Context Navigation

Snoopy.class.inc

Revision 1477, 30.3 kB (checked in by sdcosta, 2 years ago)
Fix for #409 The website was sending the gzip header as "Content-encoding: gzip" instead of "Content-Encoding: gzip". If firefox fixes it, then so should we.
Property svn:eol-style set to `native` Property svn:keywords set to `Author Date Id Revision`

Line
1	<?php
2
3	/*************************************************
4
5	Snoopy - the PHP net client
6	Author: Monte Ohrt <monte@ispi.net>
7	Copyright (c): 1999-2000 ispi, all rights reserved
8	Version: 1.0 (plus - see SJM comments below)
9
10	* This library is free software; you can redistribute it and/or
11	* modify it under the terms of the GNU Lesser General Public
12	* License as published by the Free Software Foundation; either
13	* version 2.1 of the License, or (at your option) any later version.
14	*
15	* This library 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 GNU
18	* Lesser General Public License for more details.
19	*
20	* You should have received a copy of the GNU Lesser General Public
21	* License along with this library; if not, write to the Free Software
22	* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
23
24	You may contact the author of Snoopy by e-mail at:
25	monte@ispi.net
26
27	Or, write to:
28	Monte Ohrt
29	CTO, ispi
30	237 S. 70th suite 220
31	Lincoln, NE 68510
32
33	The latest version of Snoopy can be obtained from:
34	http://snoopy.sourceforge.com
35
36
37
38	SJM - alpha-grade changes based on the version of Snoopy released with MagpieRSS 0.7
39
40	comments to steve@minutillo.com
41
42	Two additions:
43
44	1) If this is PHP 4.3 or greater, and 'openssl' is available,
45	use the PHP built in SSL support for "https" instead of calling curl externally.
46	Use of external curl can still be forced by setting $use_curl = true.
47
48	ref: http://us2.php.net/fsockopen
49
50	2) HTTP Digest Authentication. If you set a username and password, basic auth
51	will be tried first. If that fails, and the server sends back an
52	WWW-Authenticate: Digest header, the request will be retried with the appropriate
53	digest response. Only qop=auth is supported, with MD5 as the algorithm.
54	I realize that sending basic auth first, and then following up with a digest
55	challenge-response kind of defeats the purpose in terms of security.
56
57	ref: http://www.faqs.org/rfcs/rfc2617.html
58
59	*************************************************/
60
61	class Snoopy
62	{
63	/** Public variables **/
64
65	/* user definable vars */
66
67	var $host = "www.php.net"; // host name we are connecting to
68	var $port = 80; // port we are connecting to
69	var $proxy_host = ""; // proxy host to use
70	var $proxy_port = ""; // proxy port to use
71	var $agent = "Snoopy v1.0"; // agent we masquerade as
72	var $referer = ""; // referer info to pass
73	var $cookies = array(); // array of cookies to pass
74	// $cookies["username"]="joe";
75	var $rawheaders = array(); // array of raw headers to send
76	// $rawheaders["Content-type"]="text/html";
77
78	var $maxredirs = 5; // http redirection depth maximum. 0 = disallow
79	var $lastredirectaddr = ""; // contains address of last redirected address
80	var $offsiteok = true; // allows redirection off-site
81	var $maxframes = 0; // frame content depth maximum. 0 = disallow
82	var $expandlinks = true; // expand links to fully qualified URLs.
83	// this only applies to fetchlinks()
84	// or submitlinks()
85	var $passcookies = true; // pass set cookies back through redirects
86	// NOTE: this currently does not respect
87	// dates, domains or paths.
88
89	var $user = ""; // user for http authentication
90	var $pass = ""; // password for http authentication
91
92	// http accept types
93	var $accept = "image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, /";
94
95	var $results = ""; // where the content is put
96
97	var $error = ""; // error messages sent here
98	var $response_code = ""; // response code returned from server
99	var $headers = array(); // headers returned from server sent here
100	var $maxlength = 500000; // max return data length (body)
101	var $read_timeout = 0; // timeout on read operations, in seconds
102	// supported only since PHP 4 Beta 4
103	// set to 0 to disallow timeouts
104	var $timed_out = false; // if a read operation timed out
105	var $status = 0; // http request status
106
107	var $curl_path = "/usr/bin/curl";
108	// Snoopy will use cURL for fetching
109	// SSL content if a full system path to
110	// the cURL binary is supplied here.
111	// set to false if you do not have
112	// cURL installed. See http://curl.haxx.se
113	// for details on installing cURL.
114	// Snoopy does not use the cURL
115	// library functions built into php,
116	// as these functions are not stable
117	// as of this Snoopy release.
118
119	// SJM - always use curl for HTTPS requests?
120	var $use_curl = false;
121
122
123	// send Accept-encoding: gzip?
124	var $use_gzip = true;
125
126	/** Private variables **/
127
128	var $_maxlinelen = 4096; // max line length (headers)
129
130	var $_scheme = "http"; // default scheme
131	var $_httpmethod = "GET"; // default http request method
132	var $_httpversion = "HTTP/1.0"; // default http request version
133	var $_submit_method = "POST"; // default submit method
134	var $_submit_type = "application/x-www-form-urlencoded"; // default submit type
135	var $_mime_boundary = ""; // MIME boundary for multipart/form-data submit type
136	var $_redirectaddr = false; // will be set if page fetched is a redirect
137	var $_redirectdepth = 0; // increments on an http redirect
138	var $_trieddigest = false; // have we tried Digest auth yet?
139	var $_frameurls = array(); // frame src urls
140	var $_framedepth = 0; // increments on frame depth
141
142	var $_isproxy = false; // set if using a proxy server
143	var $_fp_timeout = 30; // timeout for socket connection
144	var $tmpdir = "/tmp/";
145
146	function Snoopy() {
147	if (defined('CURL_PATH')) {
148	$this -> curl_path = CURL_PATH;
149	$this -> use_curl = true;
150	}
151	/*
152	if (function_exists('getConfig') && ($tmp = getConfig('rss.output.cachedir')) != null) {
153	if (substr($tmp, -1, 1) != "/") {
154	$tmp .= "/";
155	}
156	$this -> tmpdir = $tmp;
157	}
158	*/
159	}
160	/======================================================================\
161	Function: fetch
162	Purpose: fetch the contents of a web page
163	(and possibly other protocols in the
164	future like ftp, nntp, gopher, etc.)
165	Input: $URI the location of the page to fetch
166	Output: $this->results the output text from the fetch
167	\======================================================================/
168
169	function fetch($URI)
170	{
171
172	//preg_match("\|^([^:]+)://([^:/]+)(:[\d]+)(.)\|",$URI,$URI_PARTS);
173	$URI_PARTS = parse_url($URI);
174	if (!empty($URI_PARTS["user"]))
175	$this->user = $URI_PARTS["user"];
176	if (!empty($URI_PARTS["pass"]))
177	$this->pass = $URI_PARTS["pass"];
178
179	$this->_scheme = $URI_PARTS["scheme"];
180
181	switch($URI_PARTS["scheme"])
182	{
183	case "http":
184	case "https":
185	break;
186
187	default:
188	// not a valid protocol
189	$this->error = 'Invalid protocol "'.$URI_PARTS["scheme"].'"\n';
190	return false;
191	}
192
193	if($URI_PARTS["scheme"] == "https")
194	{
195
196	// SJM - if they really want curl, or it isn't PHP 4.3 yet, or openssl extension isn't loaded
197	if($use_curl \|\| !function_exists('file_get_contents') \|\| !extension_loaded('openssl'))
198	{
199
200	// MBI: is_executable would break on some installs
201	if(!$this->curl_path \|\| (!file_exists($this->curl_path))) {
202	$this->error = "Bad curl ($this->curl_path), can't fetch HTTPS \n";
203	return false;
204	}
205
206	$this->host = $URI_PARTS["host"];
207	if(!empty($URI_PARTS["port"]))
208	$this->port = $URI_PARTS["port"];
209	if($this->_isproxy)
210	{
211	// using proxy, send entire URI
212	$this->_curlrequest($URI,$URI,$this->_httpmethod);
213	}
214	else
215	{
216	$path = $URI_PARTS["path"].($URI_PARTS["query"] ? "?".$URI_PARTS["query"] : "");
217	// no proxy, send only the path
218	$this->_curlrequest($path, $URI, $this->_httpmethod);
219	}
220
221	if($this->_redirectaddr)
222	{
223	/* url was redirected, check if we've hit the max depth */
224	if($this->maxredirs > $this->_redirectdepth)
225	{
226	// only follow redirect if it's on this site, or offsiteok is true
227	if(preg_match("\|^http://".preg_quote($this->host)."\|i",$this->_redirectaddr) \|\| $this->offsiteok)
228	{
229	/* follow the redirect */
230	$this->_redirectdepth++;
231	$this->lastredirectaddr=$this->_redirectaddr;
232	$this->fetch($this->_redirectaddr);
233	}
234	}
235	}
236
237	if($this->_framedepth < $this->maxframes && count($this->_frameurls) > 0)
238	{
239	$frameurls = $this->_frameurls;
240	$this->_frameurls = array();
241
242	while(list(,$frameurl) = each($frameurls))
243	{
244	if($this->_framedepth < $this->maxframes)
245	{
246	$this->fetch($frameurl);
247	$this->_framedepth++;
248	}
249	else
250	break;
251	}
252	}
253	return true;
254	}
255	}
256
257	// SJM - else drop through and treat https as http
258
259	$this->host = $URI_PARTS["host"];
260	if(!empty($URI_PARTS["port"]))
261	$this->port = $URI_PARTS["port"];
262
263	// SJM - if it's https, default the port to 443
264	if($URI_PARTS["scheme"] == "https")
265	{
266	if(empty($URI_PARTS["port"]))
267	{
268	$this->port = 443;
269	}
270	}
271
272	if($this->_connect($fp))
273	{
274	if($this->_isproxy)
275	{
276	// using proxy, send entire URI
277	$this->_httprequest($URI,$fp,$URI,$this->_httpmethod);
278	}
279	else
280	{
281	$path = $URI_PARTS["path"].(isset($URI_PARTS["query"]) ? "?".$URI_PARTS["query"] : "");
282	// no proxy, send only the path
283	$this->_httprequest($path, $fp, $URI, $this->_httpmethod);
284	}
285
286	$this->_disconnect($fp);
287
288	if($this->_redirectaddr)
289	{
290	/* url was redirected, check if we've hit the max depth */
291	if($this->maxredirs > $this->_redirectdepth)
292	{
293	// only follow redirect if it's on this site, or offsiteok is true
294	if(preg_match("\|^http://".preg_quote($this->host)."\|i",$this->_redirectaddr) \|\| $this->offsiteok)
295	{
296	/* follow the redirect */
297	$this->_redirectdepth++;
298	$this->lastredirectaddr=$this->_redirectaddr;
299	$this->fetch($this->_redirectaddr);
300	}
301	}
302	}
303
304	if($this->_framedepth < $this->maxframes && count($this->_frameurls) > 0)
305	{
306	$frameurls = $this->_frameurls;
307	$this->_frameurls = array();
308
309	while(list(,$frameurl) = each($frameurls))
310	{
311	if($this->_framedepth < $this->maxframes)
312	{
313	$this->fetch($frameurl);
314	$this->_framedepth++;
315	}
316	else
317	break;
318	}
319	}
320	}
321	else
322	{
323	return false;
324	}
325	return true;
326	}
327
328
329
330	/======================================================================\
331	Private functions
332	\======================================================================/
333
334
335	/======================================================================\
336	Function: _striplinks
337	Purpose: strip the hyperlinks from an html document
338	Input: $document document to strip.
339	Output: $match an array of the links
340	\======================================================================/
341
342	function _striplinks($document)
343	{
344	preg_match_all("'<\sa\s+.href\s=\s # find <a href=
345	([\"\'])? # find single or double quote
346	(?(1) (.*?)\\1 \| ([^\s\>]+)) # if quote found, match up to next matching
347	# quote, otherwise match up to next space
348	'isx",$document,$links);
349
350
351	// catenate the non-empty matches from the conditional subpattern
352
353	while(list($key,$val) = each($links[2]))
354	{
355	if(!empty($val))
356	$match[] = $val;
357	}
358
359	while(list($key,$val) = each($links[3]))
360	{
361	if(!empty($val))
362	$match[] = $val;
363	}
364
365	// return the links
366	return $match;
367	}
368
369	/======================================================================\
370	Function: _stripform
371	Purpose: strip the form elements from an html document
372	Input: $document document to strip.
373	Output: $match an array of the links
374	\======================================================================/
375
376	function _stripform($document)
377	{
378	preg_match_all("'<\/?(FORM\|INPUT\|SELECT\|TEXTAREA\|(OPTION))[^<>]>(?(2)(.(?=<\/?(option\|select)[^<>]>[\r\n])\|(?=[\r\n]))\|(?=[\r\n]))'Usi",$document,$elements);
379
380	// catenate the matches
381	$match = implode("\r\n",$elements[0]);
382
383	// return the links
384	return $match;
385	}
386
387
388
389	/======================================================================\
390	Function: _striptext
391	Purpose: strip the text from an html document
392	Input: $document document to strip.
393	Output: $text the resulting text
394	\======================================================================/
395
396	function _striptext($document)
397	{
398
399	// I didn't use preg eval (//e) since that is only available in PHP 4.0.
400	// so, list your entities one by one here. I included some of the
401	// more common ones.
402
403	$search = array("'<script[^>]?>.?</script>'si", // strip out javascript
404	"'<[\/\!]?[^<>]?>'si", // strip out html tags
405	"'([\r\n])[\s]+'", // strip out white space
406	"'&(quote\|#34);'i", // replace html entities
407	"'&(amp\|#38);'i",
408	"'&(lt\|#60);'i",
409	"'&(gt\|#62);'i",
410	"'&(nbsp\|#160);'i",
411	"'&(iexcl\|#161);'i",
412	"'&(cent\|#162);'i",
413	"'&(pound\|#163);'i",
414	"'&(copy\|#169);'i"
415	);
416	$replace = array( "",
417	"",
418	"\\1",
419	"\"",
420	"&",
421	"<",
422	">",
423	" ",
424	chr(161),
425	chr(162),
426	chr(163),
427	chr(169));
428
429	$text = preg_replace($search,$replace,$document);
430
431	return $text;
432	}
433
434	/======================================================================\
435	Function: _expandlinks
436	Purpose: expand each link into a fully qualified URL
437	Input: $links the links to qualify
438	$URI the full URI to get the base from
439	Output: $expandedLinks the expanded links
440	\======================================================================/
441
442	function _expandlinks($links,$URI)
443	{
444
445	preg_match("/^[^\?]+/",$URI,$match);
446
447	$match = preg_replace("\|/[^\/\.]+\.[^\/\.]+$\|","",$match[0]);
448
449	$search = array( "\|^http://".preg_quote($this->host)."\|i",
450	"\|^(?!http://)(\/)?(?!mailto:)\|i",
451	"\|/\./\|",
452	"\|/[^\/]+/\.\./\|"
453	);
454
455	$replace = array( "",
456	$match."/",
457	"/",
458	"/"
459	);
460
461	$expandedLinks = preg_replace($search,$replace,$links);
462
463	return $expandedLinks;
464	}
465
466	/======================================================================\
467	Function: _httprequest
468	Purpose: go get the http data from the server
469	Input: $url the url to fetch
470	$fp the current open file pointer
471	$URI the full URI
472	$body body contents to send if any (POST)
473	Output:
474	\======================================================================/
475
476	function _httprequest($url,$fp,$URI,$http_method,$content_type="",$body="")
477	{
478	if($this->passcookies && $this->_redirectaddr)
479	$this->setcookies();
480
481	$URI_PARTS = parse_url($URI);
482	if(empty($url))
483	$url = "/";
484	$headers = $http_method." ".$url." ".$this->_httpversion."\r\n";
485	if(!empty($this->agent))
486	$headers .= "User-Agent: ".$this->agent."\r\n";
487	if(!empty($this->host) && !isset($this->rawheaders['Host']))
488	$headers .= "Host: ".$this->host."\r\n";
489	if(!empty($this->accept))
490	$headers .= "Accept: ".$this->accept."\r\n";
491
492	if($this->use_gzip) {
493	// make sure PHP was built with --with-zlib
494	// and we can handle gzipp'ed data
495	if ( function_exists(gzinflate) ) {
496	$headers .= "Accept-encoding: gzip\r\n";
497	}
498	else {
499	trigger_error(
500	"use_gzip is on, but PHP was built without zlib support.".
501	" Requesting file(s) without gzip encoding.",
502	E_USER_NOTICE);
503	}
504	}
505
506	if(!empty($this->referer))
507	$headers .= "Referer: ".$this->referer."\r\n";
508	if(!empty($this->cookies))
509	{
510	if(!is_array($this->cookies))
511	$this->cookies = (array)$this->cookies;
512
513	reset($this->cookies);
514	if ( count($this->cookies) > 0 ) {
515	$cookie_headers .= 'Cookie: ';
516	foreach ( $this->cookies as $cookieKey => $cookieVal ) {
517	$cookie_headers .= $cookieKey."=".urlencode($cookieVal)."; ";
518	}
519	$headers .= substr($cookie_headers,0,-2) . "\r\n";
520	}
521	}
522	if(!empty($this->rawheaders))
523	{
524	if(!is_array($this->rawheaders))
525	$this->rawheaders = (array)$this->rawheaders;
526	while(list($headerKey,$headerVal) = each($this->rawheaders))
527	$headers .= $headerKey.": ".$headerVal."\r\n";
528	}
529	if(!empty($content_type)) {
530	$headers .= "Content-type: $content_type";
531	if ($content_type == "multipart/form-data")
532	$headers .= "; boundary=".$this->_mime_boundary;
533	$headers .= "\r\n";
534	}
535	if(!empty($body))
536	$headers .= "Content-length: ".strlen($body)."\r\n";
537	if(!empty($this->user) \|\| !empty($this->pass))
538	$headers .= "Authorization: BASIC ".base64_encode(urldecode($this->user).":".urldecode($this->pass))."\r\n";
539
540	$headers .= "\r\n";
541
542	// set the read timeout if needed
543	if ($this->read_timeout > 0)
544	socket_set_timeout($fp, $this->read_timeout);
545	$this->timed_out = false;
546
547	fwrite($fp,$headers.$body,strlen($headers.$body));
548
549	$this->_redirectaddr = false;
550	unset($this->headers);
551
552	// content was returned gzip encoded?
553	$is_gzipped = false;
554
555	while($currentHeader = fgets($fp,$this->_maxlinelen))
556	{
557	if ($this->read_timeout > 0 && $this->_check_timeout($fp))
558	{
559	$this->status=-100;
560	return false;
561	}
562
563	// if($currentHeader == "\r\n")
564	if(preg_match("/^\r?\n$/", $currentHeader) )
565	break;
566
567	if(!$this->_tried_digest && preg_match("/^WWW-Authenticate: Digest (.*)/", $currentHeader, $matches))
568	{
569	// SJM - we got a Digest challenge. Try to respond...
570
571	$digestheader = $matches[1];
572
573	preg_match("/nonce=\"(.*?)\"/", $digestheader, $matches);
574	$nonce = $matches[1];
575
576	preg_match("/realm=\"(.*?)\"/", $digestheader, $matches);
577	$realm = $matches[1];
578
579	$cnonce = md5(microtime());
580
581	$a1 = $this->user . ":" . $realm . ":" . $this->pass;
582	$a2 = $http_method . ":" . $url;
583
584	$ha1 = md5($a1);
585	$ha2 = md5($a2);
586
587	$response = md5($ha1 . ":" . $nonce . ":00000001:" . $cnonce . ":auth:" . $ha2);
588
589	$auth = 'Digest username="' . $this->user . '", ';
590	$auth .= 'realm="' . $realm . '", ';
591	$auth .= 'nonce="' . $nonce . '", ';
592	$auth .= 'uri="' . $url . '", ';
593	$auth .= 'response="' . $response . '", ';
594	$auth .= 'algorithm="MD5", ';
595	$auth .= 'cnonce="' . $cnonce . '", ';
596	$auth .= 'nc=00000001, ';
597	$auth .= 'qop="auth"';
598
599	// SJM - treat Digest challenge as a redirect. set flag so we don't keep retrying.
600
601	$this->_tried_digest = true;
602
603	$this->rawheaders["Authorization"]=$auth;
604	$this->user = "";
605	$this->pass = "";
606
607	$this->_redirectaddr = $URI_PARTS['scheme'] . '://' . $this->host . $url;
608	}
609
610	// if a header begins with Location: or URI:, set the redirect
611	if(preg_match("/^(Location:\|URI:)/i",$currentHeader))
612	{
613	// get URL portion of the redirect
614	preg_match("/^(Location:\|URI:)\s+(.*)/i",chop($currentHeader),$matches);
615	// look for :// in the Location header to see if hostname is included
616	if(!preg_match("\|\:\/\/\|",$matches[2]))
617	{
618	// no host in the path, so prepend
619	$this->_redirectaddr = $URI_PARTS["scheme"]."://".$this->host.":".$this->port;
620	// eliminate double slash
621	if(!preg_match("\|^/\|",$matches[2]))
622	$this->_redirectaddr .= "/".$matches[2];
623	else
624	$this->_redirectaddr .= $matches[2];
625	}
626	else
627	$this->_redirectaddr = $matches[2];
628	}
629
630	if(preg_match("\|^HTTP/\|",$currentHeader))
631	{
632	if(preg_match("\|^HTTP/[^\s]\s(.?)\s\|",$currentHeader, $status))
633	{
634	$this->status= $status[1];
635	}
636	$this->response_code = $currentHeader;
637	}
638
639	if (preg_match("/Content-Encoding: gzip/i", $currentHeader) ) {
640	$is_gzipped = true;
641	}
642
643	$this->headers[] = $currentHeader;
644	}
645
646	# $results = fread($fp, $this->maxlength);
647	$results = "";
648	while ( $data = fread($fp, $this->maxlength) ) {
649	$results .= $data;
650	if (
651	strlen($results) > $this->maxlength ) {
652	break;
653	}
654	}
655
656	// gunzip
657	if ( $is_gzipped ) {
658	// per http://www.php.net/manual/en/function.gzencode.php
659	$results = substr($results, 10);
660	$results = gzinflate($results);
661	}
662
663	if ($this->read_timeout > 0 && $this->_check_timeout($fp))
664	{
665	$this->status=-100;
666	return false;
667	}
668
669	// check if there is a a redirect meta tag
670
671	if(preg_match("'<meta[\s]http-equiv[^>]?content[\s]=[\s][\"\']?\d+;[\s]+URL[\s]=[\s]([^\"\']*?)[\"\']?>'i",$results,$match))
672	{
673	$this->_redirectaddr = $this->_expandlinks($match[1],$URI);
674	}
675
676	// have we hit our frame depth and is there frame src to fetch?
677	if(($this->_framedepth < $this->maxframes) && preg_match_all("'<frame\s+.src[\s]=[\'\"]?([^\'\"\>]+)'i",$results,$match))
678	{
679	$this->results[] = $results;
680	for($x=0; $x<count($match[1]); $x++)
681	$this->_frameurls[] = $this->_expandlinks($match[1][$x],$URI_PARTS["scheme"]."://".$this->host);
682	}
683	// have we already fetched framed content?
684	elseif(is_array($this->results))
685	$this->results[] = $results;
686	// no framed content
687	else
688	$this->results = $results;
689
690	return true;
691	}
692
693	/======================================================================\
694	Function: _curlrequest
695	Purpose: go get the https data from the server using curl
696	Input: $url the url to fetch
697	$URI the full URI
698	$body body contents to send if any (POST)
699	Output:
700	\======================================================================/
701
702	function _curlrequest($url,$URI,$http_method,$content_type="",$body="")
703	{
704	if($this->passcookies && $this->_redirectaddr)
705	$this->setcookies();
706
707	$headers = array();
708
709	$URI_PARTS = parse_url($URI);
710	if(empty($url))
711	$url = "/";
712	// GET ... header not needed for curl
713	//$headers[] = $http_method." ".$url." ".$this->_httpversion;
714	if(!empty($this->agent))
715	$headers[] = "User-Agent: ".$this->agent;
716	if(!empty($this->host))
717	$headers[] = "Host: ".$this->host;
718	if(!empty($this->accept))
719	$headers[] = "Accept: ".$this->accept;
720	if(!empty($this->referer))
721	$headers[] = "Referer: ".$this->referer;
722	if(!empty($this->cookies))
723	{
724	if(!is_array($this->cookies))
725	$this->cookies = (array)$this->cookies;
726
727	reset($this->cookies);
728	if ( count($this->cookies) > 0 ) {
729	$cookie_str = 'Cookie: ';
730	foreach ( $this->cookies as $cookieKey => $cookieVal ) {
731	$cookie_str .= $cookieKey."=".urlencode($cookieVal)."; ";
732	}
733	$headers[] = substr($cookie_str,0,-2);
734	}
735	}
736	if(!empty($this->rawheaders))
737	{
738	if(!is_array($this->rawheaders))
739	$this->rawheaders = (array)$this->rawheaders;
740	while(list($headerKey,$headerVal) = each($this->rawheaders))
741	$headers[] = $headerKey.": ".$headerVal;
742	}
743	if(!empty($content_type)) {
744	if ($content_type == "multipart/form-data")
745	$headers[] = "Content-type: $content_type; boundary=".$this->_mime_boundary;
746	else
747	$headers[] = "Content-type: $content_type";
748	}
749	if(!empty($body))
750	$headers[] = "Content-length: ".strlen($body);
751	if(!empty($this->user) \|\| !empty($this->pass))
752	$headers[] = "Authorization: BASIC ".base64_encode(urldecode($this->user).":".urldecode($this->pass));
753
754	for($curr_header = 0; $curr_header < count($headers); $curr_header++) {
755	$cmdline_params .= " -H \"".$headers[$curr_header]."\"";
756	}
757
758	if(!empty($body))
759	$cmdline_params .= " -d \"$body\"";
760
761	if($this->read_timeout > 0)
762	$cmdline_params .= " -m ".$this->read_timeout;
763
764	$headerfile = uniqid(time());
765
766	# accept self-signed certs
767	// mbi: removed, as it breaks on older cURL's
768	//$cmdline_params .= " -k";
769	if (defined ('RSS_SNOOPY_EXTRA_PARAMS')) {
770	$cmdline_params .= (" " . RSS_SNOOPY_EXTRA_PARAMS ." ");
771	}
772	$full_cmdline = $this->curl_path
773	." -D \"".$this -> tmpdir
774	."$headerfile\""
775	.escapeshellcmd($cmdline_params)." "
776	.escapeshellcmd($URI);
777	//die($full_cmdline);
778	exec($full_cmdline,$results,$return);
779
780	if($return)
781	{
782	$this->error = "Error: cURL could not retrieve the document, error $return.";
783	return false;
784	}
785
786
787	$results = implode("\r\n",$results);
788
789	$result_headers = file($this -> tmpdir . $headerfile);
790
791	$this->_redirectaddr = false;
792	unset($this->headers);
793
794	for($currentHeader = 0; $currentHeader < count($result_headers); $currentHeader++)
795	{
796
797	// if a header begins with Location: or URI:, set the redirect
798	if(preg_match("/^(Location: \|URI: )/i",$result_headers[$currentHeader]))
799	{
800	// get URL portion of the redirect
801	preg_match("/^(Location: \|URI:)(.*)/",chop($result_headers[$currentHeader]),$matches);
802	// look for :// in the Location header to see if hostname is included
803	if(!preg_match("\|\:\/\/\|",$matches[2]))
804	{
805	// no host in the path, so prepend
806	$this->_redirectaddr = $URI_PARTS["scheme"]."://".$this->host.":".$this->port;
807	// eliminate double slash
808	if(!preg_match("\|^/\|",$matches[2]))
809	$this->_redirectaddr .= "/".$matches[2];
810	else
811	$this->_redirectaddr .= $matches[2];
812	}
813	else
814	$this->_redirectaddr = $matches[2];
815	}
816
817	if(preg_match("\|^HTTP/\|",$result_headers[$currentHeader]))
818	{
819	$this->response_code = $result_headers[$currentHeader];
820	if(preg_match("\|^HTTP/[^\s]\s(.?)\s\|",$this->response_code, $match))
821	{
822	$this->status= $match[1];
823	}
824	}
825	$this->headers[] = $result_headers[$currentHeader];
826	}
827
828	// check if there is a a redirect meta tag
829
830	if(preg_match("'<meta[\s]http-equiv[^>]?content[\s]=[\s][\"\']?\d+;[\s]+URL[\s]=[\s]([^\"\']*?)[\"\']?>'i",$results,$match))
831	{
832	$this->_redirectaddr = $this->_expandlinks($match[1],$URI);
833	}
834
835	// have we hit our frame depth and is there frame src to fetch?
836	if(($this->_framedepth < $this->maxframes) && preg_match_all("'<frame\s+.src[\s]=[\'\"]?([^\'\"\>]+)'i",$results,$match))
837	{
838	$this->results[] = $results;
839	for($x=0; $x<count($match[1]); $x++)
840	$this->_frameurls[] = $this->_expandlinks($match[1][$x],$URI_PARTS["scheme"]."://".$this->host);
841	}
842	// have we already fetched framed content?
843	elseif(is_array($this->results))
844	$this->results[] = $results;
845	// no framed content
846	else
847	$this->results = $results;
848
849	unlink($this -> tmpdir . $headerfile);
850
851	return true;
852	}
853
854	/======================================================================\
855	Function: setcookies()
856	Purpose: set cookies for a redirection
857	\======================================================================/
858
859	function setcookies()
860	{
861	for($x=0; $x<count($this->headers); $x++)
862	{
863	if(preg_match("/^set-cookie:[\s]+([^=]+)=([^;]+)/i", $this->headers[$x],$match))
864	$this->cookies[$match[1]] = $match[2];
865	}
866	}
867
868
869	/======================================================================\
870	Function: _check_timeout
871	Purpose: checks whether timeout has occurred
872	Input: $fp file pointer
873	\======================================================================/
874
875	function _check_timeout($fp)
876	{
877	if ($this->read_timeout > 0) {
878	$fp_status = socket_get_status($fp);
879	if ($fp_status["timed_out"]) {
880	$this->timed_out = true;
881	return true;
882	}
883	}
884	return false;
885	}
886
887	/======================================================================\
888	Function: _connect
889	Purpose: make a socket connection
890	Input: $fp file pointer
891	\======================================================================/
892
893	function _connect(&$fp)
894	{
895	if(!empty($this->proxy_host) && !empty($this->proxy_port))
896	{
897	$this->_isproxy = true;
898	$host = $this->proxy_host;
899	$port = $this->proxy_port;
900	}
901	else
902	{
903	$host = $this->host;
904	$port = $this->port;
905	}
906
907	$this->status = 0;
908
909	if($this->_scheme == "https")
910	{
911	$host = "ssl://" . $host;
912	}
913
914	if($fp = fsockopen(
915	$host,
916	$port,
917	$errno,
918	$errstr,
919	$this->_fp_timeout
920	))
921	{
922	// socket connection succeeded
923
924	return true;
925	}
926	else
927	{
928	// socket connection failed
929	$this->status = $errno;
930	switch($errno)
931	{
932	case -3:
933	$this->error="socket creation failed (-3)";
934	case -4:
935	$this->error="dns lookup failure (-4)";
936	case -5:
937	$this->error="connection refused or timed out (-5)";
938	default:
939	$this->error="connection failed (".$errno.")";
940	}
941	return false;
942	}
943	}
944	/======================================================================\
945	Function: _disconnect
946	Purpose: disconnect a socket connection
947	Input: $fp file pointer
948	\======================================================================/
949
950	function _disconnect($fp)
951	{
952	return(fclose($fp));
953	}
954
955
956	/======================================================================\
957	Function: _prepare_post_body
958	Purpose: Prepare post body according to encoding type
959	Input: $formvars - form variables
960	$formfiles - form upload files
961	Output: post body
962	\======================================================================/
963
964	function _prepare_post_body($formvars, $formfiles)
965	{
966	settype($formvars, "array");
967	settype($formfiles, "array");
968
969	if (count($formvars) == 0 && count($formfiles) == 0)
970	return;
971
972	switch ($this->_submit_type) {
973	case "application/x-www-form-urlencoded":
974	reset($formvars);
975	while(list($key,$val) = each($formvars)) {
976	if (is_array($val) \|\| is_object($val)) {
977	while (list($cur_key, $cur_val) = each($val)) {
978	$postdata .= urlencode($key)."[]=".urlencode($cur_val)."&";
979	}
980	} else
981	$postdata .= urlencode($key)."=".urlencode($val)."&";
982	}
983	break;
984
985	case "multipart/form-data":
986	$this->_mime_boundary = "Snoopy".md5(uniqid(microtime()));
987
988	reset($formvars);
989	whi