Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 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 | |
Facundo Batista | 2ac5de2 | 2008-07-07 18:24:11 +0000 | [diff] [blame] | 7 | import sys |
| 8 | |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 9 | __all__ = ["urlparse", "urlunparse", "urljoin", "urldefrag", |
| 10 | "urlsplit", "urlunsplit"] |
| 11 | |
| 12 | # A classification of schemes ('' means apply by default) |
| 13 | uses_relative = ['ftp', 'http', 'gopher', 'nntp', 'imap', |
| 14 | 'wais', 'file', 'https', 'shttp', 'mms', |
| 15 | 'prospero', 'rtsp', 'rtspu', '', 'sftp'] |
| 16 | uses_netloc = ['ftp', 'http', 'gopher', 'nntp', 'telnet', |
| 17 | 'imap', 'wais', 'file', 'mms', 'https', 'shttp', |
| 18 | 'snews', 'prospero', 'rtsp', 'rtspu', 'rsync', '', |
| 19 | 'svn', 'svn+ssh', 'sftp'] |
| 20 | non_hierarchical = ['gopher', 'hdl', 'mailto', 'news', |
| 21 | 'telnet', 'wais', 'imap', 'snews', 'sip', 'sips'] |
| 22 | uses_params = ['ftp', 'hdl', 'prospero', 'http', 'imap', |
| 23 | 'https', 'shttp', 'rtsp', 'rtspu', 'sip', 'sips', |
| 24 | 'mms', '', 'sftp'] |
| 25 | uses_query = ['http', 'wais', 'imap', 'https', 'shttp', 'mms', |
| 26 | 'gopher', 'rtsp', 'rtspu', 'sip', 'sips', ''] |
| 27 | uses_fragment = ['ftp', 'hdl', 'http', 'gopher', 'news', |
| 28 | 'nntp', 'wais', 'https', 'shttp', 'snews', |
| 29 | 'file', 'prospero', ''] |
| 30 | |
| 31 | # Characters valid in scheme names |
| 32 | scheme_chars = ('abcdefghijklmnopqrstuvwxyz' |
| 33 | 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' |
| 34 | '0123456789' |
| 35 | '+-.') |
| 36 | |
| 37 | MAX_CACHE_SIZE = 20 |
| 38 | _parse_cache = {} |
| 39 | |
| 40 | def clear_cache(): |
| 41 | """Clear the parse cache.""" |
| 42 | _parse_cache.clear() |
| 43 | |
| 44 | |
| 45 | class ResultMixin(object): |
| 46 | """Shared methods for the parsed result objects.""" |
| 47 | |
| 48 | @property |
| 49 | def username(self): |
| 50 | netloc = self.netloc |
| 51 | if "@" in netloc: |
| 52 | userinfo = netloc.rsplit("@", 1)[0] |
| 53 | if ":" in userinfo: |
| 54 | userinfo = userinfo.split(":", 1)[0] |
| 55 | return userinfo |
| 56 | return None |
| 57 | |
| 58 | @property |
| 59 | def password(self): |
| 60 | netloc = self.netloc |
| 61 | if "@" in netloc: |
| 62 | userinfo = netloc.rsplit("@", 1)[0] |
| 63 | if ":" in userinfo: |
| 64 | return userinfo.split(":", 1)[1] |
| 65 | return None |
| 66 | |
| 67 | @property |
| 68 | def hostname(self): |
| 69 | netloc = self.netloc |
| 70 | if "@" in netloc: |
| 71 | netloc = netloc.rsplit("@", 1)[1] |
| 72 | if ":" in netloc: |
| 73 | netloc = netloc.split(":", 1)[0] |
| 74 | return netloc.lower() or None |
| 75 | |
| 76 | @property |
| 77 | def port(self): |
| 78 | netloc = self.netloc |
| 79 | if "@" in netloc: |
| 80 | netloc = netloc.rsplit("@", 1)[1] |
| 81 | if ":" in netloc: |
| 82 | port = netloc.split(":", 1)[1] |
| 83 | return int(port, 10) |
| 84 | return None |
| 85 | |
| 86 | from collections import namedtuple |
| 87 | |
| 88 | class SplitResult(namedtuple('SplitResult', 'scheme netloc path query fragment'), ResultMixin): |
| 89 | |
| 90 | __slots__ = () |
| 91 | |
| 92 | def geturl(self): |
| 93 | return urlunsplit(self) |
| 94 | |
| 95 | |
| 96 | class ParseResult(namedtuple('ParseResult', 'scheme netloc path params query fragment'), ResultMixin): |
| 97 | |
| 98 | __slots__ = () |
| 99 | |
| 100 | def geturl(self): |
| 101 | return urlunparse(self) |
| 102 | |
| 103 | |
| 104 | def urlparse(url, scheme='', allow_fragments=True): |
| 105 | """Parse a URL into 6 components: |
| 106 | <scheme>://<netloc>/<path>;<params>?<query>#<fragment> |
| 107 | Return a 6-tuple: (scheme, netloc, path, params, query, fragment). |
| 108 | Note that we don't break the components up in smaller bits |
| 109 | (e.g. netloc is a single string) and we don't expand % escapes.""" |
| 110 | tuple = urlsplit(url, scheme, allow_fragments) |
| 111 | scheme, netloc, url, query, fragment = tuple |
| 112 | if scheme in uses_params and ';' in url: |
| 113 | url, params = _splitparams(url) |
| 114 | else: |
| 115 | params = '' |
| 116 | return ParseResult(scheme, netloc, url, params, query, fragment) |
| 117 | |
| 118 | def _splitparams(url): |
| 119 | if '/' in url: |
| 120 | i = url.find(';', url.rfind('/')) |
| 121 | if i < 0: |
| 122 | return url, '' |
| 123 | else: |
| 124 | i = url.find(';') |
| 125 | return url[:i], url[i+1:] |
| 126 | |
| 127 | def _splitnetloc(url, start=0): |
| 128 | delim = len(url) # position of end of domain part of url, default is end |
| 129 | for c in '/?#': # look for delimiters; the order is NOT important |
| 130 | wdelim = url.find(c, start) # find first of this delim |
| 131 | if wdelim >= 0: # if found |
| 132 | delim = min(delim, wdelim) # use earliest delim position |
| 133 | return url[start:delim], url[delim:] # return (domain, rest) |
| 134 | |
| 135 | def urlsplit(url, scheme='', allow_fragments=True): |
| 136 | """Parse a URL into 5 components: |
| 137 | <scheme>://<netloc>/<path>?<query>#<fragment> |
| 138 | Return a 5-tuple: (scheme, netloc, path, query, fragment). |
| 139 | Note that we don't break the components up in smaller bits |
| 140 | (e.g. netloc is a single string) and we don't expand % escapes.""" |
| 141 | allow_fragments = bool(allow_fragments) |
| 142 | key = url, scheme, allow_fragments, type(url), type(scheme) |
| 143 | cached = _parse_cache.get(key, None) |
| 144 | if cached: |
| 145 | return cached |
| 146 | if len(_parse_cache) >= MAX_CACHE_SIZE: # avoid runaway growth |
| 147 | clear_cache() |
| 148 | netloc = query = fragment = '' |
| 149 | i = url.find(':') |
| 150 | if i > 0: |
| 151 | if url[:i] == 'http': # optimize the common case |
| 152 | scheme = url[:i].lower() |
| 153 | url = url[i+1:] |
| 154 | if url[:2] == '//': |
| 155 | netloc, url = _splitnetloc(url, 2) |
| 156 | if allow_fragments and '#' in url: |
| 157 | url, fragment = url.split('#', 1) |
| 158 | if '?' in url: |
| 159 | url, query = url.split('?', 1) |
| 160 | v = SplitResult(scheme, netloc, url, query, fragment) |
| 161 | _parse_cache[key] = v |
| 162 | return v |
| 163 | for c in url[:i]: |
| 164 | if c not in scheme_chars: |
| 165 | break |
| 166 | else: |
| 167 | scheme, url = url[:i].lower(), url[i+1:] |
| 168 | if scheme in uses_netloc and url[:2] == '//': |
| 169 | netloc, url = _splitnetloc(url, 2) |
| 170 | if allow_fragments and scheme in uses_fragment and '#' in url: |
| 171 | url, fragment = url.split('#', 1) |
| 172 | if scheme in uses_query and '?' in url: |
| 173 | url, query = url.split('?', 1) |
| 174 | v = SplitResult(scheme, netloc, url, query, fragment) |
| 175 | _parse_cache[key] = v |
| 176 | return v |
| 177 | |
| 178 | def urlunparse(components): |
| 179 | """Put a parsed URL back together again. This may result in a |
| 180 | slightly different, but equivalent URL, if the URL that was parsed |
| 181 | originally had redundant delimiters, e.g. a ? with an empty query |
| 182 | (the draft states that these are equivalent).""" |
| 183 | scheme, netloc, url, params, query, fragment = components |
| 184 | if params: |
| 185 | url = "%s;%s" % (url, params) |
| 186 | return urlunsplit((scheme, netloc, url, query, fragment)) |
| 187 | |
| 188 | def urlunsplit(components): |
| 189 | scheme, netloc, url, query, fragment = components |
| 190 | if netloc or (scheme and scheme in uses_netloc and url[:2] != '//'): |
| 191 | if url and url[:1] != '/': url = '/' + url |
| 192 | url = '//' + (netloc or '') + url |
| 193 | if scheme: |
| 194 | url = scheme + ':' + url |
| 195 | if query: |
| 196 | url = url + '?' + query |
| 197 | if fragment: |
| 198 | url = url + '#' + fragment |
| 199 | return url |
| 200 | |
| 201 | def urljoin(base, url, allow_fragments=True): |
| 202 | """Join a base URL and a possibly relative URL to form an absolute |
| 203 | interpretation of the latter.""" |
| 204 | if not base: |
| 205 | return url |
| 206 | if not url: |
| 207 | return base |
| 208 | bscheme, bnetloc, bpath, bparams, bquery, bfragment = \ |
| 209 | urlparse(base, '', allow_fragments) |
| 210 | scheme, netloc, path, params, query, fragment = \ |
| 211 | urlparse(url, bscheme, allow_fragments) |
| 212 | if scheme != bscheme or scheme not in uses_relative: |
| 213 | return url |
| 214 | if scheme in uses_netloc: |
| 215 | if netloc: |
| 216 | return urlunparse((scheme, netloc, path, |
| 217 | params, query, fragment)) |
| 218 | netloc = bnetloc |
| 219 | if path[:1] == '/': |
| 220 | return urlunparse((scheme, netloc, path, |
| 221 | params, query, fragment)) |
| 222 | if not (path or params or query): |
| 223 | return urlunparse((scheme, netloc, bpath, |
| 224 | bparams, bquery, fragment)) |
| 225 | segments = bpath.split('/')[:-1] + path.split('/') |
| 226 | # XXX The stuff below is bogus in various ways... |
| 227 | if segments[-1] == '.': |
| 228 | segments[-1] = '' |
| 229 | while '.' in segments: |
| 230 | segments.remove('.') |
| 231 | while 1: |
| 232 | i = 1 |
| 233 | n = len(segments) - 1 |
| 234 | while i < n: |
| 235 | if (segments[i] == '..' |
| 236 | and segments[i-1] not in ('', '..')): |
| 237 | del segments[i-1:i+1] |
| 238 | break |
| 239 | i = i+1 |
| 240 | else: |
| 241 | break |
| 242 | if segments == ['', '..']: |
| 243 | segments[-1] = '' |
| 244 | elif len(segments) >= 2 and segments[-1] == '..': |
| 245 | segments[-2:] = [''] |
| 246 | return urlunparse((scheme, netloc, '/'.join(segments), |
| 247 | params, query, fragment)) |
| 248 | |
| 249 | def urldefrag(url): |
| 250 | """Removes any existing fragment from URL. |
| 251 | |
| 252 | Returns a tuple of the defragmented URL and the fragment. If |
| 253 | the URL contained no fragments, the second element is the |
| 254 | empty string. |
| 255 | """ |
| 256 | if '#' in url: |
| 257 | s, n, p, a, q, frag = urlparse(url) |
| 258 | defrag = urlunparse((s, n, p, a, q, '')) |
| 259 | return defrag, frag |
| 260 | else: |
| 261 | return url, '' |
| 262 | |
| 263 | |
Guido van Rossum | df9f1ec | 2008-08-06 19:31:34 +0000 | [diff] [blame^] | 264 | _hextochr = dict(('%02x' % i, chr(i)) for i in range(256)) |
| 265 | _hextochr.update(('%02X' % i, chr(i)) for i in range(256)) |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 266 | |
| 267 | def unquote(s): |
Guido van Rossum | df9f1ec | 2008-08-06 19:31:34 +0000 | [diff] [blame^] | 268 | """unquote('abc%20def') -> 'abc def'.""" |
| 269 | res = s.split('%') |
| 270 | for i in range(1, len(res)): |
| 271 | item = res[i] |
| 272 | try: |
| 273 | res[i] = _hextochr[item[:2]] + item[2:] |
| 274 | except KeyError: |
| 275 | res[i] = '%' + item |
| 276 | except UnicodeDecodeError: |
| 277 | res[i] = chr(int(item[:2], 16)) + item[2:] |
| 278 | return "".join(res) |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 279 | |
| 280 | def unquote_plus(s): |
Guido van Rossum | df9f1ec | 2008-08-06 19:31:34 +0000 | [diff] [blame^] | 281 | """unquote('%7e/abc+def') -> '~/abc def'""" |
| 282 | s = s.replace('+', ' ') |
| 283 | return unquote(s) |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 284 | |
Guido van Rossum | df9f1ec | 2008-08-06 19:31:34 +0000 | [diff] [blame^] | 285 | always_safe = ('ABCDEFGHIJKLMNOPQRSTUVWXYZ' |
| 286 | 'abcdefghijklmnopqrstuvwxyz' |
| 287 | '0123456789' '_.-') |
| 288 | _safe_quoters= {} |
| 289 | |
| 290 | class Quoter: |
| 291 | def __init__(self, safe): |
| 292 | self.cache = {} |
| 293 | self.safe = safe + always_safe |
| 294 | |
| 295 | def __call__(self, c): |
| 296 | try: |
| 297 | return self.cache[c] |
| 298 | except KeyError: |
| 299 | if ord(c) < 256: |
| 300 | res = (c in self.safe) and c or ('%%%02X' % ord(c)) |
| 301 | self.cache[c] = res |
| 302 | return res |
| 303 | else: |
| 304 | return "".join(['%%%02X' % i for i in c.encode("utf-8")]) |
| 305 | |
| 306 | def quote(s, safe = '/'): |
| 307 | """quote('abc def') -> 'abc%20def' |
| 308 | |
| 309 | Each part of a URL, e.g. the path info, the query, etc., has a |
| 310 | different set of reserved characters that must be quoted. |
| 311 | |
| 312 | RFC 2396 Uniform Resource Identifiers (URI): Generic Syntax lists |
| 313 | the following reserved characters. |
| 314 | |
| 315 | reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" | |
| 316 | "$" | "," |
| 317 | |
| 318 | Each of these characters is reserved in some component of a URL, |
| 319 | but not necessarily in all of them. |
| 320 | |
| 321 | By default, the quote function is intended for quoting the path |
| 322 | section of a URL. Thus, it will not encode '/'. This character |
| 323 | is reserved, but in typical usage the quote function is being |
| 324 | called on a path where the existing slash characters are used as |
| 325 | reserved characters. |
| 326 | """ |
| 327 | cachekey = (safe, always_safe) |
| 328 | try: |
| 329 | quoter = _safe_quoters[cachekey] |
| 330 | except KeyError: |
| 331 | quoter = Quoter(safe) |
| 332 | _safe_quoters[cachekey] = quoter |
| 333 | res = map(quoter, s) |
| 334 | return ''.join(res) |
| 335 | |
| 336 | def quote_plus(s, safe = ''): |
| 337 | """Quote the query fragment of a URL; replacing ' ' with '+'""" |
| 338 | if ' ' in s: |
| 339 | s = quote(s, safe + ' ') |
| 340 | return s.replace(' ', '+') |
| 341 | return quote(s, safe) |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 342 | |
| 343 | def urlencode(query,doseq=0): |
| 344 | """Encode a sequence of two-element tuples or dictionary into a URL query string. |
| 345 | |
| 346 | If any values in the query arg are sequences and doseq is true, each |
| 347 | sequence element is converted to a separate parameter. |
| 348 | |
| 349 | If the query arg is a sequence of two-element tuples, the order of the |
| 350 | parameters in the output will match the order of parameters in the |
| 351 | input. |
| 352 | """ |
| 353 | |
| 354 | if hasattr(query,"items"): |
| 355 | # mapping objects |
| 356 | query = query.items() |
| 357 | else: |
| 358 | # it's a bother at times that strings and string-like objects are |
| 359 | # sequences... |
| 360 | try: |
| 361 | # non-sequence items should not work with len() |
| 362 | # non-empty strings will fail this |
| 363 | if len(query) and not isinstance(query[0], tuple): |
| 364 | raise TypeError |
| 365 | # zero-length sequences of all types will get here and succeed, |
| 366 | # but that's a minor nit - since the original implementation |
| 367 | # allowed empty dicts that type of behavior probably should be |
| 368 | # preserved for consistency |
| 369 | except TypeError: |
| 370 | ty,va,tb = sys.exc_info() |
| 371 | raise TypeError("not a valid non-string sequence or mapping object").with_traceback(tb) |
| 372 | |
| 373 | l = [] |
| 374 | if not doseq: |
| 375 | # preserve old behavior |
| 376 | for k, v in query: |
| 377 | k = quote_plus(str(k)) |
| 378 | v = quote_plus(str(v)) |
| 379 | l.append(k + '=' + v) |
| 380 | else: |
| 381 | for k, v in query: |
| 382 | k = quote_plus(str(k)) |
| 383 | if isinstance(v, str): |
| 384 | v = quote_plus(v) |
| 385 | l.append(k + '=' + v) |
| 386 | elif isinstance(v, str): |
| 387 | # is there a reasonable way to convert to ASCII? |
| 388 | # encode generates a string, but "replace" or "ignore" |
| 389 | # lose information and "strict" can raise UnicodeError |
Guido van Rossum | df9f1ec | 2008-08-06 19:31:34 +0000 | [diff] [blame^] | 390 | v = quote_plus(v.encode("ASCII","replace")) |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 391 | l.append(k + '=' + v) |
| 392 | else: |
| 393 | try: |
| 394 | # is this a sufficient test for sequence-ness? |
| 395 | x = len(v) |
| 396 | except TypeError: |
| 397 | # not a sequence |
| 398 | v = quote_plus(str(v)) |
| 399 | l.append(k + '=' + v) |
| 400 | else: |
| 401 | # loop over the sequence |
| 402 | for elt in v: |
| 403 | l.append(k + '=' + quote_plus(str(elt))) |
| 404 | return '&'.join(l) |
| 405 | |
| 406 | # Utilities to parse URLs (most of these return None for missing parts): |
| 407 | # unwrap('<URL:type://host/path>') --> 'type://host/path' |
| 408 | # splittype('type:opaquestring') --> 'type', 'opaquestring' |
| 409 | # splithost('//host[:port]/path') --> 'host[:port]', '/path' |
| 410 | # splituser('user[:passwd]@host[:port]') --> 'user[:passwd]', 'host[:port]' |
| 411 | # splitpasswd('user:passwd') -> 'user', 'passwd' |
| 412 | # splitport('host:port') --> 'host', 'port' |
| 413 | # splitquery('/path?query') --> '/path', 'query' |
| 414 | # splittag('/path#tag') --> '/path', 'tag' |
| 415 | # splitattr('/path;attr1=value1;attr2=value2;...') -> |
| 416 | # '/path', ['attr1=value1', 'attr2=value2', ...] |
| 417 | # splitvalue('attr=value') --> 'attr', 'value' |
| 418 | # urllib.parse.unquote('abc%20def') -> 'abc def' |
| 419 | # quote('abc def') -> 'abc%20def') |
| 420 | |
Georg Brandl | 13e8946 | 2008-07-01 19:56:00 +0000 | [diff] [blame] | 421 | def to_bytes(url): |
| 422 | """to_bytes(u"URL") --> 'URL'.""" |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 423 | # Most URL schemes require ASCII. If that changes, the conversion |
| 424 | # can be relaxed. |
Georg Brandl | 13e8946 | 2008-07-01 19:56:00 +0000 | [diff] [blame] | 425 | # XXX get rid of to_bytes() |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 426 | if isinstance(url, str): |
| 427 | try: |
| 428 | url = url.encode("ASCII").decode() |
| 429 | except UnicodeError: |
| 430 | raise UnicodeError("URL " + repr(url) + |
| 431 | " contains non-ASCII characters") |
| 432 | return url |
| 433 | |
| 434 | def unwrap(url): |
| 435 | """unwrap('<URL:type://host/path>') --> 'type://host/path'.""" |
| 436 | url = str(url).strip() |
| 437 | if url[:1] == '<' and url[-1:] == '>': |
| 438 | url = url[1:-1].strip() |
| 439 | if url[:4] == 'URL:': url = url[4:].strip() |
| 440 | return url |
| 441 | |
| 442 | _typeprog = None |
| 443 | def splittype(url): |
| 444 | """splittype('type:opaquestring') --> 'type', 'opaquestring'.""" |
| 445 | global _typeprog |
| 446 | if _typeprog is None: |
| 447 | import re |
| 448 | _typeprog = re.compile('^([^/:]+):') |
| 449 | |
| 450 | match = _typeprog.match(url) |
| 451 | if match: |
| 452 | scheme = match.group(1) |
| 453 | return scheme.lower(), url[len(scheme) + 1:] |
| 454 | return None, url |
| 455 | |
| 456 | _hostprog = None |
| 457 | def splithost(url): |
| 458 | """splithost('//host[:port]/path') --> 'host[:port]', '/path'.""" |
| 459 | global _hostprog |
| 460 | if _hostprog is None: |
| 461 | import re |
| 462 | _hostprog = re.compile('^//([^/?]*)(.*)$') |
| 463 | |
| 464 | match = _hostprog.match(url) |
| 465 | if match: return match.group(1, 2) |
| 466 | return None, url |
| 467 | |
| 468 | _userprog = None |
| 469 | def splituser(host): |
| 470 | """splituser('user[:passwd]@host[:port]') --> 'user[:passwd]', 'host[:port]'.""" |
| 471 | global _userprog |
| 472 | if _userprog is None: |
| 473 | import re |
| 474 | _userprog = re.compile('^(.*)@(.*)$') |
| 475 | |
| 476 | match = _userprog.match(host) |
Guido van Rossum | df9f1ec | 2008-08-06 19:31:34 +0000 | [diff] [blame^] | 477 | if match: return map(unquote, match.group(1, 2)) |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 478 | return None, host |
| 479 | |
| 480 | _passwdprog = None |
| 481 | def splitpasswd(user): |
| 482 | """splitpasswd('user:passwd') -> 'user', 'passwd'.""" |
| 483 | global _passwdprog |
| 484 | if _passwdprog is None: |
| 485 | import re |
| 486 | _passwdprog = re.compile('^([^:]*):(.*)$') |
| 487 | |
| 488 | match = _passwdprog.match(user) |
| 489 | if match: return match.group(1, 2) |
| 490 | return user, None |
| 491 | |
| 492 | # splittag('/path#tag') --> '/path', 'tag' |
| 493 | _portprog = None |
| 494 | def splitport(host): |
| 495 | """splitport('host:port') --> 'host', 'port'.""" |
| 496 | global _portprog |
| 497 | if _portprog is None: |
| 498 | import re |
| 499 | _portprog = re.compile('^(.*):([0-9]+)$') |
| 500 | |
| 501 | match = _portprog.match(host) |
| 502 | if match: return match.group(1, 2) |
| 503 | return host, None |
| 504 | |
| 505 | _nportprog = None |
| 506 | def splitnport(host, defport=-1): |
| 507 | """Split host and port, returning numeric port. |
| 508 | Return given default port if no ':' found; defaults to -1. |
| 509 | Return numerical port if a valid number are found after ':'. |
| 510 | Return None if ':' but not a valid number.""" |
| 511 | global _nportprog |
| 512 | if _nportprog is None: |
| 513 | import re |
| 514 | _nportprog = re.compile('^(.*):(.*)$') |
| 515 | |
| 516 | match = _nportprog.match(host) |
| 517 | if match: |
| 518 | host, port = match.group(1, 2) |
| 519 | try: |
| 520 | if not port: raise ValueError("no digits") |
| 521 | nport = int(port) |
| 522 | except ValueError: |
| 523 | nport = None |
| 524 | return host, nport |
| 525 | return host, defport |
| 526 | |
| 527 | _queryprog = None |
| 528 | def splitquery(url): |
| 529 | """splitquery('/path?query') --> '/path', 'query'.""" |
| 530 | global _queryprog |
| 531 | if _queryprog is None: |
| 532 | import re |
| 533 | _queryprog = re.compile('^(.*)\?([^?]*)$') |
| 534 | |
| 535 | match = _queryprog.match(url) |
| 536 | if match: return match.group(1, 2) |
| 537 | return url, None |
| 538 | |
| 539 | _tagprog = None |
| 540 | def splittag(url): |
| 541 | """splittag('/path#tag') --> '/path', 'tag'.""" |
| 542 | global _tagprog |
| 543 | if _tagprog is None: |
| 544 | import re |
| 545 | _tagprog = re.compile('^(.*)#([^#]*)$') |
| 546 | |
| 547 | match = _tagprog.match(url) |
| 548 | if match: return match.group(1, 2) |
| 549 | return url, None |
| 550 | |
| 551 | def splitattr(url): |
| 552 | """splitattr('/path;attr1=value1;attr2=value2;...') -> |
| 553 | '/path', ['attr1=value1', 'attr2=value2', ...].""" |
| 554 | words = url.split(';') |
| 555 | return words[0], words[1:] |
| 556 | |
| 557 | _valueprog = None |
| 558 | def splitvalue(attr): |
| 559 | """splitvalue('attr=value') --> 'attr', 'value'.""" |
| 560 | global _valueprog |
| 561 | if _valueprog is None: |
| 562 | import re |
| 563 | _valueprog = re.compile('^([^=]*)=(.*)$') |
| 564 | |
| 565 | match = _valueprog.match(attr) |
| 566 | if match: return match.group(1, 2) |
| 567 | return attr, None |
| 568 | |
| 569 | test_input = """ |
| 570 | http://a/b/c/d |
| 571 | |
| 572 | g:h = <URL:g:h> |
| 573 | http:g = <URL:http://a/b/c/g> |
| 574 | http: = <URL:http://a/b/c/d> |
| 575 | g = <URL:http://a/b/c/g> |
| 576 | ./g = <URL:http://a/b/c/g> |
| 577 | g/ = <URL:http://a/b/c/g/> |
| 578 | /g = <URL:http://a/g> |
| 579 | //g = <URL:http://g> |
| 580 | ?y = <URL:http://a/b/c/d?y> |
| 581 | g?y = <URL:http://a/b/c/g?y> |
| 582 | g?y/./x = <URL:http://a/b/c/g?y/./x> |
| 583 | . = <URL:http://a/b/c/> |
| 584 | ./ = <URL:http://a/b/c/> |
| 585 | .. = <URL:http://a/b/> |
| 586 | ../ = <URL:http://a/b/> |
| 587 | ../g = <URL:http://a/b/g> |
| 588 | ../.. = <URL:http://a/> |
| 589 | ../../g = <URL:http://a/g> |
| 590 | ../../../g = <URL:http://a/../g> |
| 591 | ./../g = <URL:http://a/b/g> |
| 592 | ./g/. = <URL:http://a/b/c/g/> |
| 593 | /./g = <URL:http://a/./g> |
| 594 | g/./h = <URL:http://a/b/c/g/h> |
| 595 | g/../h = <URL:http://a/b/c/h> |
| 596 | http:g = <URL:http://a/b/c/g> |
| 597 | http: = <URL:http://a/b/c/d> |
| 598 | http:?y = <URL:http://a/b/c/d?y> |
| 599 | http:g?y = <URL:http://a/b/c/g?y> |
| 600 | http:g?y/./x = <URL:http://a/b/c/g?y/./x> |
| 601 | """ |
| 602 | |
| 603 | def test(): |
Jeremy Hylton | 1afc169 | 2008-06-18 20:49:58 +0000 | [diff] [blame] | 604 | base = '' |
| 605 | if sys.argv[1:]: |
| 606 | fn = sys.argv[1] |
| 607 | if fn == '-': |
| 608 | fp = sys.stdin |
| 609 | else: |
| 610 | fp = open(fn) |
| 611 | else: |
| 612 | from io import StringIO |
| 613 | fp = StringIO(test_input) |
| 614 | for line in fp: |
| 615 | words = line.split() |
| 616 | if not words: |
| 617 | continue |
| 618 | url = words[0] |
| 619 | parts = urlparse(url) |
| 620 | print('%-10s : %s' % (url, parts)) |
| 621 | abs = urljoin(base, url) |
| 622 | if not base: |
| 623 | base = abs |
| 624 | wrapped = '<URL:%s>' % abs |
| 625 | print('%-10s = %s' % (url, wrapped)) |
| 626 | if len(words) == 3 and words[1] == '=': |
| 627 | if wrapped != words[2]: |
| 628 | print('EXPECTED', words[2], '!!!!!!!!!!') |
| 629 | |
| 630 | if __name__ == '__main__': |
| 631 | test() |