| blob | 655ea387c8dd4f30c6cc8723059140c90de6b4e4 |
1 ;;; url-http.el --- HTTP retrieval routines
3 ;; Copyright (C) 1999, 2001, 2004, 2005, 2006, 2007, 2008,
4 ;; 2009, 2010 Free Software Foundation, Inc.
6 ;; Author: Bill Perry <wmperry@gnu.org>
7 ;; Keywords: comm, data, processes
9 ;; This file is part of GNU Emacs.
10 ;;
11 ;; GNU Emacs is free software: you can redistribute it and/or modify
12 ;; it under the terms of the GNU General Public License as published by
13 ;; the Free Software Foundation, either version 3 of the License, or
14 ;; (at your option) any later version.
16 ;; GNU Emacs is distributed in the hope that it will be useful,
17 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
19 ;; GNU General Public License for more details.
21 ;; You should have received a copy of the GNU General Public License
22 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
24 ;;; Commentary:
26 ;;; Code:
54 "What version of HTTP we advertise, as a string.
55 Valid values are 1.1 and 1.0.
56 This is only useful when debugging the HTTP subsystem.
58 Setting this to 1.0 will tell servers not to send chunked encoding,
62 "Whether to use a single TCP connection multiple times in HTTP.
63 This is only useful when debugging the HTTP subsystem. Setting to
64 nil will explicitly close the connection to the server after every
116 ;(eval-when-compile
117 ;; These are all macros so that they are hidden from external sight
118 ;; when the file is byte-compiled.
119 ;;
120 ;; This allows us to expose just the entry points we want.
122 ;; These routines will allow us to implement persistent HTTP
123 ;; connections.
127 ;; The user hit C-g, honor it! Some things can get in an
128 ;; incredibly tight loop (chunked encoding)
141 url-http-open-connections)
142 proc)
152 url-http-open-connections))
153 nil)
172 host port
175 ;; `url-open-stream' needs a buffer in which to do things
176 ;; like authentication. But we use another buffer afterwards.
179 ;; url-open-stream might return nil.
181 ;; Drop the temp buffer link before killing the buffer.
183 proc)
186 ;; Building an HTTP request
191 ""
196 url-version
205 "Create an HTTP request for `url-http-target-url', referred to by REF-URL."
207 url-http-method url-http-data
208 url-http-extra-headers))
214 url-http-extra-headers))
216 nil
224 nil
227 proxy-info)
237 ;; Protection against stupid values in the referer
242 ;; We do not want to expose the referer if the user is paranoid.
248 ;; url-http-extra-headers contains an assoc-list of
249 ;; header/value pairs that we need to put into the request.
257 ;; This was done with a call to `format'. Concatting parts has
258 ;; the advantage of keeping the parts of each header together and
259 ;; allows us to elide null lines directly, at the cost of making
260 ;; the layout less clear.
262 ;; We used to concat directly, but if one of the strings happens
263 ;; to being multibyte (even if it only contains pure ASCII) then
264 ;; every string gets converted with `string-MAKE-multibyte' which
265 ;; turns the 127-255 codes into things like latin-1 accented chars
266 ;; (it would work right if it used `string-TO-multibyte' instead).
267 ;; So to avoid the problem we force every string to be unibyte.
269 ;; FIXME: Instead of `string-AS-unibyte' we'd want
270 ;; `string-to-unibyte', so as to properly signal an error if one
271 ;; of the strings contains a multibyte char.
272 'string-as-unibyte
275 ;; The request
279 ;; Version of MIME we speak
281 ;; (maybe) Try to keep the connection open
285 ;; HTTP extensions we support
289 ;; Who we want to talk to
296 ;; Who its from
300 ;; Encodings we understand
307 ;; Languages we understand
311 ;; Types we understand
313 ;; User agent
315 ;; Proxy Authorization
316 proxy-auth
317 ;; Authorization
318 auth
319 ;; Cookies
322 ;; If-modified-since
329 ;; Whence we came
332 extra-headers
333 ;; Length of data
339 ;; End request
341 ;; Any data
345 request))
347 ;; Parsing routines
350 This allows us to use `mail-fetch-field', etc."
358 url-callback-function url-callback-arguments))
363 nil nil t))
369 url-http-proxy
370 url-current-object)))
372 ;; Cheating, but who cares? :)
373 'url-http-proxy-basic-auth-storage
375 auth
378 ;; find strongest supported auth
382 this-auth)))
391 type this-type
392 strength this-strength))))
399 " authentication. If you'd like to write it,"
405 type t args)))
409 url-http-extra-headers)
414 url-callback-arguments)))))))
417 "Parse just the response code."
419 url-http-response-version))
434 "Handle all set-cookie / set-cookie2 headers in an HTTP response.
435 The buffer must already be narrowed to the headers, so `mail-fetch-field' will
436 work correctly."
443 ;;; (while cookies2
444 ;;; (url-cookie-handle-set-cookie2 (pop cookies)))
445 )
446 )
449 "Parse and handle HTTP specific headers.
450 Return t if and only if the current buffer is still active and
451 should be shown to the user."
452 ;; The comments after each status code handled are taken from RFC
453 ;; 2616 (HTTP/1.1)
455 url-http-response-version
456 url-http-method url-http-data url-http-process
457 url-callback-function url-callback-arguments))
461 url-http-process)
470 ;;(narrow-to-region (point-min) url-http-end-of-headers)
472 ;; In HTTP 1.0, keep the connection only if there is a
473 ;; "Connection: keep-alive" header.
474 ;; In HTTP 1.1 (and greater), keep the connection unless there is a
475 ;; "Connection: close" header
488 ;; other status symbols: jewelry and luxury cars
490 ;; The filename part of a URL could be in remote file syntax,
491 ;; see Bug#6717 for an example. We disable file name
492 ;; handlers, therefore.
499 ;; Classes of response codes
500 ;;
501 ;; 5xx = Server Error
502 ;; 4xx = Client Error
503 ;; 3xx = Redirection
504 ;; 2xx = Successful
505 ;; 1xx = Informational
507 ;; 100 = Continue with request
508 ;; 101 = Switching protocols
509 ;; 102 = Processing (Added by DAV)
513 ;; 200 Ok
514 ;; 201 Created
515 ;; 202 Accepted
516 ;; 203 Non-authoritative information
517 ;; 204 No content
518 ;; 205 Reset content
519 ;; 206 Partial content
520 ;; 207 Multi-status (Added by DAV)
523 ;; No new data, just stay at the same document
527 ;; Generic success for all others. Store in the cache, and
528 ;; mark it as successful.
534 ;; 300 Multiple choices
535 ;; 301 Moved permanently
536 ;; 302 Found
537 ;; 303 See other
538 ;; 304 Not modified
539 ;; 305 Use proxy
540 ;; 307 Temporary redirect
545 ;; Quoth the spec (section 10.3.1)
546 ;; -------------------------------
547 ;; The requested resource corresponds to any one of a set of
548 ;; representations, each with its own specific location and
549 ;; agent-driven negotiation information is being provided so
550 ;; that the user can select a preferred representation and
551 ;; redirect its request to that location.
552 ;; [...]
553 ;; If the server has a preferred choice of representation, it
554 ;; SHOULD include the specific URI for that representation in
555 ;; the Location field; user agents MAY use the Location field
556 ;; value for automatic redirection.
557 ;; -------------------------------
558 ;; We do not support agent-driven negotiation, so we just
559 ;; redirect to the preferred URI if one is provided.
560 nil)
562 ;; If the 301|302 status code is received in response to a
563 ;; request other than GET or HEAD, the user agent MUST NOT
564 ;; automatically redirect the request unless it can be
565 ;; confirmed by the user, since this might change the
566 ;; conditions under which the request was issued.
568 ;; Automatic redirection is ok
569 nil
570 ;; It is just too big of a pain in the ass to get this
571 ;; prompt all the time. We will just silently lose our
572 ;; data and convert to a GET method.
574 url-http-method url-http-response-status)
576 url-http-data nil)))
578 ;; The response to the request can be found under a different
579 ;; URI and SHOULD be retrieved using a GET method on that
580 ;; resource.
582 url-http-data nil))
584 ;; The 304 response MUST NOT contain a message-body.
589 success t))
591 ;; The requested resource MUST be accessed through the
592 ;; proxy given by the Location field. The Location field
593 ;; gives the URI of the proxy. The recipient is expected
594 ;; to repeat this single request via the proxy. 305
595 ;; responses MUST only be generated by origin servers.
597 redirect-uri))
599 ;; Treat everything like '300'
600 nil))
602 ;; Clean off any whitespace and/or <...> cruft.
608 ;; Some stupid sites (like sourceforge) send a
609 ;; non-fully-qualified URL (ie: /), which royally confuses
610 ;; the URL library.
612 ;; Be careful to use the real target URL, otherwise we may
613 ;; compute the redirection relative to the URL of the proxy.
619 ;; Check existing number of redirects
630 ;; url-max-redirections hasn't been reached, so go
631 ;; ahead and redirect.
633 ;; Remember that the request was redirected.
637 ;; Put in the current buffer a forwarding pointer to the new
638 ;; destination buffer.
639 ;; FIXME: This is a hack to fix url-retrieve-synchronously
640 ;; without changing the API. Instead url-retrieve should
641 ;; either simply not return the "destination" buffer, or it
642 ;; should take an optional `dest-buf' argument.
645 redirect-uri url-callback-function
646 url-callback-arguments))
648 ;; We hit url-max-redirections, so issue an error and
649 ;; stop redirecting.
653 redirect-uri))
657 ;; 400 Bad Request
658 ;; 401 Unauthorized
659 ;; 402 Payment required
660 ;; 403 Forbidden
661 ;; 404 Not found
662 ;; 405 Method not allowed
663 ;; 406 Not acceptable
664 ;; 407 Proxy authentication required
665 ;; 408 Request time-out
666 ;; 409 Conflict
667 ;; 410 Gone
668 ;; 411 Length required
669 ;; 412 Precondition failed
670 ;; 413 Request entity too large
671 ;; 414 Request-URI too large
672 ;; 415 Unsupported media type
673 ;; 416 Requested range not satisfiable
674 ;; 417 Expectation failed
675 ;; 422 Unprocessable Entity (Added by DAV)
676 ;; 423 Locked
677 ;; 424 Failed Dependency
680 ;; The request requires user authentication. The response
681 ;; MUST include a WWW-Authenticate header field containing a
682 ;; challenge applicable to the requested resource. The
683 ;; client MAY repeat the request with a suitable
684 ;; Authorization header field.
687 ;; This code is reserved for future use
691 ;; The server understood the request, but is refusing to
692 ;; fulfill it. Authorization will not help and the request
693 ;; SHOULD NOT be repeated.
696 ;; Not found
699 ;; The method specified in the Request-Line is not allowed
700 ;; for the resource identified by the Request-URI. The
701 ;; response MUST include an Allow header containing a list of
702 ;; valid methods for the requested resource.
705 ;; The resource identified by the request is only capable of
706 ;; generating response entities which have content
707 ;; characteristics nota cceptable according to the accept
708 ;; headers sent in the request.
711 ;; This code is similar to 401 (Unauthorized), but indicates
712 ;; that the client must first authenticate itself with the
713 ;; proxy. The proxy MUST return a Proxy-Authenticate header
714 ;; field containing a challenge applicable to the proxy for
715 ;; the requested resource.
718 ;; The client did not produce a request within the time that
719 ;; the server was prepared to wait. The client MAY repeat
720 ;; the request without modifications at any later time.
723 ;; The request could not be completed due to a conflict with
724 ;; the current state of the resource. This code is only
725 ;; allowed in situations where it is expected that the user
726 ;; mioght be able to resolve the conflict and resubmit the
727 ;; request. The response body SHOULD include enough
728 ;; information for the user to recognize the source of the
729 ;; conflict.
732 ;; The requested resource is no longer available at the
733 ;; server and no forwarding address is known.
736 ;; The server refuses to accept the request without a defined
737 ;; Content-Length. The client MAY repeat the request if it
738 ;; adds a valid Content-Length header field containing the
739 ;; length of the message-body in the request message.
740 ;;
741 ;; NOTE - this will never happen because
742 ;; `url-http-create-request' automatically calculates the
743 ;; content-length.
746 ;; The precondition given in one or more of the
747 ;; request-header fields evaluated to false when it was
748 ;; tested on the server.
751 ;; The server is refusing to process a request because the
752 ;; request entity|URI is larger than the server is willing or
753 ;; able to process.
756 ;; The server is refusing to service the request because the
757 ;; entity of the request is in a format not supported by the
758 ;; requested resource for the requested method.
761 ;; A server SHOULD return a response with this status code if
762 ;; a request included a Range request-header field, and none
763 ;; of the range-specifier values in this field overlap the
764 ;; current extent of the selected resource, and the request
765 ;; did not include an If-Range request-header field.
768 ;; The expectation given in an Expect request-header field
769 ;; could not be met by this server, or, if the server is a
770 ;; proxy, the server has unambiguous evidence that the
771 ;; request could not be met by the next-hop server.
774 ;; The request could not be understood by the server due to
775 ;; malformed syntax. The client SHOULD NOT repeat the
776 ;; request without modifications.
778 ;; Tell the callback that an error occurred, and what the
779 ;; status code was.
785 ;; 500 Internal server error
786 ;; 501 Not implemented
787 ;; 502 Bad gateway
788 ;; 503 Service unavailable
789 ;; 504 Gateway time-out
790 ;; 505 HTTP version not supported
791 ;; 507 Insufficient storage
795 ;; The server does not support the functionality required to
796 ;; fulfill the request.
797 nil)
799 ;; The server, while acting as a gateway or proxy, received
800 ;; an invalid response from the upstream server it accessed
801 ;; in attempting to fulfill the request.
802 nil)
804 ;; The server is currently unable to handle the request due
805 ;; to a temporary overloading or maintenance of the server.
806 ;; The implication is that this is a temporary condition
807 ;; which will be alleviated after some delay. If known, the
808 ;; length of the delay MAY be indicated in a Retry-After
809 ;; header. If no Retry-After is given, the client SHOULD
810 ;; handle the response as it would for a 500 response.
811 nil)
813 ;; The server, while acting as a gateway or proxy, did not
814 ;; receive a timely response from the upstream server
815 ;; specified by the URI (e.g. HTTP, FTP, LDAP) or some other
816 ;; auxiliary server (e.g. DNS) it needed to access in
817 ;; attempting to complete the request.
818 nil)
820 ;; The server does not support, or refuses to support, the
821 ;; HTTP protocol version that was used in the request
822 ;; message.
823 nil)
825 ;; The method could not be performed on the resource
826 ;; because the server is unable to store the representation
827 ;; needed to successfully complete the request. This
828 ;; condition is considered to be temporary. If the request
829 ;; which received this status code was the result of a user
830 ;; action, the request MUST NOT be repeated until it is
831 ;; requested by a separate user action.
832 nil))
833 ;; Tell the callback that an error occurred, and what the
834 ;; status code was.
841 class url-http-response-status)))
846 success))
848 ;; Miscellaneous
850 "Activate callback specified when this buffer was created."
852 url-callback-function
853 url-callback-arguments))
856 url-http-process)
860 ;; )
862 ;; These unfortunately cannot be macros... please ignore them!
864 "Remove (now defunct) process PROC from the list of open connections."
868 url-http-open-connections))
871 ;; Sentinel used for old HTTP/0.9 or connections we know are going
872 ;; to die as the 'end of document' notifier.
879 ;; HTTP/0.9 just gets passed back no matter what
885 ;; Function used when we do NOT know how long the document is going to be
886 ;; Just _very_ simple 'downloaded %d' type of info.
891 "Function used when we DO know how long the document is going to be.
892 More sophisticated percentage downloaded, etc.
893 Also does minimal parsing of HTTP headers and will actually cause
894 the callback to be triggered."
896 url-http-end-of-headers
897 url-http-content-length
898 url-http-content-type
899 url-http-process))
902 "Reading [%s]... %s of %s (%d%%)"
904 url-http-content-length)
905 url-http-content-type
909 url-http-content-length))
911 "Reading... %s of %s (%d%%)"
913 url-http-content-length)
917 url-http-content-length)))
921 ;; Found the end of the document! Wheee!
928 "Function used when dealing with 'chunked' encoding.
929 Cannot give a sophisticated percentage, but we need a different
930 function to look for the special 0-length chunk that signifies
931 the end of the document."
933 url-http-end-of-headers
934 url-http-content-type
935 url-http-chunked-length
936 url-http-chunked-counter
937 url-http-process url-http-chunked-start))
944 ;; We need to loop thru looking for more chunks even within
945 ;; one after-change-function call.
950 "Reading [%s]... chunk #%d"
951 url-http-content-type url-http-chunked-counter)
953 "Reading... chunk #%d"
954 url-http-chunked-counter))
956 url-http-chunked-counter st nd length)
962 ;; We know how long the chunk is supposed to be, skip over
963 ;; leading crap if possible.
967 url-http-chunked-counter)
969 url-http-chunked-length)))
972 url-http-chunked-length)
973 nd))
980 ;; Must not have received the entirety of the chunk header,
981 ;; need to spin some more.
986 'end-open t
987 'chunked-encoding t
999 ; (if (not url-http-debug)
1002 url-http-chunked-counter url-http-chunked-length
1006 ;; Found the end of the document! Wheee!
1010 ;; Every chunk, even the last 0-length one, is
1011 ;; terminated by CRLF. Skip it.
1021 ;; This will wait for the headers to arrive and then splice in the
1022 ;; next appropriate after-change-function, etc.
1024 url-http-end-of-headers
1025 url-http-content-type
1026 url-http-content-length
1027 url-http-transfer-encoding
1028 url-callback-function
1029 url-callback-arguments
1030 url-http-process
1031 url-http-method
1032 url-http-after-change-function
1033 url-http-response-status))
1043 ;; Not HTTP/x.y data, must be 0.9
1044 ;; God, I wish this could die.
1046 url-http-end-of-headers 0
1047 old-http t)
1049 ;; Saw the end of the headers
1053 end-of-headers t)
1057 ;; Haven't seen the end of the headers yet, need to wait
1058 ;; for more data to arrive.
1059 nil
1065 ;;(narrow-to-region (point-min) url-http-end-of-headers)
1085 ;; A HEAD request is _ALWAYS_ terminated by the header
1086 ;; information, regardless of any entity headers,
1087 ;; according to section 4.4 of the HTTP/1.1 draft.
1093 ;; A CONNECT request is finished, but we cannot stick this
1094 ;; back on the free connectin list
1099 ;; Only allowed to have a header section. We have to handle
1100 ;; this here instead of in url-http-parse-headers because if
1101 ;; you have a cached copy of something without a known
1102 ;; content-length, and try to retrieve it from the cache, we'd
1103 ;; fall into the 'being dumb' section and wait for the
1104 ;; connection to terminate, which means we'd wait for 10
1105 ;; seconds for the keep-alives to time out on some servers.
1109 ;; HTTP/0.9 always signaled end-of-connection by closing the
1110 ;; connection.
1132 ;; We got a NULL body! Activate the callback
1133 ;; immediately!
1139 ;; Have some leftover data
1143 nd
1146 nil)))
1151 ;; We are still at the beginning of the buffer... must just be
1152 ;; waiting for a response.
1156 ;;;###autoload
1158 "Retrieve URL via HTTP asynchronously.
1159 URL must be a parsed URL. See `url-generic-parse-url' for details.
1160 When retrieval is completed, the function CALLBACK is executed with
1161 CBARGS as the arguments."
1164 url-http-end-of-headers
1165 url-http-content-type
1166 url-http-content-length
1167 url-http-transfer-encoding
1168 url-http-after-change-function
1169 url-callback-function
1170 url-callback-arguments
1171 url-http-method
1172 url-http-extra-headers
1173 url-http-data
1174 url-http-chunked-length
1175 url-http-chunked-start
1176 url-http-chunked-counter
1177 url-http-process))
1183 ;; Failed to open the connection for some reason
1194 url-http-content-type
1195 url-http-content-length
1196 url-http-transfer-encoding
1197 url-http-after-change-function
1198 url-http-response-version
1199 url-http-response-status
1200 url-http-chunked-length
1201 url-http-chunked-counter
1202 url-http-chunked-start
1203 url-callback-function
1204 url-callback-arguments
1205 url-http-process
1206 url-http-method
1207 url-http-extra-headers
1208 url-http-data
1209 url-http-target-url
1210 url-http-connection-opened
1211 url-http-proxy))
1215 url-http-extra-headers url-request-extra-headers
1216 url-http-data url-request-data
1217 url-http-process connection
1218 url-http-chunked-length nil
1219 url-http-chunked-start nil
1220 url-http-chunked-counter 0
1221 url-callback-function callback
1222 url-callback-arguments cbargs
1223 url-http-after-change-function 'url-http-wait-for-headers-change-function
1224 url-http-target-url url-current-object
1225 url-http-connection-opened nil
1226 url-http-proxy url-using-proxy)
1233 ;; Asynchronous connection
1236 ;; Asynchronous connection failed
1241 buffer))
1245 ;; We are performing an asynchronous connection, and a status change
1246 ;; has occurred.
1263 ;; Since Emacs 19/20 does not allow you to change the
1264 ;; `after-change-functions' hook in the midst of running them, we fake
1265 ;; an after change by hooking into the process filter and inserting
1266 ;; the data ourselves. This is slightly less efficient, but there
1267 ;; were tons of weird ways the after-change code was biting us in the
1268 ;; shorts.
1270 ;; Sometimes we get a zero-length data chunk after the process has
1271 ;; been changed to 'free', which means it has no buffer associated
1272 ;; with it. Do nothing if there is no buffer, or 0 length data.
1277 (url-http-debug "Calling after change function `%s' for `%S'" url-http-after-change-function proc)
1286 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
1287 ;;; file-name-handler stuff from here on out
1288 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
1291 'symbol-value-in-buffer
1293 "Return the value of SYMBOL in BUFFER, or UNBOUND-VALUE if it is unbound."
1296 unbound-value
1304 ;;;###autoload
1316 exists))
1318 ;;;###autoload
1326 nil ;dir / link / normal file
1329 nil nil nil ;atime ; mtime ; ctime
1338 ;;;###autoload
1344 ;;;###autoload
1346 "Return a property list describing options available for URL.
1347 This list is retrieved using the `OPTIONS' HTTP method.
1349 Property list members:
1351 methods
1352 A list of symbols specifying what HTTP methods the resource
1353 supports.
1355 dav
1356 A list of numbers specifying what DAV protocol/schema versions are
1357 supported.
1359 dasl
1360 A list of supported DASL search types supported (string form)
1362 ranges
1363 A list of the units available for use in partial document fetches.
1365 p3p
1366 The `Platform For Privacy Protection' description for the resource.
1367 Currently this is just the raw header contents. This is likely to
1368 change once P3P is formally supported by the URL package or
1369 Emacs/W3."
1377 ;; Only parse the options if we got a 2xx response code!
1383 ;; Figure out what methods are supported.
1386 options 'methods
1389 ;; Check for DAV
1392 options 'dav
1397 ;; Now for DASL
1400 options 'dasl
1403 ;; P3P - should get more detailed here. FIXME
1407 ;; Check for whether they accept byte-range requests.
1410 options 'ranges
1414 ))))
1416 options))
1418 ;; HTTPS. This used to be in url-https.el, but that file collides
1419 ;; with url-http.el on systems with 8-character file names.
1422 ;;;###autoload
1424 ;;;###autoload
1427 ;; FIXME what is the point of this alias being an autoload?
1428 ;; Trying to use it will not cause url-http to be loaded,
1429 ;; since the full alias just gets dumped into loaddefs.el.
1431 ;;;###autoload (autoload 'url-default-expander "url-expand")
1432 ;;;###autoload
1442 ;;;###autoload (autoload 'url-https "url-http")
1444 ;;;###autoload (autoload 'url-https-file-exists-p "url-http")
1446 ;;;###autoload (autoload 'url-https-file-readable-p "url-http")
1448 ;;;###autoload (autoload 'url-https-file-attributes "url-http")
1453 ;; arch-tag: ba7c59ae-c0f4-4a31-9617-d85f221732ee
1454 ;;; url-http.el ends here