| blob | 28071e7165a3ffabddb668a22347c7f17ec22c62 |
1 ;;; url-http.el --- HTTP retrieval routines
3 ;; Copyright (C) 1999, 2001, 2004-2011 Free Software Foundation, Inc.
5 ;; Author: Bill Perry <wmperry@gnu.org>
6 ;; Keywords: comm, data, processes
8 ;; This file is part of GNU Emacs.
9 ;;
10 ;; GNU Emacs is free software: you can redistribute it and/or modify
11 ;; it under the terms of the GNU General Public License as published by
12 ;; the Free Software Foundation, either version 3 of the License, or
13 ;; (at your option) any later version.
15 ;; GNU Emacs is distributed in the hope that it will be useful,
16 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
17 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 ;; GNU General Public License for more details.
20 ;; You should have received a copy of the GNU General Public License
21 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
23 ;;; Commentary:
25 ;;; Code:
53 "What version of HTTP we advertise, as a string.
54 Valid values are 1.1 and 1.0.
55 This is only useful when debugging the HTTP subsystem.
57 Setting this to 1.0 will tell servers not to send chunked encoding,
61 "Whether to use a single TCP connection multiple times in HTTP.
62 This is only useful when debugging the HTTP subsystem. Setting to
63 nil will explicitly close the connection to the server after every
115 ;(eval-when-compile
116 ;; These are all macros so that they are hidden from external sight
117 ;; when the file is byte-compiled.
118 ;;
119 ;; This allows us to expose just the entry points we want.
121 ;; These routines will allow us to implement persistent HTTP
122 ;; connections.
126 ;; The user hit C-g, honor it! Some things can get in an
127 ;; incredibly tight loop (chunked encoding)
140 url-http-open-connections)
141 proc)
151 url-http-open-connections))
152 nil)
171 host port
174 ;; `url-open-stream' needs a buffer in which to do things
175 ;; like authentication. But we use another buffer afterwards.
178 ;; url-open-stream might return nil.
180 ;; Drop the temp buffer link before killing the buffer.
182 proc)
185 ;; Building an HTTP request
190 ""
195 url-version
204 "Create an HTTP request for `url-http-target-url', referred to by REF-URL."
206 url-http-method url-http-data
207 url-http-extra-headers))
213 url-http-extra-headers))
215 nil
223 nil
226 proxy-info)
236 ;; Protection against stupid values in the referer
241 ;; We do not want to expose the referer if the user is paranoid.
247 ;; url-http-extra-headers contains an assoc-list of
248 ;; header/value pairs that we need to put into the request.
256 ;; This was done with a call to `format'. Concatting parts has
257 ;; the advantage of keeping the parts of each header together and
258 ;; allows us to elide null lines directly, at the cost of making
259 ;; the layout less clear.
261 ;; We used to concat directly, but if one of the strings happens
262 ;; to being multibyte (even if it only contains pure ASCII) then
263 ;; every string gets converted with `string-MAKE-multibyte' which
264 ;; turns the 127-255 codes into things like latin-1 accented chars
265 ;; (it would work right if it used `string-TO-multibyte' instead).
266 ;; So to avoid the problem we force every string to be unibyte.
268 ;; FIXME: Instead of `string-AS-unibyte' we'd want
269 ;; `string-to-unibyte', so as to properly signal an error if one
270 ;; of the strings contains a multibyte char.
271 'string-as-unibyte
274 ;; The request
278 ;; Version of MIME we speak
280 ;; (maybe) Try to keep the connection open
284 ;; HTTP extensions we support
288 ;; Who we want to talk to
295 ;; Who its from
299 ;; Encodings we understand
306 ;; Languages we understand
310 ;; Types we understand
312 ;; User agent
314 ;; Proxy Authorization
315 proxy-auth
316 ;; Authorization
317 auth
318 ;; Cookies
321 ;; If-modified-since
328 ;; Whence we came
331 extra-headers
332 ;; Length of data
338 ;; End request
340 ;; Any data
344 request))
346 ;; Parsing routines
349 This allows us to use `mail-fetch-field', etc."
357 url-callback-function url-callback-arguments))
362 nil nil t))
368 url-http-proxy
369 url-current-object)))
371 ;; Cheating, but who cares? :)
372 'url-http-proxy-basic-auth-storage
374 auth
377 ;; find strongest supported auth
381 this-auth)))
390 type this-type
391 strength this-strength))))
398 " authentication. If you'd like to write it,"
404 type t args)))
408 url-http-extra-headers)
413 url-callback-arguments)))))))
416 "Parse just the response code."
418 url-http-response-version))
433 "Handle all set-cookie / set-cookie2 headers in an HTTP response.
434 The buffer must already be narrowed to the headers, so `mail-fetch-field' will
435 work correctly."
442 ;;; (while cookies2
443 ;;; (url-cookie-handle-set-cookie2 (pop cookies)))
444 )
445 )
448 "Parse and handle HTTP specific headers.
449 Return t if and only if the current buffer is still active and
450 should be shown to the user."
451 ;; The comments after each status code handled are taken from RFC
452 ;; 2616 (HTTP/1.1)
454 url-http-response-version
455 url-http-method url-http-data url-http-process
456 url-callback-function url-callback-arguments))
460 url-http-process)
469 ;;(narrow-to-region (point-min) url-http-end-of-headers)
471 ;; In HTTP 1.0, keep the connection only if there is a
472 ;; "Connection: keep-alive" header.
473 ;; In HTTP 1.1 (and greater), keep the connection unless there is a
474 ;; "Connection: close" header
487 ;; other status symbols: jewelry and luxury cars
489 ;; The filename part of a URL could be in remote file syntax,
490 ;; see Bug#6717 for an example. We disable file name
491 ;; handlers, therefore.
498 ;; Classes of response codes
499 ;;
500 ;; 5xx = Server Error
501 ;; 4xx = Client Error
502 ;; 3xx = Redirection
503 ;; 2xx = Successful
504 ;; 1xx = Informational
506 ;; 100 = Continue with request
507 ;; 101 = Switching protocols
508 ;; 102 = Processing (Added by DAV)
512 ;; 200 Ok
513 ;; 201 Created
514 ;; 202 Accepted
515 ;; 203 Non-authoritative information
516 ;; 204 No content
517 ;; 205 Reset content
518 ;; 206 Partial content
519 ;; 207 Multi-status (Added by DAV)
522 ;; No new data, just stay at the same document
526 ;; Generic success for all others. Store in the cache, and
527 ;; mark it as successful.
533 ;; 300 Multiple choices
534 ;; 301 Moved permanently
535 ;; 302 Found
536 ;; 303 See other
537 ;; 304 Not modified
538 ;; 305 Use proxy
539 ;; 307 Temporary redirect
544 ;; Quoth the spec (section 10.3.1)
545 ;; -------------------------------
546 ;; The requested resource corresponds to any one of a set of
547 ;; representations, each with its own specific location and
548 ;; agent-driven negotiation information is being provided so
549 ;; that the user can select a preferred representation and
550 ;; redirect its request to that location.
551 ;; [...]
552 ;; If the server has a preferred choice of representation, it
553 ;; SHOULD include the specific URI for that representation in
554 ;; the Location field; user agents MAY use the Location field
555 ;; value for automatic redirection.
556 ;; -------------------------------
557 ;; We do not support agent-driven negotiation, so we just
558 ;; redirect to the preferred URI if one is provided.
559 nil)
561 ;; If the 301|302 status code is received in response to a
562 ;; request other than GET or HEAD, the user agent MUST NOT
563 ;; automatically redirect the request unless it can be
564 ;; confirmed by the user, since this might change the
565 ;; conditions under which the request was issued.
567 ;; Automatic redirection is ok
568 nil
569 ;; It is just too big of a pain in the ass to get this
570 ;; prompt all the time. We will just silently lose our
571 ;; data and convert to a GET method.
573 url-http-method url-http-response-status)
575 url-http-data nil)))
577 ;; The response to the request can be found under a different
578 ;; URI and SHOULD be retrieved using a GET method on that
579 ;; resource.
581 url-http-data nil))
583 ;; The 304 response MUST NOT contain a message-body.
588 success t))
590 ;; The requested resource MUST be accessed through the
591 ;; proxy given by the Location field. The Location field
592 ;; gives the URI of the proxy. The recipient is expected
593 ;; to repeat this single request via the proxy. 305
594 ;; responses MUST only be generated by origin servers.
596 redirect-uri))
598 ;; Treat everything like '300'
599 nil))
601 ;; Clean off any whitespace and/or <...> cruft.
607 ;; Some stupid sites (like sourceforge) send a
608 ;; non-fully-qualified URL (ie: /), which royally confuses
609 ;; the URL library.
611 ;; Be careful to use the real target URL, otherwise we may
612 ;; compute the redirection relative to the URL of the proxy.
618 ;; Check existing number of redirects
629 ;; url-max-redirections hasn't been reached, so go
630 ;; ahead and redirect.
632 ;; Remember that the request was redirected.
636 ;; Put in the current buffer a forwarding pointer to the new
637 ;; destination buffer.
638 ;; FIXME: This is a hack to fix url-retrieve-synchronously
639 ;; without changing the API. Instead url-retrieve should
640 ;; either simply not return the "destination" buffer, or it
641 ;; should take an optional `dest-buf' argument.
644 redirect-uri url-callback-function
645 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.
880 ;; HTTP/0.9 just gets passed back no matter what
886 ;; Function used when we do NOT know how long the document is going to be
887 ;; Just _very_ simple 'downloaded %d' type of info.
892 "Function used when we DO know how long the document is going to be.
893 More sophisticated percentage downloaded, etc.
894 Also does minimal parsing of HTTP headers and will actually cause
895 the callback to be triggered."
897 url-http-end-of-headers
898 url-http-content-length
899 url-http-content-type
900 url-http-process))
903 "Reading [%s]... %s of %s (%d%%)"
905 url-http-content-length)
906 url-http-content-type
910 url-http-content-length))
912 "Reading... %s of %s (%d%%)"
914 url-http-content-length)
918 url-http-content-length)))
922 ;; Found the end of the document! Wheee!
929 "Function used when dealing with 'chunked' encoding.
930 Cannot give a sophisticated percentage, but we need a different
931 function to look for the special 0-length chunk that signifies
932 the end of the document."
934 url-http-end-of-headers
935 url-http-content-type
936 url-http-chunked-length
937 url-http-chunked-counter
938 url-http-process url-http-chunked-start))
945 ;; We need to loop thru looking for more chunks even within
946 ;; one after-change-function call.
951 "Reading [%s]... chunk #%d"
952 url-http-content-type url-http-chunked-counter)
954 "Reading... chunk #%d"
955 url-http-chunked-counter))
957 url-http-chunked-counter st nd length)
963 ;; We know how long the chunk is supposed to be, skip over
964 ;; leading crap if possible.
968 url-http-chunked-counter)
970 url-http-chunked-length)))
973 url-http-chunked-length)
974 nd))
981 ;; Must not have received the entirety of the chunk header,
982 ;; need to spin some more.
987 'end-open t
988 'chunked-encoding t
1000 ; (if (not url-http-debug)
1003 url-http-chunked-counter url-http-chunked-length
1007 ;; Found the end of the document! Wheee!
1011 ;; Every chunk, even the last 0-length one, is
1012 ;; terminated by CRLF. Skip it.
1022 ;; This will wait for the headers to arrive and then splice in the
1023 ;; next appropriate after-change-function, etc.
1025 url-http-end-of-headers
1026 url-http-content-type
1027 url-http-content-length
1028 url-http-transfer-encoding
1029 url-callback-function
1030 url-callback-arguments
1031 url-http-process
1032 url-http-method
1033 url-http-after-change-function
1034 url-http-response-status))
1045 ;; Not HTTP/x.y data, must be 0.9
1046 ;; God, I wish this could die.
1048 url-http-end-of-headers 0
1049 old-http t)
1051 ;; Saw the end of the headers
1055 end-of-headers t)
1059 ;; Haven't seen the end of the headers yet, need to wait
1060 ;; for more data to arrive.
1061 nil
1067 ;;(narrow-to-region (point-min) url-http-end-of-headers)
1081 ;; We got back a headerless malformed response from the
1082 ;; server.
1091 ;; A HEAD request is _ALWAYS_ terminated by the header
1092 ;; information, regardless of any entity headers,
1093 ;; according to section 4.4 of the HTTP/1.1 draft.
1099 ;; A CONNECT request is finished, but we cannot stick this
1100 ;; back on the free connectin list
1105 ;; Only allowed to have a header section. We have to handle
1106 ;; this here instead of in url-http-parse-headers because if
1107 ;; you have a cached copy of something without a known
1108 ;; content-length, and try to retrieve it from the cache, we'd
1109 ;; fall into the 'being dumb' section and wait for the
1110 ;; connection to terminate, which means we'd wait for 10
1111 ;; seconds for the keep-alives to time out on some servers.
1115 ;; HTTP/0.9 always signaled end-of-connection by closing the
1116 ;; connection.
1138 ;; We got a NULL body! Activate the callback
1139 ;; immediately!
1145 ;; Have some leftover data
1149 nd
1152 nil)))
1157 ;; We are still at the beginning of the buffer... must just be
1158 ;; waiting for a response.
1163 ;;;###autoload
1165 "Retrieve URL via HTTP asynchronously.
1166 URL must be a parsed URL. See `url-generic-parse-url' for details.
1167 When retrieval is completed, the function CALLBACK is executed with
1168 CBARGS as the arguments."
1171 url-http-end-of-headers
1172 url-http-content-type
1173 url-http-content-length
1174 url-http-transfer-encoding
1175 url-http-after-change-function
1176 url-callback-function
1177 url-callback-arguments
1178 url-http-method
1179 url-http-extra-headers
1180 url-http-data
1181 url-http-chunked-length
1182 url-http-chunked-start
1183 url-http-chunked-counter
1184 url-http-process))
1190 ;; Failed to open the connection for some reason
1201 url-http-content-type
1202 url-http-content-length
1203 url-http-transfer-encoding
1204 url-http-after-change-function
1205 url-http-response-version
1206 url-http-response-status
1207 url-http-chunked-length
1208 url-http-chunked-counter
1209 url-http-chunked-start
1210 url-callback-function
1211 url-callback-arguments
1212 url-http-process
1213 url-http-method
1214 url-http-extra-headers
1215 url-http-data
1216 url-http-target-url
1217 url-http-connection-opened
1218 url-http-proxy))
1222 url-http-extra-headers url-request-extra-headers
1223 url-http-data url-request-data
1224 url-http-process connection
1225 url-http-chunked-length nil
1226 url-http-chunked-start nil
1227 url-http-chunked-counter 0
1228 url-callback-function callback
1229 url-callback-arguments cbargs
1230 url-http-after-change-function 'url-http-wait-for-headers-change-function
1231 url-http-target-url url-current-object
1232 url-http-connection-opened nil
1233 url-http-proxy url-using-proxy)
1240 ;; Asynchronous connection
1243 ;; Asynchronous connection failed
1248 buffer))
1252 ;; We are performing an asynchronous connection, and a status change
1253 ;; has occurred.
1270 ;; Since Emacs 19/20 does not allow you to change the
1271 ;; `after-change-functions' hook in the midst of running them, we fake
1272 ;; an after change by hooking into the process filter and inserting
1273 ;; the data ourselves. This is slightly less efficient, but there
1274 ;; were tons of weird ways the after-change code was biting us in the
1275 ;; shorts.
1276 ;; FIXME this can probably be simplified since the above is no longer true.
1278 ;; Sometimes we get a zero-length data chunk after the process has
1279 ;; been changed to 'free', which means it has no buffer associated
1280 ;; with it. Do nothing if there is no buffer, or 0 length data.
1285 (url-http-debug "Calling after change function `%s' for `%S'" url-http-after-change-function proc)
1294 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
1295 ;;; file-name-handler stuff from here on out
1296 ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
1299 'symbol-value-in-buffer
1301 "Return the value of SYMBOL in BUFFER, or UNBOUND-VALUE if it is unbound."
1304 unbound-value
1312 ;;;###autoload
1324 exists))
1326 ;;;###autoload
1334 nil ;dir / link / normal file
1337 nil nil nil ;atime ; mtime ; ctime
1346 ;;;###autoload
1352 ;;;###autoload
1354 "Return a property list describing options available for URL.
1355 This list is retrieved using the `OPTIONS' HTTP method.
1357 Property list members:
1359 methods
1360 A list of symbols specifying what HTTP methods the resource
1361 supports.
1363 dav
1364 A list of numbers specifying what DAV protocol/schema versions are
1365 supported.
1367 dasl
1368 A list of supported DASL search types supported (string form)
1370 ranges
1371 A list of the units available for use in partial document fetches.
1373 p3p
1374 The `Platform For Privacy Protection' description for the resource.
1375 Currently this is just the raw header contents. This is likely to
1376 change once P3P is formally supported by the URL package or
1377 Emacs/W3."
1385 ;; Only parse the options if we got a 2xx response code!
1391 ;; Figure out what methods are supported.
1394 options 'methods
1397 ;; Check for DAV
1400 options 'dav
1405 ;; Now for DASL
1408 options 'dasl
1411 ;; P3P - should get more detailed here. FIXME
1415 ;; Check for whether they accept byte-range requests.
1418 options 'ranges
1422 ))))
1424 options))
1426 ;; HTTPS. This used to be in url-https.el, but that file collides
1427 ;; with url-http.el on systems with 8-character file names.
1430 ;;;###autoload
1432 ;;;###autoload
1435 ;; FIXME what is the point of this alias being an autoload?
1436 ;; Trying to use it will not cause url-http to be loaded,
1437 ;; since the full alias just gets dumped into loaddefs.el.
1439 ;;;###autoload (autoload 'url-default-expander "url-expand")
1440 ;;;###autoload
1450 ;;;###autoload (autoload 'url-https "url-http")
1452 ;;;###autoload (autoload 'url-https-file-exists-p "url-http")
1454 ;;;###autoload (autoload 'url-https-file-readable-p "url-http")
1456 ;;;###autoload (autoload 'url-https-file-attributes "url-http")
1461 ;;; url-http.el ends here