| 1 | """Open an arbitrary URL.
|
|---|
| 2 |
|
|---|
| 3 | See the following document for more info on URLs:
|
|---|
| 4 | "Names and Addresses, URIs, URLs, URNs, URCs", at
|
|---|
| 5 | http://www.w3.org/pub/WWW/Addressing/Overview.html
|
|---|
| 6 |
|
|---|
| 7 | See also the HTTP spec (from which the error codes are derived):
|
|---|
| 8 | "HTTP - Hypertext Transfer Protocol", at
|
|---|
| 9 | http://www.w3.org/pub/WWW/Protocols/
|
|---|
| 10 |
|
|---|
| 11 | Related standards and specs:
|
|---|
| 12 | - RFC1808: the "relative URL" spec. (authoritative status)
|
|---|
| 13 | - RFC1738 - the "URL standard". (authoritative status)
|
|---|
| 14 | - RFC1630 - the "URI spec". (informational status)
|
|---|
| 15 |
|
|---|
| 16 | The object returned by URLopener().open(file) will differ per
|
|---|
| 17 | protocol. All you know is that is has methods read(), readline(),
|
|---|
| 18 | readlines(), fileno(), close() and info(). The read*(), fileno()
|
|---|
| 19 | and close() methods work like those of open files.
|
|---|
| 20 | The info() method returns a mimetools.Message object which can be
|
|---|
| 21 | used to query various info about the object, if available.
|
|---|
| 22 | (mimetools.Message objects are queried with the getheader() method.)
|
|---|
| 23 | """
|
|---|
| 24 |
|
|---|
| 25 | import string
|
|---|
| 26 | import socket
|
|---|
| 27 | import os
|
|---|
| 28 | import time
|
|---|
| 29 | import sys
|
|---|
| 30 | from urlparse import urljoin as basejoin
|
|---|
| 31 |
|
|---|
| 32 | __all__ = ["urlopen", "URLopener", "FancyURLopener", "urlretrieve",
|
|---|
| 33 | "urlcleanup", "quote", "quote_plus", "unquote", "unquote_plus",
|
|---|
| 34 | "urlencode", "url2pathname", "pathname2url", "splittag",
|
|---|
| 35 | "localhost", "thishost", "ftperrors", "basejoin", "unwrap",
|
|---|
| 36 | "splittype", "splithost", "splituser", "splitpasswd", "splitport",
|
|---|
| 37 | "splitnport", "splitquery", "splitattr", "splitvalue",
|
|---|
| 38 | "splitgophertype", "getproxies"]
|
|---|
| 39 |
|
|---|
| 40 | __version__ = '1.17' # XXX This version is not always updated :-(
|
|---|
| 41 |
|
|---|
| 42 | MAXFTPCACHE = 10 # Trim the ftp cache beyond this size
|
|---|
| 43 |
|
|---|
| 44 | # Helper for non-unix systems
|
|---|
| 45 | if os.name == 'mac':
|
|---|
| 46 | from macurl2path import url2pathname, pathname2url
|
|---|
| 47 | elif os.name == 'nt':
|
|---|
| 48 | from nturl2path import url2pathname, pathname2url
|
|---|
| 49 | elif os.name == 'riscos':
|
|---|
| 50 | from rourl2path import url2pathname, pathname2url
|
|---|
| 51 | else:
|
|---|
| 52 | def url2pathname(pathname):
|
|---|
| 53 | """OS-specific conversion from a relative URL of the 'file' scheme
|
|---|
| 54 | to a file system path; not recommended for general use."""
|
|---|
| 55 | return unquote(pathname)
|
|---|
| 56 |
|
|---|
| 57 | def pathname2url(pathname):
|
|---|
| 58 | """OS-specific conversion from a file system path to a relative URL
|
|---|
| 59 | of the 'file' scheme; not recommended for general use."""
|
|---|
| 60 | return quote(pathname)
|
|---|
| 61 |
|
|---|
| 62 | # This really consists of two pieces:
|
|---|
| 63 | # (1) a class which handles opening of all sorts of URLs
|
|---|
| 64 | # (plus assorted utilities etc.)
|
|---|
| 65 | # (2) a set of functions for parsing URLs
|
|---|
| 66 | # XXX Should these be separated out into different modules?
|
|---|
| 67 |
|
|---|
| 68 |
|
|---|
| 69 | # Shortcut for basic usage
|
|---|
| 70 | _urlopener = None
|
|---|
| 71 | def urlopen(url, data=None, proxies=None):
|
|---|
| 72 | """urlopen(url [, data]) -> open file-like object"""
|
|---|
| 73 | global _urlopener
|
|---|
| 74 | if proxies is not None:
|
|---|
| 75 | opener = FancyURLopener(proxies=proxies)
|
|---|
| 76 | elif not _urlopener:
|
|---|
| 77 | opener = FancyURLopener()
|
|---|
| 78 | _urlopener = opener
|
|---|
| 79 | else:
|
|---|
| 80 | opener = _urlopener
|
|---|
| 81 | if data is None:
|
|---|
| 82 | return opener.open(url)
|
|---|
| 83 | else:
|
|---|
| 84 | return opener.open(url, data)
|
|---|
| 85 | def urlretrieve(url, filename=None, reporthook=None, data=None):
|
|---|
| 86 | global _urlopener
|
|---|
| 87 | if not _urlopener:
|
|---|
| 88 | _urlopener = FancyURLopener()
|
|---|
| 89 | return _urlopener.retrieve(url, filename, reporthook, data)
|
|---|
| 90 | def urlcleanup():
|
|---|
| 91 | if _urlopener:
|
|---|
| 92 | _urlopener.cleanup()
|
|---|
| 93 |
|
|---|
| 94 | # exception raised when downloaded size does not match content-length
|
|---|
| 95 | class ContentTooShortError(IOError):
|
|---|
| 96 | def __init__(self, message, content):
|
|---|
| 97 | IOError.__init__(self, message)
|
|---|
| 98 | self.content = content
|
|---|
| 99 |
|
|---|
| 100 | ftpcache = {}
|
|---|
| 101 | class URLopener:
|
|---|
| 102 | """Class to open URLs.
|
|---|
| 103 | This is a class rather than just a subroutine because we may need
|
|---|
| 104 | more than one set of global protocol-specific options.
|
|---|
| 105 | Note -- this is a base class for those who don't want the
|
|---|
| 106 | automatic handling of errors type 302 (relocated) and 401
|
|---|
| 107 | (authorization needed)."""
|
|---|
| 108 |
|
|---|
| 109 | __tempfiles = None
|
|---|
| 110 |
|
|---|
| 111 | version = "Python-urllib/%s" % __version__
|
|---|
| 112 |
|
|---|
| 113 | # Constructor
|
|---|
| 114 | def __init__(self, proxies=None, **x509):
|
|---|
| 115 | if proxies is None:
|
|---|
| 116 | proxies = getproxies()
|
|---|
| 117 | assert hasattr(proxies, 'has_key'), "proxies must be a mapping"
|
|---|
| 118 | self.proxies = proxies
|
|---|
| 119 | self.key_file = x509.get('key_file')
|
|---|
| 120 | self.cert_file = x509.get('cert_file')
|
|---|
| 121 | self.addheaders = [('User-Agent', self.version)]
|
|---|
| 122 | self.__tempfiles = []
|
|---|
| 123 | self.__unlink = os.unlink # See cleanup()
|
|---|
| 124 | self.tempcache = None
|
|---|
| 125 | # Undocumented feature: if you assign {} to tempcache,
|
|---|
| 126 | # it is used to cache files retrieved with
|
|---|
| 127 | # self.retrieve(). This is not enabled by default
|
|---|
| 128 | # since it does not work for changing documents (and I
|
|---|
| 129 | # haven't got the logic to check expiration headers
|
|---|
| 130 | # yet).
|
|---|
| 131 | self.ftpcache = ftpcache
|
|---|
| 132 | # Undocumented feature: you can use a different
|
|---|
| 133 | # ftp cache by assigning to the .ftpcache member;
|
|---|
| 134 | # in case you want logically independent URL openers
|
|---|
| 135 | # XXX This is not threadsafe. Bah.
|
|---|
| 136 |
|
|---|
| 137 | def __del__(self):
|
|---|
| 138 | self.close()
|
|---|
| 139 |
|
|---|
| 140 | def close(self):
|
|---|
| 141 | self.cleanup()
|
|---|
| 142 |
|
|---|
| 143 | def cleanup(self):
|
|---|
| 144 | # This code sometimes runs when the rest of this module
|
|---|
| 145 | # has already been deleted, so it can't use any globals
|
|---|
| 146 | # or import anything.
|
|---|
| 147 | if self.__tempfiles:
|
|---|
| 148 | for file in self.__tempfiles:
|
|---|
| 149 | try:
|
|---|
| 150 | self.__unlink(file)
|
|---|
| 151 | except OSError:
|
|---|
| 152 | pass
|
|---|
| 153 | del self.__tempfiles[:]
|
|---|
| 154 | if self.tempcache:
|
|---|
| 155 | self.tempcache.clear()
|
|---|
| 156 |
|
|---|
| 157 | def addheader(self, *args):
|
|---|
| 158 | """Add a header to be used by the HTTP interface only
|
|---|
| 159 | e.g. u.addheader('Accept', 'sound/basic')"""
|
|---|
| 160 | self.addheaders.append(args)
|
|---|
| 161 |
|
|---|
| 162 | # External interface
|
|---|
| 163 | def open(self, fullurl, data=None):
|
|---|
| 164 | """Use URLopener().open(file) instead of open(file, 'r')."""
|
|---|
| 165 | fullurl = unwrap(toBytes(fullurl))
|
|---|
| 166 | if self.tempcache and fullurl in self.tempcache:
|
|---|
| 167 | filename, headers = self.tempcache[fullurl]
|
|---|
| 168 | fp = open(filename, 'rb')
|
|---|
| 169 | return addinfourl(fp, headers, fullurl)
|
|---|
| 170 | urltype, url = splittype(fullurl)
|
|---|
| 171 | if not urltype:
|
|---|
| 172 | urltype = 'file'
|
|---|
| 173 | if urltype in self.proxies:
|
|---|
| 174 | proxy = self.proxies[urltype]
|
|---|
| 175 | urltype, proxyhost = splittype(proxy)
|
|---|
| 176 | host, selector = splithost(proxyhost)
|
|---|
| 177 | url = (host, fullurl) # Signal special case to open_*()
|
|---|
| 178 | else:
|
|---|
| 179 | proxy = None
|
|---|
| 180 | name = 'open_' + urltype
|
|---|
| 181 | self.type = urltype
|
|---|
| 182 | name = name.replace('-', '_')
|
|---|
| 183 | if not hasattr(self, name):
|
|---|
| 184 | if proxy:
|
|---|
| 185 | return self.open_unknown_proxy(proxy, fullurl, data)
|
|---|
| 186 | else:
|
|---|
| 187 | return self.open_unknown(fullurl, data)
|
|---|
| 188 | try:
|
|---|
| 189 | if data is None:
|
|---|
| 190 | return getattr(self, name)(url)
|
|---|
| 191 | else:
|
|---|
| 192 | return getattr(self, name)(url, data)
|
|---|
| 193 | except socket.error, msg:
|
|---|
| 194 | raise IOError, ('socket error', msg), sys.exc_info()[2]
|
|---|
| 195 |
|
|---|
| 196 | def open_unknown(self, fullurl, data=None):
|
|---|
| 197 | """Overridable interface to open unknown URL type."""
|
|---|
| 198 | type, url = splittype(fullurl)
|
|---|
| 199 | raise IOError, ('url error', 'unknown url type', type)
|
|---|
| 200 |
|
|---|
| 201 | def open_unknown_proxy(self, proxy, fullurl, data=None):
|
|---|
| 202 | """Overridable interface to open unknown URL type."""
|
|---|
| 203 | type, url = splittype(fullurl)
|
|---|
| 204 | raise IOError, ('url error', 'invalid proxy for %s' % type, proxy)
|
|---|
| 205 |
|
|---|
| 206 | # External interface
|
|---|
| 207 | def retrieve(self, url, filename=None, reporthook=None, data=None):
|
|---|
| 208 | """retrieve(url) returns (filename, headers) for a local object
|
|---|
| 209 | or (tempfilename, headers) for a remote object."""
|
|---|
| 210 | url = unwrap(toBytes(url))
|
|---|
| 211 | if self.tempcache and url in self.tempcache:
|
|---|
| 212 | return self.tempcache[url]
|
|---|
| 213 | type, url1 = splittype(url)
|
|---|
| 214 | if filename is None and (not type or type == 'file'):
|
|---|
| 215 | try:
|
|---|
| 216 | fp = self.open_local_file(url1)
|
|---|
| 217 | hdrs = fp.info()
|
|---|
| 218 | del fp
|
|---|
| 219 | return url2pathname(splithost(url1)[1]), hdrs
|
|---|
| 220 | except IOError, msg:
|
|---|
| 221 | pass
|
|---|
| 222 | fp = self.open(url, data)
|
|---|
| 223 | headers = fp.info()
|
|---|
| 224 | if filename:
|
|---|
| 225 | tfp = open(filename, 'wb')
|
|---|
| 226 | else:
|
|---|
| 227 | import tempfile
|
|---|
| 228 | garbage, path = splittype(url)
|
|---|
| 229 | garbage, path = splithost(path or "")
|
|---|
| 230 | path, garbage = splitquery(path or "")
|
|---|
| 231 | path, garbage = splitattr(path or "")
|
|---|
| 232 | suffix = os.path.splitext(path)[1]
|
|---|
| 233 | (fd, filename) = tempfile.mkstemp(suffix)
|
|---|
| 234 | self.__tempfiles.append(filename)
|
|---|
| 235 | tfp = os.fdopen(fd, 'wb')
|
|---|
| 236 | result = filename, headers
|
|---|
| 237 | if self.tempcache is not None:
|
|---|
| 238 | self.tempcache[url] = result
|
|---|
| 239 | bs = 1024*8
|
|---|
| 240 | size = -1
|
|---|
| 241 | read = 0
|
|---|
| 242 | blocknum = 0
|
|---|
| 243 | if reporthook:
|
|---|
| 244 | if "content-length" in headers:
|
|---|
| 245 | size = int(headers["Content-Length"])
|
|---|
| 246 | reporthook(blocknum, bs, size)
|
|---|
| 247 | while 1:
|
|---|
| 248 | block = fp.read(bs)
|
|---|
| 249 | if block == "":
|
|---|
| 250 | break
|
|---|
| 251 | read += len(block)
|
|---|
| 252 | tfp.write(block)
|
|---|
| 253 | blocknum += 1
|
|---|
| 254 | if reporthook:
|
|---|
| 255 | reporthook(blocknum, bs, size)
|
|---|
| 256 | fp.close()
|
|---|
| 257 | tfp.close()
|
|---|
| 258 | del fp
|
|---|
| 259 | del tfp
|
|---|
| 260 |
|
|---|
| 261 | # raise exception if actual size does not match content-length header
|
|---|
| 262 | if size >= 0 and read < size:
|
|---|
| 263 | raise ContentTooShortError("retrieval incomplete: got only %i out "
|
|---|
| 264 | "of %i bytes" % (read, size), result)
|
|---|
| 265 |
|
|---|
| 266 | return result
|
|---|
| 267 |
|
|---|
| 268 | # Each method named open_<type> knows how to open that type of URL
|
|---|
| 269 |
|
|---|
| 270 | def open_http(self, url, data=None):
|
|---|
| 271 | """Use HTTP protocol."""
|
|---|
| 272 | import httplib
|
|---|
| 273 | user_passwd = None
|
|---|
| 274 | proxy_passwd= None
|
|---|
| 275 | if isinstance(url, str):
|
|---|
| 276 | host, selector = splithost(url)
|
|---|
| 277 | if host:
|
|---|
| 278 | user_passwd, host = splituser(host)
|
|---|
| 279 | host = unquote(host)
|
|---|
| 280 | realhost = host
|
|---|
| 281 | else:
|
|---|
| 282 | host, selector = url
|
|---|
| 283 | # check whether the proxy contains authorization information
|
|---|
| 284 | proxy_passwd, host = splituser(host)
|
|---|
| 285 | # now we proceed with the url we want to obtain
|
|---|
| 286 | urltype, rest = splittype(selector)
|
|---|
| 287 | url = rest
|
|---|
| 288 | user_passwd = None
|
|---|
| 289 | if urltype.lower() != 'http':
|
|---|
| 290 | realhost = None
|
|---|
| 291 | else:
|
|---|
| 292 | realhost, rest = splithost(rest)
|
|---|
| 293 | if realhost:
|
|---|
| 294 | user_passwd, realhost = splituser(realhost)
|
|---|
| 295 | if user_passwd:
|
|---|
| 296 | selector = "%s://%s%s" % (urltype, realhost, rest)
|
|---|
| 297 | if proxy_bypass(realhost):
|
|---|
| 298 | host = realhost
|
|---|
| 299 |
|
|---|
| 300 | #print "proxy via http:", host, selector
|
|---|
| 301 | if not host: raise IOError, ('http error', 'no host given')
|
|---|
| 302 |
|
|---|
| 303 | if proxy_passwd:
|
|---|
| 304 | import base64
|
|---|
| 305 | proxy_auth = base64.encodestring(proxy_passwd).strip()
|
|---|
| 306 | else:
|
|---|
| 307 | proxy_auth = None
|
|---|
| 308 |
|
|---|
| 309 | if user_passwd:
|
|---|
| 310 | import base64
|
|---|
| 311 | auth = base64.encodestring(user_passwd).strip()
|
|---|
| 312 | else:
|
|---|
| 313 | auth = None
|
|---|
| 314 | h = httplib.HTTP(host)
|
|---|
| 315 | if data is not None:
|
|---|
| 316 | h.putrequest('POST', selector)
|
|---|
| 317 | h.putheader('Content-Type', 'application/x-www-form-urlencoded')
|
|---|
| 318 | h.putheader('Content-Length', '%d' % len(data))
|
|---|
| 319 | else:
|
|---|
| 320 | h.putrequest('GET', selector)
|
|---|
| 321 | if proxy_auth: h.putheader('Proxy-Authorization', 'Basic %s' % proxy_auth)
|
|---|
| 322 | if auth: h.putheader('Authorization', 'Basic %s' % auth)
|
|---|
| 323 | if realhost: h.putheader('Host', realhost)
|
|---|
| 324 | for args in self.addheaders: h.putheader(*args)
|
|---|
| 325 | h.endheaders()
|
|---|
| 326 | if data is not None:
|
|---|
| 327 | h.send(data)
|
|---|
| 328 | errcode, errmsg, headers = h.getreply()
|
|---|
| 329 | fp = h.getfile()
|
|---|
| 330 | if errcode == 200:
|
|---|
| 331 | return addinfourl(fp, headers, "http:" + url)
|
|---|
| 332 | else:
|
|---|
| 333 | if data is None:
|
|---|
| 334 | return self.http_error(url, fp, errcode, errmsg, headers)
|
|---|
| 335 | else:
|
|---|
| 336 | return self.http_error(url, fp, errcode, errmsg, headers, data)
|
|---|
| 337 |
|
|---|
| 338 | def http_error(self, url, fp, errcode, errmsg, headers, data=None):
|
|---|
| 339 | """Handle http errors.
|
|---|
| 340 | Derived class can override this, or provide specific handlers
|
|---|
| 341 | named http_error_DDD where DDD is the 3-digit error code."""
|
|---|
| 342 | # First check if there's a specific handler for this error
|
|---|
| 343 | name = 'http_error_%d' % errcode
|
|---|
| 344 | if hasattr(self, name):
|
|---|
| 345 | method = getattr(self, name)
|
|---|
| 346 | if data is None:
|
|---|
| 347 | result = method(url, fp, errcode, errmsg, headers)
|
|---|
| 348 | else:
|
|---|
| 349 | result = method(url, fp, errcode, errmsg, headers, data)
|
|---|
| 350 | if result: return result
|
|---|
| 351 | return self.http_error_default(url, fp, errcode, errmsg, headers)
|
|---|
| 352 |
|
|---|
| 353 | def http_error_default(self, url, fp, errcode, errmsg, headers):
|
|---|
| 354 | """Default error handler: close the connection and raise IOError."""
|
|---|
| 355 | void = fp.read()
|
|---|
| 356 | fp.close()
|
|---|
| 357 | raise IOError, ('http error', errcode, errmsg, headers)
|
|---|
| 358 |
|
|---|
| 359 | if hasattr(socket, "ssl"):
|
|---|
| 360 | def open_https(self, url, data=None):
|
|---|
| 361 | """Use HTTPS protocol."""
|
|---|
| 362 | import httplib
|
|---|
| 363 | user_passwd = None
|
|---|
| 364 | proxy_passwd = None
|
|---|
| 365 | if isinstance(url, str):
|
|---|
| 366 | host, selector = splithost(url)
|
|---|
| 367 | if host:
|
|---|
| 368 | user_passwd, host = splituser(host)
|
|---|
| 369 | host = unquote(host)
|
|---|
| 370 | realhost = host
|
|---|
| 371 | else:
|
|---|
| 372 | host, selector = url
|
|---|
| 373 | # here, we determine, whether the proxy contains authorization information
|
|---|
| 374 | proxy_passwd, host = splituser(host)
|
|---|
| 375 | urltype, rest = splittype(selector)
|
|---|
| 376 | url = rest
|
|---|
| 377 | user_passwd = None
|
|---|
| 378 | if urltype.lower() != 'https':
|
|---|
| 379 | realhost = None
|
|---|
| 380 | else:
|
|---|
| 381 | realhost, rest = splithost(rest)
|
|---|
| 382 | if realhost:
|
|---|
| 383 | user_passwd, realhost = splituser(realhost)
|
|---|
| 384 | if user_passwd:
|
|---|
| 385 | selector = "%s://%s%s" % (urltype, realhost, rest)
|
|---|
| 386 | #print "proxy via https:", host, selector
|
|---|
| 387 | if not host: raise IOError, ('https error', 'no host given')
|
|---|
| 388 | if proxy_passwd:
|
|---|
| 389 | import base64
|
|---|
| 390 | proxy_auth = base64.encodestring(proxy_passwd).strip()
|
|---|
| 391 | else:
|
|---|
| 392 | proxy_auth = None
|
|---|
| 393 | if user_passwd:
|
|---|
| 394 | import base64
|
|---|
| 395 | auth = base64.encodestring(user_passwd).strip()
|
|---|
| 396 | else:
|
|---|
| 397 | auth = None
|
|---|
| 398 | h = httplib.HTTPS(host, 0,
|
|---|
| 399 | key_file=self.key_file,
|
|---|
| 400 | cert_file=self.cert_file)
|
|---|
| 401 | if data is not None:
|
|---|
| 402 | h.putrequest('POST', selector)
|
|---|
| 403 | h.putheader('Content-Type',
|
|---|
| 404 | 'application/x-www-form-urlencoded')
|
|---|
| 405 | h.putheader('Content-Length', '%d' % len(data))
|
|---|
| 406 | else:
|
|---|
| 407 | h.putrequest('GET', selector)
|
|---|
| 408 | if proxy_auth: h.putheader('Proxy-Authorization: Basic %s' % proxy_auth)
|
|---|
| 409 | if auth: h.putheader('Authorization: Basic %s' % auth)
|
|---|
| 410 | if realhost: h.putheader('Host', realhost)
|
|---|
| 411 | for args in self.addheaders: h.putheader(*args)
|
|---|
| 412 | h.endheaders()
|
|---|
| 413 | if data is not None:
|
|---|
| 414 | h.send(data)
|
|---|
| 415 | errcode, errmsg, headers = h.getreply()
|
|---|
| 416 | fp = h.getfile()
|
|---|
| 417 | if errcode == 200:
|
|---|
| 418 | return addinfourl(fp, headers, "https:" + url)
|
|---|
| 419 | else:
|
|---|
| 420 | if data is None:
|
|---|
| 421 | return self.http_error(url, fp, errcode, errmsg, headers)
|
|---|
| 422 | else:
|
|---|
| 423 | return self.http_error(url, fp, errcode, errmsg, headers,
|
|---|
| 424 | data)
|
|---|
| 425 |
|
|---|
| 426 | def open_gopher(self, url):
|
|---|
| 427 | """Use Gopher protocol."""
|
|---|
| 428 | if not isinstance(url, str):
|
|---|
| 429 | raise IOError, ('gopher error', 'proxy support for gopher protocol currently not implemented')
|
|---|
| 430 | import gopherlib
|
|---|
| 431 | host, selector = splithost(url)
|
|---|
| 432 | if not host: raise IOError, ('gopher error', 'no host given')
|
|---|
| 433 | host = unquote(host)
|
|---|
| 434 | type, selector = splitgophertype(selector)
|
|---|
| 435 | selector, query = splitquery(selector)
|
|---|
| 436 | selector = unquote(selector)
|
|---|
| 437 | if query:
|
|---|
| 438 | query = unquote(query)
|
|---|
| 439 | fp = gopherlib.send_query(selector, query, host)
|
|---|
| 440 | else:
|
|---|
| 441 | fp = gopherlib.send_selector(selector, host)
|
|---|
| 442 | return addinfourl(fp, noheaders(), "gopher:" + url)
|
|---|
| 443 |
|
|---|
| 444 | def open_file(self, url):
|
|---|
| 445 | """Use local file or FTP depending on form of URL."""
|
|---|
| 446 | if not isinstance(url, str):
|
|---|
| 447 | raise IOError, ('file error', 'proxy support for file protocol currently not implemented')
|
|---|
| 448 | if url[:2] == '//' and url[2:3] != '/' and url[2:12].lower() != 'localhost/':
|
|---|
| 449 | return self.open_ftp(url)
|
|---|
| 450 | else:
|
|---|
| 451 | return self.open_local_file(url)
|
|---|
| 452 |
|
|---|
| 453 | def open_local_file(self, url):
|
|---|
| 454 | """Use local file."""
|
|---|
| 455 | import mimetypes, mimetools, email.Utils
|
|---|
| 456 | try:
|
|---|
| 457 | from cStringIO import StringIO
|
|---|
| 458 | except ImportError:
|
|---|
| 459 | from StringIO import StringIO
|
|---|
| 460 | host, file = splithost(url)
|
|---|
| 461 | localname = url2pathname(file)
|
|---|
| 462 | try:
|
|---|
| 463 | stats = os.stat(localname)
|
|---|
| 464 | except OSError, e:
|
|---|
| 465 | raise IOError(e.errno, e.strerror, e.filename)
|
|---|
| 466 | size = stats.st_size
|
|---|
| 467 | modified = email.Utils.formatdate(stats.st_mtime, usegmt=True)
|
|---|
| 468 | mtype = mimetypes.guess_type(url)[0]
|
|---|
| 469 | headers = mimetools.Message(StringIO(
|
|---|
| 470 | 'Content-Type: %s\nContent-Length: %d\nLast-modified: %s\n' %
|
|---|
| 471 | (mtype or 'text/plain', size, modified)))
|
|---|
| 472 | if not host:
|
|---|
| 473 | urlfile = file
|
|---|
| 474 | if file[:1] == '/':
|
|---|
| 475 | urlfile = 'file://' + file
|
|---|
| 476 | return addinfourl(open(localname, 'rb'),
|
|---|
| 477 | headers, urlfile)
|
|---|
| 478 | host, port = splitport(host)
|
|---|
| 479 | if not port \
|
|---|
| 480 | and socket.gethostbyname(host) in (localhost(), thishost()):
|
|---|
| 481 | urlfile = file
|
|---|
| 482 | if file[:1] == '/':
|
|---|
| 483 | urlfile = 'file://' + file
|
|---|
| 484 | return addinfourl(open(localname, 'rb'),
|
|---|
| 485 | headers, urlfile)
|
|---|
| 486 | raise IOError, ('local file error', 'not on local host')
|
|---|
| 487 |
|
|---|
| 488 | def open_ftp(self, url):
|
|---|
| 489 | """Use FTP protocol."""
|
|---|
| 490 | if not isinstance(url, str):
|
|---|
| 491 | raise IOError, ('ftp error', 'proxy support for ftp protocol currently not implemented')
|
|---|
| 492 | import mimetypes, mimetools
|
|---|
| 493 | try:
|
|---|
| 494 | from cStringIO import StringIO
|
|---|
| 495 | except ImportError:
|
|---|
| 496 | from StringIO import StringIO
|
|---|
| 497 | host, path = splithost(url)
|
|---|
| 498 | if not host: raise IOError, ('ftp error', 'no host given')
|
|---|
| 499 | host, port = splitport(host)
|
|---|
| 500 | user, host = splituser(host)
|
|---|
| 501 | if user: user, passwd = splitpasswd(user)
|
|---|
| 502 | else: passwd = None
|
|---|
| 503 | host = unquote(host)
|
|---|
| 504 | user = unquote(user or '')
|
|---|
| 505 | passwd = unquote(passwd or '')
|
|---|
| 506 | host = socket.gethostbyname(host)
|
|---|
| 507 | if not port:
|
|---|
| 508 | import ftplib
|
|---|
| 509 | port = ftplib.FTP_PORT
|
|---|
| 510 | else:
|
|---|
| 511 | port = int(port)
|
|---|
| 512 | path, attrs = splitattr(path)
|
|---|
| 513 | path = unquote(path)
|
|---|
| 514 | dirs = path.split('/')
|
|---|
| 515 | dirs, file = dirs[:-1], dirs[-1]
|
|---|
| 516 | if dirs and not dirs[0]: dirs = dirs[1:]
|
|---|
| 517 | if dirs and not dirs[0]: dirs[0] = '/'
|
|---|
| 518 | key = user, host, port, '/'.join(dirs)
|
|---|
|
|---|
