1 """Open an arbitrary URL.
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
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/
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)
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.)
30 from urlparse
import urljoin
as basejoin
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"]
40 __version__
= '1.17' # XXX This version is not always updated :-(
42 MAXFTPCACHE
= 10 # Trim the ftp cache beyond this size
44 # Helper for non-unix systems
46 from macurl2path
import url2pathname
, pathname2url
48 from nturl2path
import url2pathname
, pathname2url
49 elif os
.name
== 'riscos':
50 from rourl2path
import url2pathname
, pathname2url
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
)
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
)
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?
69 # Shortcut for basic usage
71 def urlopen(url
, data
=None, proxies
=None):
72 """urlopen(url [, data]) -> open file-like object"""
74 if proxies
is not None:
75 opener
= FancyURLopener(proxies
=proxies
)
77 opener
= FancyURLopener()
82 return opener
.open(url
)
84 return opener
.open(url
, data
)
85 def urlretrieve(url
, filename
=None, reporthook
=None, data
=None):
88 _urlopener
= FancyURLopener()
89 return _urlopener
.retrieve(url
, filename
, reporthook
, data
)
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
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)."""
111 version
= "Python-urllib/%s" % __version__
114 def __init__(self
, proxies
=None, **x509
):
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
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.
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.
148 for file in self
.__tempfiles
:
153 del self
.__tempfiles
[:]
155 self
.tempcache
.clear()
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
)
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
)
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_*()
180 name
= 'open_' + urltype
182 name
= name
.replace('-', '_')
183 if not hasattr(self
, name
):
185 return self
.open_unknown_proxy(proxy
, fullurl
, data
)
187 return self
.open_unknown(fullurl
, data
)
190 return getattr(self
, name
)(url
)
192 return getattr(self
, name
)(url
, data
)
193 except socket
.error
, msg
:
194 raise IOError, ('socket error', msg
), sys
.exc_info()[2]
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)
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
)
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'):
216 fp
= self
.open_local_file(url1
)
219 return url2pathname(splithost(url1
)[1]), hdrs
222 fp
= self
.open(url
, data
)
225 tfp
= open(filename
, 'wb')
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
244 if "content-length" in headers
:
245 size
= int(headers
["Content-Length"])
246 reporthook(blocknum
, bs
, size
)
255 reporthook(blocknum
, bs
, size
)
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
)
268 # Each method named open_<type> knows how to open that type of URL
270 def open_http(self
, url
, data
=None):
271 """Use HTTP protocol."""
275 if isinstance(url
, str):
276 host
, selector
= splithost(url
)
278 user_passwd
, host
= splituser(host
)
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
)
289 if urltype
.lower() != 'http':
292 realhost
, rest
= splithost(rest
)
294 user_passwd
, realhost
= splituser(realhost
)
296 selector
= "%s://%s%s" % (urltype
, realhost
, rest
)
297 if proxy_bypass(realhost
):
300 #print "proxy via http:", host, selector
301 if not host
: raise IOError, ('http error', 'no host given')
305 proxy_auth
= base64
.encodestring(proxy_passwd
).strip()
311 auth
= base64
.encodestring(user_passwd
).strip()
314 h
= httplib
.HTTP(host
)
316 h
.putrequest('POST', selector
)
317 h
.putheader('Content-type', 'application/x-www-form-urlencoded')
318 h
.putheader('Content-length', '%d' % len(data
))
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
)
328 errcode
, errmsg
, headers
= h
.getreply()
331 return addinfourl(fp
, headers
, "http:" + url
)
334 return self
.http_error(url
, fp
, errcode
, errmsg
, headers
)
336 return self
.http_error(url
, fp
, errcode
, errmsg
, headers
, data
)
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
)
347 result
= method(url
, fp
, errcode
, errmsg
, headers
)
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
)
353 def http_error_default(self
, url
, fp
, errcode
, errmsg
, headers
):
354 """Default error handler: close the connection and raise IOError."""
357 raise IOError, ('http error', errcode
, errmsg
, headers
)
359 if hasattr(socket
, "ssl"):
360 def open_https(self
, url
, data
=None):
361 """Use HTTPS protocol."""
365 if isinstance(url
, str):
366 host
, selector
= splithost(url
)
368 user_passwd
, host
= splituser(host
)
373 # here, we determine, whether the proxy contains authorization information
374 proxy_passwd
, host
= splituser(host
)
375 urltype
, rest
= splittype(selector
)
378 if urltype
.lower() != 'https':
381 realhost
, rest
= splithost(rest
)
383 user_passwd
, realhost
= splituser(realhost
)
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')
390 proxy_auth
= base64
.encodestring(proxy_passwd
).strip()
395 auth
= base64
.encodestring(user_passwd
).strip()
398 h
= httplib
.HTTPS(host
, 0,
399 key_file
=self
.key_file
,
400 cert_file
=self
.cert_file
)
402 h
.putrequest('POST', selector
)
403 h
.putheader('Content-type',
404 'application/x-www-form-urlencoded')
405 h
.putheader('Content-length', '%d' % len(data
))
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
)
415 errcode
, errmsg
, headers
= h
.getreply()
418 return addinfourl(fp
, headers
, "https:" + url
)
421 return self
.http_error(url
, fp
, errcode
, errmsg
, headers
)
423 return self
.http_error(url
, fp
, errcode
, errmsg
, headers
,
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')
431 host
, selector
= splithost(url
)
432 if not host
: raise IOError, ('gopher error', 'no host given')
434 type, selector
= splitgophertype(selector
)
435 selector
, query
= splitquery(selector
)
436 selector
= unquote(selector
)
438 query
= unquote(query
)
439 fp
= gopherlib
.send_query(selector
, query
, host
)
441 fp
= gopherlib
.send_selector(selector
, host
)
442 return addinfourl(fp
, noheaders(), "gopher:" + url
)
444 def open_file(self
, url
):
445 if not isinstance(url
, str):
446 raise IOError, ('file error', 'proxy support for file protocol currently not implemented')
447 """Use local file or FTP depending on form of URL."""
448 if url
[:2] == '//' and url
[2:3] != '/' and url
[2:12].lower() != 'localhost/':
449 return self
.open_ftp(url
)
451 return self
.open_local_file(url
)
453 def open_local_file(self
, url
):
454 """Use local file."""
455 import mimetypes
, mimetools
, email
.Utils
457 from cStringIO
import StringIO
459 from StringIO
import StringIO
460 host
, file = splithost(url
)
461 localname
= url2pathname(file)
463 stats
= os
.stat(localname
)
465 raise IOError(e
.errno
, e
.strerror
, e
.filename
)
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
)))
475 urlfile
= 'file://' + file
476 return addinfourl(open(localname
, 'rb'),
478 host
, port
= splitport(host
)
480 and socket
.gethostbyname(host
) in (localhost(), thishost()):
483 urlfile
= 'file://' + file
484 return addinfourl(open(localname
, 'rb'),
486 raise IOError, ('local file error', 'not on local host')
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
494 from cStringIO
import StringIO
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
)
504 user
= unquote(user
or '')
505 passwd
= unquote(passwd
or '')
506 host
= socket
.gethostbyname(host
)
509 port
= ftplib
.FTP_PORT
512 path
, attrs
= splitattr(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
)
520 if len(self
.ftpcache
) > MAXFTPCACHE
:
521 # Prune the cache, rather arbitrarily
522 for k
in self
.ftpcache
.keys():
528 if not key
in self
.ftpcache
:
529 self
.ftpcache
[key
] = \
530 ftpwrapper(user
, passwd
, host
, port
, dirs
)
531 if not file: type = 'D'
534 attr
, value
= splitvalue(attr
)
535 if attr
.lower() == 'type' and \
536 value
in ('a', 'A', 'i', 'I', 'd', 'D'):
538 (fp
, retrlen
) = self
.ftpcache
[key
].retrfile(file, type)
539 mtype
= mimetypes
.guess_type("ftp:" + url
)[0]
542 headers
+= "Content-Type: %s\n" % mtype
543 if retrlen
is not None and retrlen
>= 0:
544 headers
+= "Content-Length: %d\n" % retrlen
545 headers
= mimetools
.Message(StringIO(headers
))
546 return addinfourl(fp
, headers
, "ftp:" + url
)
547 except ftperrors(), msg
:
548 raise IOError, ('ftp error', msg
), sys
.exc_info()[2]
550 def open_data(self
, url
, data
=None):
551 """Use "data" URL."""
552 if not isinstance(url
, str):
553 raise IOError, ('data error', 'proxy support for data protocol currently not implemented')
556 # syntax of data URLs:
557 # dataurl := "data:" [ mediatype ] [ ";base64" ] "," data
558 # mediatype := [ type "/" subtype ] *( ";" parameter )
560 # parameter := attribute "=" value
563 from cStringIO
import StringIO
565 from StringIO
import StringIO
567 [type, data
] = url
.split(',', 1)
569 raise IOError, ('data error', 'bad data URL')
571 type = 'text/plain;charset=US-ASCII'
572 semi
= type.rfind(';')
573 if semi
>= 0 and '=' not in type[semi
:]:
574 encoding
= type[semi
+1:]
579 msg
.append('Date: %s'%time
.strftime('%a, %d %b %Y %T GMT',
580 time
.gmtime(time
.time())))
581 msg
.append('Content-type: %s' % type)
582 if encoding
== 'base64':
584 data
= base64
.decodestring(data
)
587 msg
.append('Content-length: %d' % len(data
))
592 headers
= mimetools
.Message(f
, 0)
593 #f.fileno = None # needed for addinfourl
594 return addinfourl(f
, headers
, url
)
597 class FancyURLopener(URLopener
):
598 """Derived class with handlers for errors we can handle (perhaps)."""
600 def __init__(self
, *args
, **kwargs
):
601 URLopener
.__init
__(self
, *args
, **kwargs
)
606 def http_error_default(self
, url
, fp
, errcode
, errmsg
, headers
):
607 """Default error handling -- don't raise an exception."""
608 return addinfourl(fp
, headers
, "http:" + url
)
610 def http_error_302(self
, url
, fp
, errcode
, errmsg
, headers
, data
=None):
611 """Error 302 -- relocated (temporarily)."""
613 if self
.maxtries
and self
.tries
>= self
.maxtries
:
614 if hasattr(self
, "http_error_500"):
615 meth
= self
.http_error_500
617 meth
= self
.http_error_default
619 return meth(url
, fp
, 500,
620 "Internal Server Error: Redirect Recursion", headers
)
621 result
= self
.redirect_internal(url
, fp
, errcode
, errmsg
, headers
,
626 def redirect_internal(self
, url
, fp
, errcode
, errmsg
, headers
, data
):
627 if 'location' in headers
:
628 newurl
= headers
['location']
629 elif 'uri' in headers
:
630 newurl
= headers
['uri']
635 # In case the server sent a relative URL, join with original:
636 newurl
= basejoin(self
.type + ":" + url
, newurl
)
637 return self
.open(newurl
)
639 def http_error_301(self
, url
, fp
, errcode
, errmsg
, headers
, data
=None):
640 """Error 301 -- also relocated (permanently)."""
641 return self
.http_error_302(url
, fp
, errcode
, errmsg
, headers
, data
)
643 def http_error_303(self
, url
, fp
, errcode
, errmsg
, headers
, data
=None):
644 """Error 303 -- also relocated (essentially identical to 302)."""
645 return self
.http_error_302(url
, fp
, errcode
, errmsg
, headers
, data
)
647 def http_error_307(self
, url
, fp
, errcode
, errmsg
, headers
, data
=None):
648 """Error 307 -- relocated, but turn POST into error."""
650 return self
.http_error_302(url
, fp
, errcode
, errmsg
, headers
, data
)
652 return self
.http_error_default(url
, fp
, errcode
, errmsg
, headers
)
654 def http_error_401(self
, url
, fp
, errcode
, errmsg
, headers
, data
=None):
655 """Error 401 -- authentication required.
656 This function supports Basic authentication only."""
657 if not 'www-authenticate' in headers
:
658 URLopener
.http_error_default(self
, url
, fp
,
659 errcode
, errmsg
, headers
)
660 stuff
= headers
['www-authenticate']
662 match
= re
.match('[ \t]*([^ \t]+)[ \t]+realm="([^"]*)"', stuff
)
664 URLopener
.http_error_default(self
, url
, fp
,
665 errcode
, errmsg
, headers
)
666 scheme
, realm
= match
.groups()
667 if scheme
.lower() != 'basic':
668 URLopener
.http_error_default(self
, url
, fp
,
669 errcode
, errmsg
, headers
)
670 name
= 'retry_' + self
.type + '_basic_auth'
672 return getattr(self
,name
)(url
, realm
)
674 return getattr(self
,name
)(url
, realm
, data
)
676 def http_error_407(self
, url
, fp
, errcode
, errmsg
, headers
, data
=None):
677 """Error 407 -- proxy authentication required.
678 This function supports Basic authentication only."""
679 if not 'proxy-authenticate' in headers
:
680 URLopener
.http_error_default(self
, url
, fp
,
681 errcode
, errmsg
, headers
)
682 stuff
= headers
['proxy-authenticate']
684 match
= re
.match('[ \t]*([^ \t]+)[ \t]+realm="([^"]*)"', stuff
)
686 URLopener
.http_error_default(self
, url
, fp
,
687 errcode
, errmsg
, headers
)
688 scheme
, realm
= match
.groups()
689 if scheme
.lower() != 'basic':
690 URLopener
.http_error_default(self
, url
, fp
,
691 errcode
, errmsg
, headers
)
692 name
= 'retry_proxy_' + self
.type + '_basic_auth'
694 return getattr(self
,name
)(url
, realm
)
696 return getattr(self
,name
)(url
, realm
, data
)
698 def retry_proxy_http_basic_auth(self
, url
, realm
, data
=None):
699 host
, selector
= splithost(url
)
700 newurl
= 'http://' + host
+ selector
701 proxy
= self
.proxies
['http']
702 urltype
, proxyhost
= splittype(proxy
)
703 proxyhost
, proxyselector
= splithost(proxyhost
)
704 i
= proxyhost
.find('@') + 1
705 proxyhost
= proxyhost
[i
:]
706 user
, passwd
= self
.get_user_passwd(proxyhost
, realm
, i
)
707 if not (user
or passwd
): return None
708 proxyhost
= quote(user
, safe
='') + ':' + quote(passwd
, safe
='') + '@' + proxyhost
709 self
.proxies
['http'] = 'http://' + proxyhost
+ proxyselector
711 return self
.open(newurl
)
713 return self
.open(newurl
, data
)
715 def retry_proxy_https_basic_auth(self
, url
, realm
, data
=None):
716 host
, selector
= splithost(url
)
717 newurl
= 'https://' + host
+ selector
718 proxy
= self
.proxies
['https']
719 urltype
, proxyhost
= splittype(proxy
)
720 proxyhost
, proxyselector
= splithost(proxyhost
)
721 i
= proxyhost
.find('@') + 1
722 proxyhost
= proxyhost
[i
:]
723 user
, passwd
= self
.get_user_passwd(proxyhost
, realm
, i
)
724 if not (user
or passwd
): return None
725 proxyhost
= quote(user
, safe
='') + ':' + quote(passwd
, safe
='') + '@' + proxyhost
726 self
.proxies
['https'] = 'https://' + proxyhost
+ proxyselector
728 return self
.open(newurl
)
730 return self
.open(newurl
, data
)
732 def retry_http_basic_auth(self
, url
, realm
, data
=None):
733 host
, selector
= splithost(url
)
734 i
= host
.find('@') + 1
736 user
, passwd
= self
.get_user_passwd(host
, realm
, i
)
737 if not (user
or passwd
): return None
738 host
= quote(user
, safe
='') + ':' + quote(passwd
, safe
='') + '@' + host
739 newurl
= 'http://' + host
+ selector
741 return self
.open(newurl
)
743 return self
.open(newurl
, data
)
745 def retry_https_basic_auth(self
, url
, realm
, data
=None):
746 host
, selector
= splithost(url
)
747 i
= host
.find('@') + 1
749 user
, passwd
= self
.get_user_passwd(host
, realm
, i
)
750 if not (user
or passwd
): return None
751 host
= quote(user
, safe
='') + ':' + quote(passwd
, safe
='') + '@' + host
752 newurl
= 'https://' + host
+ selector
754 return self
.open(newurl
)
756 return self
.open(newurl
, data
)
758 def get_user_passwd(self
, host
, realm
, clear_cache
= 0):
759 key
= realm
+ '@' + host
.lower()
760 if key
in self
.auth_cache
:
762 del self
.auth_cache
[key
]
764 return self
.auth_cache
[key
]
765 user
, passwd
= self
.prompt_user_passwd(host
, realm
)
766 if user
or passwd
: self
.auth_cache
[key
] = (user
, passwd
)
769 def prompt_user_passwd(self
, host
, realm
):
770 """Override this in a GUI environment!"""
773 user
= raw_input("Enter username for %s at %s: " % (realm
,
775 passwd
= getpass
.getpass("Enter password for %s in %s at %s: " %
778 except KeyboardInterrupt:
787 """Return the IP address of the magic hostname 'localhost'."""
789 if _localhost
is None:
790 _localhost
= socket
.gethostbyname('localhost')
795 """Return the IP address of the current host."""
797 if _thishost
is None:
798 _thishost
= socket
.gethostbyname(socket
.gethostname())
803 """Return the set of errors raised by the FTP class."""
805 if _ftperrors
is None:
807 _ftperrors
= ftplib
.all_errors
812 """Return an empty mimetools.Message object."""
814 if _noheaders
is None:
817 from cStringIO
import StringIO
819 from StringIO
import StringIO
820 _noheaders
= mimetools
.Message(StringIO(), 0)
821 _noheaders
.fp
.close() # Recycle file descriptor
828 """Class used by open_ftp() for cache of open FTP connections."""
830 def __init__(self
, user
, passwd
, host
, port
, dirs
):
841 self
.ftp
= ftplib
.FTP()
842 self
.ftp
.connect(self
.host
, self
.port
)
843 self
.ftp
.login(self
.user
, self
.passwd
)
844 for dir in self
.dirs
:
847 def retrfile(self
, file, type):
850 if type in ('d', 'D'): cmd
= 'TYPE A'; isdir
= 1
851 else: cmd
= 'TYPE ' + type; isdir
= 0
853 self
.ftp
.voidcmd(cmd
)
854 except ftplib
.all_errors
:
856 self
.ftp
.voidcmd(cmd
)
858 if file and not isdir
:
859 # Use nlst to see if the file exists at all
862 except ftplib
.error_perm
, reason
:
863 raise IOError, ('ftp error', reason
), sys
.exc_info()[2]
864 # Restore the transfer mode!
865 self
.ftp
.voidcmd(cmd
)
866 # Try to retrieve as a file
869 conn
= self
.ftp
.ntransfercmd(cmd
)
870 except ftplib
.error_perm
, reason
:
871 if str(reason
)[:3] != '550':
872 raise IOError, ('ftp error', reason
), sys
.exc_info()[2]
874 # Set transfer mode to ASCII!
875 self
.ftp
.voidcmd('TYPE A')
876 # Try a directory listing
877 if file: cmd
= 'LIST ' + file
879 conn
= self
.ftp
.ntransfercmd(cmd
)
881 # Pass back both a suitably decorated object and a retrieval length
882 return (addclosehook(conn
[0].makefile('rb'),
883 self
.endtransfer
), conn
[1])
884 def endtransfer(self
):
901 """Base class for addinfo and addclosehook."""
903 def __init__(self
, fp
):
905 self
.read
= self
.fp
.read
906 self
.readline
= self
.fp
.readline
907 if hasattr(self
.fp
, "readlines"): self
.readlines
= self
.fp
.readlines
908 if hasattr(self
.fp
, "fileno"):
909 self
.fileno
= self
.fp
.fileno
911 self
.fileno
= lambda: None
912 if hasattr(self
.fp
, "__iter__"):
913 self
.__iter
__ = self
.fp
.__iter
__
914 if hasattr(self
.fp
, "next"):
915 self
.next
= self
.fp
.next
918 return '<%s at %r whose fp = %r>' % (self
.__class
__.__name
__,
924 self
.readlines
= None
926 if self
.fp
: self
.fp
.close()
929 class addclosehook(addbase
):
930 """Class to add a close hook to an open file."""
932 def __init__(self
, fp
, closehook
, *hookargs
):
933 addbase
.__init
__(self
, fp
)
934 self
.closehook
= closehook
935 self
.hookargs
= hookargs
940 self
.closehook(*self
.hookargs
)
941 self
.closehook
= None
944 class addinfo(addbase
):
945 """class to add an info() method to an open file."""
947 def __init__(self
, fp
, headers
):
948 addbase
.__init
__(self
, fp
)
949 self
.headers
= headers
954 class addinfourl(addbase
):
955 """class to add info() and geturl() methods to an open file."""
957 def __init__(self
, fp
, headers
, url
):
958 addbase
.__init
__(self
, fp
)
959 self
.headers
= headers
969 # Utilities to parse URLs (most of these return None for missing parts):
970 # unwrap('<URL:type://host/path>') --> 'type://host/path'
971 # splittype('type:opaquestring') --> 'type', 'opaquestring'
972 # splithost('//host[:port]/path') --> 'host[:port]', '/path'
973 # splituser('user[:passwd]@host[:port]') --> 'user[:passwd]', 'host[:port]'
974 # splitpasswd('user:passwd') -> 'user', 'passwd'
975 # splitport('host:port') --> 'host', 'port'
976 # splitquery('/path?query') --> '/path', 'query'
977 # splittag('/path#tag') --> '/path', 'tag'
978 # splitattr('/path;attr1=value1;attr2=value2;...') ->
979 # '/path', ['attr1=value1', 'attr2=value2', ...]
980 # splitvalue('attr=value') --> 'attr', 'value'
981 # splitgophertype('/Xselector') --> 'X', 'selector'
982 # unquote('abc%20def') -> 'abc def'
983 # quote('abc def') -> 'abc%20def')
992 return isinstance(x
, unicode)
995 """toBytes(u"URL") --> 'URL'."""
996 # Most URL schemes require ASCII. If that changes, the conversion
1000 url
= url
.encode("ASCII")
1001 except UnicodeError:
1002 raise UnicodeError("URL " + repr(url
) +
1003 " contains non-ASCII characters")
1007 """unwrap('<URL:type://host/path>') --> 'type://host/path'."""
1009 if url
[:1] == '<' and url
[-1:] == '>':
1010 url
= url
[1:-1].strip()
1011 if url
[:4] == 'URL:': url
= url
[4:].strip()
1016 """splittype('type:opaquestring') --> 'type', 'opaquestring'."""
1018 if _typeprog
is None:
1020 _typeprog
= re
.compile('^([^/:]+):')
1022 match
= _typeprog
.match(url
)
1024 scheme
= match
.group(1)
1025 return scheme
.lower(), url
[len(scheme
) + 1:]
1030 """splithost('//host[:port]/path') --> 'host[:port]', '/path'."""
1032 if _hostprog
is None:
1034 _hostprog
= re
.compile('^//([^/]*)(.*)$')
1036 match
= _hostprog
.match(url
)
1037 if match
: return match
.group(1, 2)
1041 def splituser(host
):
1042 """splituser('user[:passwd]@host[:port]') --> 'user[:passwd]', 'host[:port]'."""
1044 if _userprog
is None:
1046 _userprog
= re
.compile('^(.*)@(.*)$')
1048 match
= _userprog
.match(host
)
1049 if match
: return map(unquote
, match
.group(1, 2))
1053 def splitpasswd(user
):
1054 """splitpasswd('user:passwd') -> 'user', 'passwd'."""
1056 if _passwdprog
is None:
1058 _passwdprog
= re
.compile('^([^:]*):(.*)$')
1060 match
= _passwdprog
.match(user
)
1061 if match
: return match
.group(1, 2)
1064 # splittag('/path#tag') --> '/path', 'tag'
1066 def splitport(host
):
1067 """splitport('host:port') --> 'host', 'port'."""
1069 if _portprog
is None:
1071 _portprog
= re
.compile('^(.*):([0-9]+)$')
1073 match
= _portprog
.match(host
)
1074 if match
: return match
.group(1, 2)
1078 def splitnport(host
, defport
=-1):
1079 """Split host and port, returning numeric port.
1080 Return given default port if no ':' found; defaults to -1.
1081 Return numerical port if a valid number are found after ':'.
1082 Return None if ':' but not a valid number."""
1084 if _nportprog
is None:
1086 _nportprog
= re
.compile('^(.*):(.*)$')
1088 match
= _nportprog
.match(host
)
1090 host
, port
= match
.group(1, 2)
1092 if not port
: raise ValueError, "no digits"
1097 return host
, defport
1100 def splitquery(url
):
1101 """splitquery('/path?query') --> '/path', 'query'."""
1103 if _queryprog
is None:
1105 _queryprog
= re
.compile('^(.*)\?([^?]*)$')
1107 match
= _queryprog
.match(url
)
1108 if match
: return match
.group(1, 2)
1113 """splittag('/path#tag') --> '/path', 'tag'."""
1115 if _tagprog
is None:
1117 _tagprog
= re
.compile('^(.*)#([^#]*)$')
1119 match
= _tagprog
.match(url
)
1120 if match
: return match
.group(1, 2)
1124 """splitattr('/path;attr1=value1;attr2=value2;...') ->
1125 '/path', ['attr1=value1', 'attr2=value2', ...]."""
1126 words
= url
.split(';')
1127 return words
[0], words
[1:]
1130 def splitvalue(attr
):
1131 """splitvalue('attr=value') --> 'attr', 'value'."""
1133 if _valueprog
is None:
1135 _valueprog
= re
.compile('^([^=]*)=(.*)$')
1137 match
= _valueprog
.match(attr
)
1138 if match
: return match
.group(1, 2)
1141 def splitgophertype(selector
):
1142 """splitgophertype('/Xselector') --> 'X', 'selector'."""
1143 if selector
[:1] == '/' and selector
[1:2]:
1144 return selector
[1], selector
[2:]
1145 return None, selector
1147 _hextochr
= dict(('%02x' % i
, chr(i
)) for i
in range(256))
1148 _hextochr
.update(('%02X' % i
, chr(i
)) for i
in range(256))
1151 """unquote('abc%20def') -> 'abc def'."""
1153 for i
in xrange(1, len(res
)):
1156 res
[i
] = _hextochr
[item
[:2]] + item
[2:]
1159 except UnicodeDecodeError:
1160 res
[i
] = unichr(int(item
[:2], 16)) + item
[2:]
1163 def unquote_plus(s
):
1164 """unquote('%7e/abc+def') -> '~/abc def'"""
1165 s
= s
.replace('+', ' ')
1168 always_safe
= ('ABCDEFGHIJKLMNOPQRSTUVWXYZ'
1169 'abcdefghijklmnopqrstuvwxyz'
1173 def quote(s
, safe
= '/'):
1174 """quote('abc def') -> 'abc%20def'
1176 Each part of a URL, e.g. the path info, the query, etc., has a
1177 different set of reserved characters that must be quoted.
1179 RFC 2396 Uniform Resource Identifiers (URI): Generic Syntax lists
1180 the following reserved characters.
1182 reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" |
1185 Each of these characters is reserved in some component of a URL,
1186 but not necessarily in all of them.
1188 By default, the quote function is intended for quoting the path
1189 section of a URL. Thus, it will not encode '/'. This character
1190 is reserved, but in typical usage the quote function is being
1191 called on a path where the existing slash characters are used as
1192 reserved characters.
1194 cachekey
= (safe
, always_safe
)
1196 safe_map
= _safemaps
[cachekey
]
1200 for i
in range(256):
1202 safe_map
[c
] = (c
in safe
) and c
or ('%%%02X' % i
)
1203 _safemaps
[cachekey
] = safe_map
1204 res
= map(safe_map
.__getitem
__, s
)
1207 def quote_plus(s
, safe
= ''):
1208 """Quote the query fragment of a URL; replacing ' ' with '+'"""
1210 s
= quote(s
, safe
+ ' ')
1211 return s
.replace(' ', '+')
1212 return quote(s
, safe
)
1214 def urlencode(query
,doseq
=0):
1215 """Encode a sequence of two-element tuples or dictionary into a URL query string.
1217 If any values in the query arg are sequences and doseq is true, each
1218 sequence element is converted to a separate parameter.
1220 If the query arg is a sequence of two-element tuples, the order of the
1221 parameters in the output will match the order of parameters in the
1225 if hasattr(query
,"items"):
1227 query
= query
.items()
1229 # it's a bother at times that strings and string-like objects are
1232 # non-sequence items should not work with len()
1233 # non-empty strings will fail this
1234 if len(query
) and not isinstance(query
[0], tuple):
1236 # zero-length sequences of all types will get here and succeed,
1237 # but that's a minor nit - since the original implementation
1238 # allowed empty dicts that type of behavior probably should be
1239 # preserved for consistency
1241 ty
,va
,tb
= sys
.exc_info()
1242 raise TypeError, "not a valid non-string sequence or mapping object", tb
1246 # preserve old behavior
1248 k
= quote_plus(str(k
))
1249 v
= quote_plus(str(v
))
1250 l
.append(k
+ '=' + v
)
1253 k
= quote_plus(str(k
))
1254 if isinstance(v
, str):
1256 l
.append(k
+ '=' + v
)
1257 elif _is_unicode(v
):
1258 # is there a reasonable way to convert to ASCII?
1259 # encode generates a string, but "replace" or "ignore"
1260 # lose information and "strict" can raise UnicodeError
1261 v
= quote_plus(v
.encode("ASCII","replace"))
1262 l
.append(k
+ '=' + v
)
1265 # is this a sufficient test for sequence-ness?
1269 v
= quote_plus(str(v
))
1270 l
.append(k
+ '=' + v
)
1272 # loop over the sequence
1274 l
.append(k
+ '=' + quote_plus(str(elt
)))
1278 def getproxies_environment():
1279 """Return a dictionary of scheme -> proxy server URL mappings.
1281 Scan the environment for variables named <scheme>_proxy;
1282 this seems to be the standard convention. If you need a
1283 different way, you can pass a proxies dictionary to the
1284 [Fancy]URLopener constructor.
1288 for name
, value
in os
.environ
.items():
1290 if value
and name
[-6:] == '_proxy':
1291 proxies
[name
[:-6]] = value
1294 if sys
.platform
== 'darwin':
1295 def getproxies_internetconfig():
1296 """Return a dictionary of scheme -> proxy server URL mappings.
1298 By convention the mac uses Internet Config to store
1299 proxies. An HTTP proxy, for instance, is stored under
1314 if 'UseHTTPProxy' in config
and config
['UseHTTPProxy']:
1316 value
= config
['HTTPProxyHost']
1320 proxies
['http'] = 'http://%s' % value
1321 # FTP: XXXX To be done.
1322 # Gopher: XXXX To be done.
1325 def proxy_bypass(x
):
1329 return getproxies_environment() or getproxies_internetconfig()
1331 elif os
.name
== 'nt':
1332 def getproxies_registry():
1333 """Return a dictionary of scheme -> proxy server URL mappings.
1335 Win32 uses the registry to store proxies.
1342 # Std module, so should be around - but you never know!
1345 internetSettings
= _winreg
.OpenKey(_winreg
.HKEY_CURRENT_USER
,
1346 r
'Software\Microsoft\Windows\CurrentVersion\Internet Settings')
1347 proxyEnable
= _winreg
.QueryValueEx(internetSettings
,
1350 # Returned as Unicode but problems if not converted to ASCII
1351 proxyServer
= str(_winreg
.QueryValueEx(internetSettings
,
1353 if '=' in proxyServer
:
1354 # Per-protocol settings
1355 for p
in proxyServer
.split(';'):
1356 protocol
, address
= p
.split('=', 1)
1357 # See if address has a type:// prefix
1359 if not re
.match('^([^/:]+)://', address
):
1360 address
= '%s://%s' % (protocol
, address
)
1361 proxies
[protocol
] = address
1363 # Use one setting for all protocols
1364 if proxyServer
[:5] == 'http:':
1365 proxies
['http'] = proxyServer
1367 proxies
['http'] = 'http://%s' % proxyServer
1368 proxies
['ftp'] = 'ftp://%s' % proxyServer
1369 internetSettings
.Close()
1370 except (WindowsError, ValueError, TypeError):
1371 # Either registry key not found etc, or the value in an
1372 # unexpected format.
1373 # proxies already set up to be empty so nothing to do
1378 """Return a dictionary of scheme -> proxy server URL mappings.
1380 Returns settings gathered from the environment, if specified,
1384 return getproxies_environment() or getproxies_registry()
1386 def proxy_bypass(host
):
1391 # Std modules, so should be around - but you never know!
1394 internetSettings
= _winreg
.OpenKey(_winreg
.HKEY_CURRENT_USER
,
1395 r
'Software\Microsoft\Windows\CurrentVersion\Internet Settings')
1396 proxyEnable
= _winreg
.QueryValueEx(internetSettings
,
1398 proxyOverride
= str(_winreg
.QueryValueEx(internetSettings
,
1399 'ProxyOverride')[0])
1400 # ^^^^ Returned as Unicode but problems if not converted to ASCII
1401 except WindowsError:
1403 if not proxyEnable
or not proxyOverride
:
1405 # try to make a host list from name and IP address.
1406 rawHost
, port
= splitport(host
)
1409 addr
= socket
.gethostbyname(rawHost
)
1412 except socket
.error
:
1415 fqdn
= socket
.getfqdn(rawHost
)
1418 except socket
.error
:
1420 # make a check value list from the registry entry: replace the
1421 # '<local>' string by the localhost entry and the corresponding
1423 proxyOverride
= proxyOverride
.split(';')
1425 while i
< len(proxyOverride
):
1426 if proxyOverride
[i
] == '<local>':
1427 proxyOverride
[i
:i
+1] = ['localhost',
1429 socket
.gethostname(),
1430 socket
.gethostbyname(
1431 socket
.gethostname())]
1433 # print proxyOverride
1434 # now check if we match one of the registry values.
1435 for test
in proxyOverride
:
1436 test
= test
.replace(".", r
"\.") # mask dots
1437 test
= test
.replace("*", r
".*") # change glob sequence
1438 test
= test
.replace("?", r
".") # change glob char
1440 # print "%s <--> %s" %( test, val )
1441 if re
.match(test
, val
, re
.I
):
1446 # By default use environment variables
1447 getproxies
= getproxies_environment
1449 def proxy_bypass(host
):
1452 # Test and time quote() and unquote()
1455 for i
in range(256): s
= s
+ chr(i
)
1466 print round(t1
- t0
, 3), 'sec'
1469 def reporthook(blocknum
, blocksize
, totalsize
):
1470 # Report during remote transfers
1471 print "Block number: %d, Block size: %d, Total size: %d" % (
1472 blocknum
, blocksize
, totalsize
)
1480 'file://localhost/etc/passwd',
1481 'ftp://ftp.python.org/pub/python/README',
1482 ## 'gopher://gopher.micro.umn.edu/1/',
1483 'http://www.python.org/index.html',
1485 if hasattr(URLopener
, "open_https"):
1486 args
.append('https://synergy.as.cmu.edu/~geek/')
1489 print '-'*10, url
, '-'*10
1490 fn
, h
= urlretrieve(url
, None, reporthook
)
1494 for k
in h
.keys(): print k
+ ':', h
[k
]
1500 table
= string
.maketrans("", "")
1501 data
= data
.translate(table
, "\r")
1511 opts
, args
= getopt
.getopt(sys
.argv
[1:], "th")
1512 except getopt
.error
, msg
:
1514 print "Use -h for help"
1521 print "Usage: python urllib.py [-t] [url ...]"
1522 print "-t runs self-test;",
1523 print "otherwise, contents of urls are printed"
1531 print "Use -h for help"
1533 print urlopen(url
).read(),
1535 # Run test program when run as a script
1536 if __name__
== '__main__':