
+ 'svn', 'svn+ssh', 'sftp','nfs','git', 'git+ssh']\r
+uses_params = ['ftp', 'hdl', 'prospero', 'http', 'imap',\r
+ 'https', 'shttp', 'rtsp', 'rtspu', 'sip', 'sips',\r
+ 'mms', '', 'sftp', 'tel']\r
+\r
+# These are not actually used anymore, but should stay for backwards\r
+# compatibility. (They are undocumented, but have a public-looking name.)\r
+non_hierarchical = ['gopher', 'hdl', 'mailto', 'news',\r
+ 'telnet', 'wais', 'imap', 'snews', 'sip', 'sips']\r
+uses_query = ['http', 'wais', 'imap', 'https', 'shttp', 'mms',\r
+ 'gopher', 'rtsp', 'rtspu', 'sip', 'sips', '']\r
+uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news',\r
+ 'nntp', 'wais', 'https', 'shttp', 'snews',\r
+ 'file', 'prospero', '']\r
+\r
+# Characters valid in scheme names\r
+scheme_chars = ('abcdefghijklmnopqrstuvwxyz'\r
+ 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'\r
+ '0123456789'\r
+ '+-.')\r
+\r
+MAX_CACHE_SIZE = 20\r
+_parse_cache = {}\r
+\r
+def clear_cache():\r
+ """Clear the parse cache."""\r
+ _parse_cache.clear()\r
+\r
+\r
+class ResultMixin(object):\r
+ """Shared methods for the parsed result objects."""\r
+\r
+ @property\r
+ def username(self):\r
+ netloc = self.netloc\r
+ if "@" in netloc:\r
+ userinfo = netloc.rsplit("@", 1)[0]\r
+ if ":" in userinfo:\r
+ userinfo = userinfo.split(":", 1)[0]\r
+ return userinfo\r
+ return None\r
+\r
+ @property\r
+ def password(self):\r
+ netloc = self.netloc\r
+ if "@" in netloc:\r
+ userinfo = netloc.rsplit("@", 1)[0]\r
+ if ":" in userinfo:\r
+ return userinfo.split(":", 1)[1]\r
+ return None\r
+\r
+ @property\r
+ def hostname(self):\r
+ netloc = self.netloc.split('@')[-1]\r
+ if '[' in netloc and ']' in netloc:\r
+ return netloc.split(']')[0][1:].lower()\r
+ elif ':' in netloc:\r
+ return netloc.split(':')[0].lower()\r
+ elif netloc == '':\r
+ return None\r
+ else:\r
+ return netloc.lower()\r
+\r
+ @property\r
+ def port(self):\r
+ netloc = self.netloc.split('@')[-1].split(']')[-1]\r
+ if ':' in netloc:\r
+ port = netloc.split(':')[1]\r
+ if port:\r
+ port = int(port, 10)\r
+ # verify legal port\r
+ if (0 <= port <= 65535):\r
+ return port\r
+ return None\r
+\r
+from collections import namedtuple\r
+\r
+class SplitResult(namedtuple('SplitResult', 'scheme netloc path query fragment'), ResultMixin):\r
+\r
+ __slots__ = ()\r
+\r
+ def geturl(self):\r
+ return urlunsplit(self)\r
+\r
+\r
+class ParseResult(namedtuple('ParseResult', 'scheme netloc path params query fragment'), ResultMixin):\r
+\r
+ __slots__ = ()\r
+\r
+ def geturl(self):\r
+ return urlunparse(self)\r
+\r
+\r
+def urlparse(url, scheme='', allow_fragments=True):\r
+ """Parse a URL into 6 components:\r
+ <scheme>://<netloc>/<path>;<params>?<query>#<fragment>\r
+ Return a 6-tuple: (scheme, netloc, path, params, query, fragment).\r
+ Note that we don't break the components up in smaller bits\r
+ (e.g. netloc is a single string) and we don't expand % escapes."""\r
+ tuple = urlsplit(url, scheme, allow_fragments)\r
+ scheme, netloc, url, query, fragment = tuple\r
+ if scheme in uses_params and ';' in url:\r
+ url, params = _splitparams(url)\r
+ else:\r
+ params = ''\r
+ return ParseResult(scheme, netloc, url, params, query, fragment)\r
+\r
+def _splitparams(url):\r
+ if '/' in url:\r
+ i = url.find(';', url.rfind('/'))\r
+ if i < 0:\r
+ return url, ''\r
+ else:\r
+ i = url.find(';')\r
+ return url[:i], url[i+1:]\r
+\r
+def _splitnetloc(url, start=0):\r
+ delim = len(url) # position of end of domain part of url, default is end\r
+ for c in '/?#': # look for delimiters; the order is NOT important\r
+ wdelim = url.find(c, start) # find first of this delim\r
+ if wdelim >= 0: # if found\r
+ delim = min(delim, wdelim) # use earliest delim position\r
+ return url[start:delim], url[delim:] # return (domain, rest)\r
+\r
+def urlsplit(url, scheme='', allow_fragments=True):\r
+ """Parse a URL into 5 components:\r
+ <scheme>://<netloc>/<path>?<query>#<fragment>\r
+ Return a 5-tuple: (scheme, netloc, path, query, fragment).\r
+ Note that we don't break the components up in smaller bits\r
+ (e.g. netloc is a single string) and we don't expand % escapes."""\r
+ allow_fragments = bool(allow_fragments)\r
+ key = url, scheme, allow_fragments, type(url), type(scheme)\r
+ cached = _parse_cache.get(key, None)\r
+ if cached:\r
+ return cached\r
+ if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth\r
+ clear_cache()\r
+ netloc = query = fragment = ''\r
+ i = url.find(':')\r
+ if i > 0:\r
+ if url[:i] == 'http': # optimize the common case\r
+ scheme = url[:i].lower()\r
+ url = url[i+1:]\r
+ if url[:2] == '//':\r
+ netloc, url = _splitnetloc(url, 2)\r
+ if (('[' in netloc and ']' not in netloc) or\r
+ (']' in netloc and '[' not in netloc)):\r
+ raise ValueError("Invalid IPv6 URL")\r
+ if allow_fragments and '#' in url:\r
+ url, fragment = url.split('#', 1)\r
+ if '?' in url:\r
+ url, query = url.split('?', 1)\r
+ v = SplitResult(scheme, netloc, url, query, fragment)\r
+ _parse_cache[key] = v\r
+ return v\r
+ for c in url[:i]:\r
+ if c not in scheme_chars:\r
+ break\r
+ else:\r
+ # make sure "url" is not actually a port number (in which case\r
+ # "scheme" is really part of the path)\r
+ rest = url[i+1:]\r
+ if not rest or any(c not in '0123456789' for c in rest):\r
+ # not a port number\r
+ scheme, url = url[:i].lower(), rest\r
+\r
+ if url[:2] == '//':\r
+ netloc, url = _splitnetloc(url, 2)\r
+ if (('[' in netloc and ']' not in netloc) or\r
+ (']' in netloc and '[' not in netloc)):\r
+ raise ValueError("Invalid IPv6 URL")\r
+ if allow_fragments and '#' in url:\r
+ url, fragment = url.split('#', 1)\r
+ if '?' in url:\r
+ url, query = url.split('?', 1)\r
+ v = SplitResult(scheme, netloc, url, query, fragment)\r
+ _parse_cache[key] = v\r
+ return v\r
+\r
+def urlunparse(data):\r
+ """Put a parsed URL back together again. This may result in a\r
+ slightly different, but equivalent URL, if the URL that was parsed\r
+ originally had redundant delimiters, e.g. a ? with an empty query\r
+ (the draft states that these are equivalent)."""\r
+ scheme, netloc, url, params, query, fragment = data\r
+ if params:\r
+ url = "%s;%s" % (url, params)\r
+ return urlunsplit((scheme, netloc, url, query, fragment))\r
+\r
+def urlunsplit(data):\r
+ """Combine the elements of a tuple as returned by urlsplit() into a\r
+ complete URL as a string. The data argument can be any five-item iterable.\r
+ This may result in a slightly different, but equivalent URL, if the URL that\r
+ was parsed originally had unnecessary delimiters (for example, a ? with an\r
+ empty query; the RFC states that these are equivalent)."""\r
+ scheme, netloc, url, query, fragment = data\r
+ if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'):\r
+ if url and url[:1] != '/': url = '/' + url\r
+ url = '//' + (netloc or '') + url\r
+ if scheme:\r
+ url = scheme + ':' + url\r
+ if query:\r
+ url = url + '?' + query\r
+ if fragment:\r
+ url = url + '#' + fragment\r
+ return url\r
+\r
+def urljoin(base, url, allow_fragments=True):\r
+ """Join a base URL and a possibly relative URL to form an absolute\r
+ interpretation of the latter."""\r
+ if not base:\r
+ return url\r
+ if not url:\r
+ return base\r
+ bscheme, bnetloc, bpath, bparams, bquery, bfragment = \\r
+ urlparse(base, '', allow_fragments)\r
+ scheme, netloc, path, params, query, fragment = \\r
+ urlparse(url, bscheme, allow_fragments)\r
+ if scheme != bscheme or scheme not in uses_relative:\r
+ return url\r
+ if scheme in uses_netloc:\r
+ if netloc:\r
+ return urlunparse((scheme, netloc, path,\r
+ params, query, fragment))\r
+ netloc = bnetloc\r
+ if path[:1] == '/':\r
+ return urlunparse((scheme, netloc, path,\r
+ params, query, fragment))\r
+ if not path and not params:\r
+ path = bpath\r
+ params = bparams\r
+ if not query:\r
+ query = bquery\r
+ return urlunparse((scheme, netloc, path,\r
+ params, query, fragment))\r
+ segments = bpath.split('/')[:-1] + path.split('/')\r
+ # XXX The stuff below is bogus in various ways...\r
+ if segments[-1] == '.':\r
+ segments[-1] = ''\r
+ while '.' in segments:\r
+ segments.remove('.')\r
+ while 1:\r
+ i = 1\r
+ n = len(segments) - 1\r
+ while i < n:\r
+ if (segments[i] == '..'\r
+ and segments[i-1] not in ('', '..')):\r
+ del segments[i-1:i+1]\r
+ break\r
+ i = i+1\r
+ else:\r
+ break\r
+ if segments == ['', '..']:\r
+ segments[-1] = ''\r
+ elif len(segments) >= 2 and segments[-1] == '..':\r
+ segments[-2:] = ['']\r
+ return urlunparse((scheme, netloc, '/'.join(segments),\r
+ params, query, fragment))\r
+\r
+def urldefrag(url):\r
+ """Removes any existing fragment from URL.\r
+\r
+ Returns a tuple of the defragmented URL and the fragment. If\r
+ the URL contained no fragments, the second element is the\r
+ empty string.\r
+ """\r
+ if '#' in url:\r
+ s, n, p, a, q, frag = urlparse(url)\r
+ defrag = urlunparse((s, n, p, a, q, ''))\r
+ return defrag, frag\r
+ else:\r
+ return url, ''\r
+\r
+try:\r
+ unicode\r
+except NameError:\r
+ def _is_unicode(x):\r
+ return 0\r
+else:\r
+ def _is_unicode(x):\r
+ return isinstance(x, unicode)\r
+\r
+# unquote method for parse_qs and parse_qsl\r
+# Cannot use directly from urllib as it would create a circular reference\r
+# because urllib uses urlparse methods (urljoin). If you update this function,\r
+# update it also in urllib. This code duplication does not existin in Python3.\r
+\r
+_hexdig = '0123456789ABCDEFabcdef'\r
+_hextochr = dict((a+b, chr(int(a+b,16)))\r
+ for a in _hexdig for b in _hexdig)\r
+_asciire = re.compile('([\x00-\x7f]+)')\r
+\r
+def unquote(s):\r
+ """unquote('abc%20def') -> 'abc def'."""\r
+ if _is_unicode(s):\r
+ if '%' not in s:\r
+ return s\r
+ bits = _asciire.split(s)\r
+ res = [bits[0]]\r
+ append = res.append\r
+ for i in range(1, len(bits), 2):\r
+ append(unquote(str(bits[i])).decode('latin1'))\r
+ append(bits[i + 1])\r
+ return ''.join(res)\r
+\r
+ bits = s.split('%')\r
+ # fastpath\r
+ if len(bits) == 1:\r
+ return s\r
+ res = [bits[0]]\r
+ append = res.append\r
+ for item in bits[1:]:\r
+ try:\r
+ append(_hextochr[item[:2]])\r
+ append(item[2:])\r
+ except KeyError:\r
+ append('%')\r
+ append(item)\r
+ return ''.join(res)\r
+\r
+def parse_qs(qs, keep_blank_values=0, strict_parsing=0):\r
+ """Parse a query given as a string argument.\r
+\r
+ Arguments:\r
+\r
+ qs: percent-encoded query string to be parsed\r
+\r
+ keep_blank_values: flag indicating whether blank values in\r
+ percent-encoded queries should be treated as blank strings.\r
+ A true value indicates that blanks should be retained as\r
+ blank strings. The default false value indicates that\r
+ blank values are to be ignored and treated as if they were\r
+ not included.\r
+\r
+ strict_parsing: flag indicating what to do with parsing errors.\r
+ If false (the default), errors are silently ignored.\r
+ If true, errors raise a ValueError exception.\r
+ """\r
+ dict = {}\r
+ for name, value in parse_qsl(qs, keep_blank_values, strict_parsing):\r
+ if name in dict:\r
+ dict[name].append(value)\r
+ else:\r
+ dict[name] = [value]\r
+ return dict\r
+\r
+def parse_qsl(qs, keep_blank_values=0, strict_parsing=0):\r
+ """Parse a query given as a string argument.\r
+\r
+ Arguments:\r
+\r
+ qs: percent-encoded query string to be parsed\r
+\r
+ keep_blank_values: flag indicating whether blank values in\r
+ percent-encoded queries should be treated as blank strings. A\r
+ true value indicates that blanks should be retained as blank\r
+ strings. The default false value indicates that blank values\r
+ are to be ignored and treated as if they were not included.\r
+\r
+ strict_parsing: flag indicating what to do with parsing errors. If\r
+ false (the default), errors are silently ignored. If true,\r
+ errors raise a ValueError exception.\r
+\r
+ Returns a list, as G-d intended.\r
+ """\r
+ pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]\r
+ r = []\r
+ for name_value in pairs:\r
+ if not name_value and not strict_parsing:\r
+ continue\r
+ nv = name_value.split('=', 1)\r
+ if len(nv) != 2:\r
+ if strict_parsing:\r
+ raise ValueError, "bad query field: %r" % (name_value,)\r
+ # Handle case of a control-name with no equal sign\r
+ if keep_blank_values:\r
+ nv.append('')\r
+ else:\r
+ continue\r
+ if len(nv[1]) or keep_blank_values:\r
+ name = unquote(nv[0].replace('+', ' '))\r
+ value = unquote(nv[1].replace('+', ' '))\r
+ r.append((name, value))\r
+\r
+ return r\r