<?php
class Snoopy
{
var $host = "www.php.net"; // host name we are connecting to
var $port = 80; // port we are connecting to
var $proxy_host = ""; // proxy host to use
var $proxy_port = ""; // proxy port to use
var $proxy_user = ""; // proxy user to use
var $proxy_pass = ""; // proxy password to use
var $agent = "Snoopy v1.2.4"; // agent we masquerade as
var $referer = ""; // referer info to pass
var $cookies = array(); // array of cookies to pass
var $rawheaders = array(); // array of raw headers to send
var $maxredirs = 5; // http redirection depth maximum. 0 = disallow
var $lastredirectaddr = ""; // contains address of last redirected address
var $offsiteok = true; // allows redirection off-site
var $maxframes = 0; // frame content depth maximum. 0 = disallow
var $expandlinks = true; // expand links to fully qualified URLs.
var $passcookies = true; // pass set cookies back through redirects
var $user = ""; // user for http authentication
var $pass = ""; // password for http authentication
var $accept = "image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, */*";
var $results = "./content"; // where the content is put
var $error = ""; // error messages sent here
var $response_code = ""; // response code returned from server
var $headers = array(); // headers returned from server sent here
var $maxlength = 500000; // max return data length (body)
var $read_timeout = 0; // timeout on read operations, in seconds
var $timed_out = false; // if a read operation timed out
var $status = 0; // http request status
var $temp_dir = "/tmp"; // temporary directory that the webserver
var $curl_path = "/usr/local/bin/curl";
var $_maxlinelen = 4096; // max line length (headers)
var $_httpmethod = "GET"; // default http request method
var $_httpversion = "HTTP/1.0"; // default http request version
var $_submit_method = "POST"; // default submit method
var $_submit_type = "application/x-www-form-urlencoded"; // default submit type
var $_mime_boundary = ""; // MIME boundary for multipart/form-data submit type
var $_redirectaddr = false; // will be set if page fetched is a redirect
var $_redirectdepth = 0; // increments on an http redirect
var $_frameurls = array(); // frame src urls
var $_framedepth = 0; // increments on frame depth
var $_isproxy = false; // set if using a proxy server
var $_fp_timeout = 30; // timeout for socket connection/*======================================================================*\
Function: fetch
Purpose: fetch the contents of a web page
(and possibly other protocols in the
future like ftp, nntp, gopher, etc.)
Input: $URI the location of the page to fetch
Output: $this->results the output text from the fetch
\*======================================================================*/ function fetch($URI)
{
//preg_match("|^([^:]+)://([^:/]+)(:[\d]+)*(.*)|",$URI,$URI_PARTS);
$URI_PARTS = parse_url($URI);
if (!empty($URI_PARTS["user"]))
$this->user = $URI_PARTS["user"];
if (!empty($URI_PARTS["pass"]))
$this->pass = $URI_PARTS["pass"];
if (empty($URI_PARTS["query"]))
$URI_PARTS["query"] = '';
if (empty($URI_PARTS["path"]))
$URI_PARTS["path"] = '';
switch(strtolower($URI_PARTS["scheme"]))
{
case "http":
$this->host = $URI_PARTS["host"];
if(!empty($URI_PARTS["port"]))
$this->port = $URI_PARTS["port"];
if($this->_connect($fp))
{
if($this->_isproxy)
{
// using proxy, send entire URI
$this->_httprequest($URI,$fp,$URI,$this->_httpmethod);
}
else
{
$path = $URI_PARTS["path"].($URI_PARTS["query"] ? "?".$URI_PARTS["query"] : "");
// no proxy, send only the path
$this->_httprequest($path, $fp, $URI, $this->_httpmethod);
}
$this->_disconnect($fp); if($this->_redirectaddr)
{
/* url was redirected, check if we've hit the max depth */
if($this->maxredirs > $this->_redirectdepth)
{
// only follow redirect if it's on this site, or offsiteok is true
if(preg_match("|^http://".preg_quote($this->host)."|i",$this->_redirectaddr) || $this->offsiteok)
{
/* follow the redirect */
$this->_redirectdepth++;
$this->lastredirectaddr=$this->_redirectaddr;
$this->fetch($this->_redirectaddr);
}
}
} if($this->_framedepth < $this->maxframes && count($this->_frameurls) > 0)
{
$frameurls = $this->_frameurls;
$this->_frameurls = array();
while(list(,$frameurl) = each($frameurls))
{
if($this->_framedepth < $this->maxframes)
{
$this->fetch($frameurl);
$this->_framedepth++;
}
else
break;
}
}
}
else
{
return false;
}
return true;
break;
case "https":
if(!$this->curl_path)
return false;
if(function_exists("is_executable"))
if (!is_executable($this->curl_path))
return false;
$this->host = $URI_PARTS["host"];
if(!empty($URI_PARTS["port"]))
$this->port = $URI_PARTS["port"];
if($this->_isproxy)
{
// using proxy, send entire URI
$this->_httpsrequest($URI,$URI,$this->_httpmethod);
}
else
{
$path = $URI_PARTS["path"].($URI_PARTS["query"] ? "?".$URI_PARTS["query"] : "");
// no proxy, send only the path
$this->_httpsrequest($path, $URI, $this->_httpmethod);
} if($this->_redirectaddr)
{
/* url was redirected, check if we've hit the max depth */
if($this->maxredirs > $this->_redirectdepth)
{
// only follow redirect if it's on this site, or offsiteok is true
if(preg_match("|^http://".preg_quote($this->host)."|i",$this->_redirectaddr) || $this->offsiteok)
{
/* follow the redirect */
$this->_redirectdepth++;
$this->lastredirectaddr=$this->_redirectaddr;
$this->fetch($this->_redirectaddr);
}
}
} if($this->_framedepth < $this->maxframes && count($this->_frameurls) > 0)
{
$frameurls = $this->_frameurls;
$this->_frameurls = array(); while(list(,$frameurl) = each($frameurls))
{
if($this->_framedepth < $this->maxframes)
{
$this->fetch($frameurl);
$this->_framedepth++;
}
else
break;
}
}
return true;
break;
default:
// not a valid protocol
$this->error = 'Invalid protocol "'.$URI_PARTS["scheme"].'"\n';
return false;
break;
}
return true;
}
class Snoopy
{
var $host = "www.php.net"; // host name we are connecting to
var $port = 80; // port we are connecting to
var $proxy_host = ""; // proxy host to use
var $proxy_port = ""; // proxy port to use
var $proxy_user = ""; // proxy user to use
var $proxy_pass = ""; // proxy password to use
var $agent = "Snoopy v1.2.4"; // agent we masquerade as
var $referer = ""; // referer info to pass
var $cookies = array(); // array of cookies to pass
var $rawheaders = array(); // array of raw headers to send
var $maxredirs = 5; // http redirection depth maximum. 0 = disallow
var $lastredirectaddr = ""; // contains address of last redirected address
var $offsiteok = true; // allows redirection off-site
var $maxframes = 0; // frame content depth maximum. 0 = disallow
var $expandlinks = true; // expand links to fully qualified URLs.
var $passcookies = true; // pass set cookies back through redirects
var $user = ""; // user for http authentication
var $pass = ""; // password for http authentication
var $accept = "image/gif, image/x-xbitmap, image/jpeg, image/pjpeg, */*";
var $results = "./content"; // where the content is put
var $error = ""; // error messages sent here
var $response_code = ""; // response code returned from server
var $headers = array(); // headers returned from server sent here
var $maxlength = 500000; // max return data length (body)
var $read_timeout = 0; // timeout on read operations, in seconds
var $timed_out = false; // if a read operation timed out
var $status = 0; // http request status
var $temp_dir = "/tmp"; // temporary directory that the webserver
var $curl_path = "/usr/local/bin/curl";
var $_maxlinelen = 4096; // max line length (headers)
var $_httpmethod = "GET"; // default http request method
var $_httpversion = "HTTP/1.0"; // default http request version
var $_submit_method = "POST"; // default submit method
var $_submit_type = "application/x-www-form-urlencoded"; // default submit type
var $_mime_boundary = ""; // MIME boundary for multipart/form-data submit type
var $_redirectaddr = false; // will be set if page fetched is a redirect
var $_redirectdepth = 0; // increments on an http redirect
var $_frameurls = array(); // frame src urls
var $_framedepth = 0; // increments on frame depth
var $_isproxy = false; // set if using a proxy server
var $_fp_timeout = 30; // timeout for socket connection/*======================================================================*\
Function: fetch
Purpose: fetch the contents of a web page
(and possibly other protocols in the
future like ftp, nntp, gopher, etc.)
Input: $URI the location of the page to fetch
Output: $this->results the output text from the fetch
\*======================================================================*/ function fetch($URI)
{
//preg_match("|^([^:]+)://([^:/]+)(:[\d]+)*(.*)|",$URI,$URI_PARTS);
$URI_PARTS = parse_url($URI);
if (!empty($URI_PARTS["user"]))
$this->user = $URI_PARTS["user"];
if (!empty($URI_PARTS["pass"]))
$this->pass = $URI_PARTS["pass"];
if (empty($URI_PARTS["query"]))
$URI_PARTS["query"] = '';
if (empty($URI_PARTS["path"]))
$URI_PARTS["path"] = '';
switch(strtolower($URI_PARTS["scheme"]))
{
case "http":
$this->host = $URI_PARTS["host"];
if(!empty($URI_PARTS["port"]))
$this->port = $URI_PARTS["port"];
if($this->_connect($fp))
{
if($this->_isproxy)
{
// using proxy, send entire URI
$this->_httprequest($URI,$fp,$URI,$this->_httpmethod);
}
else
{
$path = $URI_PARTS["path"].($URI_PARTS["query"] ? "?".$URI_PARTS["query"] : "");
// no proxy, send only the path
$this->_httprequest($path, $fp, $URI, $this->_httpmethod);
}
$this->_disconnect($fp); if($this->_redirectaddr)
{
/* url was redirected, check if we've hit the max depth */
if($this->maxredirs > $this->_redirectdepth)
{
// only follow redirect if it's on this site, or offsiteok is true
if(preg_match("|^http://".preg_quote($this->host)."|i",$this->_redirectaddr) || $this->offsiteok)
{
/* follow the redirect */
$this->_redirectdepth++;
$this->lastredirectaddr=$this->_redirectaddr;
$this->fetch($this->_redirectaddr);
}
}
} if($this->_framedepth < $this->maxframes && count($this->_frameurls) > 0)
{
$frameurls = $this->_frameurls;
$this->_frameurls = array();
while(list(,$frameurl) = each($frameurls))
{
if($this->_framedepth < $this->maxframes)
{
$this->fetch($frameurl);
$this->_framedepth++;
}
else
break;
}
}
}
else
{
return false;
}
return true;
break;
case "https":
if(!$this->curl_path)
return false;
if(function_exists("is_executable"))
if (!is_executable($this->curl_path))
return false;
$this->host = $URI_PARTS["host"];
if(!empty($URI_PARTS["port"]))
$this->port = $URI_PARTS["port"];
if($this->_isproxy)
{
// using proxy, send entire URI
$this->_httpsrequest($URI,$URI,$this->_httpmethod);
}
else
{
$path = $URI_PARTS["path"].($URI_PARTS["query"] ? "?".$URI_PARTS["query"] : "");
// no proxy, send only the path
$this->_httpsrequest($path, $URI, $this->_httpmethod);
} if($this->_redirectaddr)
{
/* url was redirected, check if we've hit the max depth */
if($this->maxredirs > $this->_redirectdepth)
{
// only follow redirect if it's on this site, or offsiteok is true
if(preg_match("|^http://".preg_quote($this->host)."|i",$this->_redirectaddr) || $this->offsiteok)
{
/* follow the redirect */
$this->_redirectdepth++;
$this->lastredirectaddr=$this->_redirectaddr;
$this->fetch($this->_redirectaddr);
}
}
} if($this->_framedepth < $this->maxframes && count($this->_frameurls) > 0)
{
$frameurls = $this->_frameurls;
$this->_frameurls = array(); while(list(,$frameurl) = each($frameurls))
{
if($this->_framedepth < $this->maxframes)
{
$this->fetch($frameurl);
$this->_framedepth++;
}
else
break;
}
}
return true;
break;
default:
// not a valid protocol
$this->error = 'Invalid protocol "'.$URI_PARTS["scheme"].'"\n';
return false;
break;
}
return true;
}
Function: _httprequest
Purpose: go get the http data from the server
Input: $url the url to fetch
$fp the current open file pointer
$URI the full URI
$body body contents to send if any (POST)
Output:
\*======================================================================*/
function _httprequest($url,$fp,$URI,$http_method,$content_type="",$body="")
{
$cookie_headers = '';
if($this->passcookies && $this->_redirectaddr)
$this->setcookies();
$URI_PARTS = parse_url($URI);
if(empty($url))
$url = "/";
$headers = $http_method." ".$url." ".$this->_httpversion."\r\n";
if(!empty($this->agent))
$headers .= "User-Agent: ".$this->agent."\r\n";
if(!empty($this->host) && !isset($this->rawheaders['Host'])) {
$headers .= "Host: ".$this->host;
if(!empty($this->port))
$headers .= ":".$this->port;
$headers .= "\r\n";
}
if(!empty($this->accept))
$headers .= "Accept: ".$this->accept."\r\n";
if(!empty($this->referer))
$headers .= "Referer: ".$this->referer."\r\n";
if(!empty($this->cookies))
{
if(!is_array($this->cookies))
$this->cookies = (array)$this->cookies;
reset($this->cookies);
if ( count($this->cookies) > 0 ) {
$cookie_headers .= 'Cookie: ';
foreach ( $this->cookies as $cookieKey => $cookieVal ) {
$cookie_headers .= $cookieKey."=".urlencode($cookieVal)."; ";
}
$headers .= substr($cookie_headers,0,-2) . "\r\n";
}
}
if(!empty($this->rawheaders))
{
if(!is_array($this->rawheaders))
$this->rawheaders = (array)$this->rawheaders;
while(list($headerKey,$headerVal) = each($this->rawheaders))
$headers .= $headerKey.": ".$headerVal."\r\n";
}
if(!empty($content_type)) {
$headers .= "Content-type: $content_type";
if ($content_type == "multipart/form-data")
$headers .= "; boundary=".$this->_mime_boundary;
$headers .= "\r\n";
}
if(!empty($body))
$headers .= "Content-length: ".strlen($body)."\r\n";
if(!empty($this->user) || !empty($this->pass))
$headers .= "Authorization: Basic ".base64_encode($this->user.":".$this->pass)."\r\n";
//add proxy auth headers
if(!empty($this->proxy_user))
$headers .= 'Proxy-Authorization: ' . 'Basic ' . base64_encode($this->proxy_user . ':' . $this->proxy_pass)."\r\n";
$headers .= "\r\n";
// set the read timeout if needed
if ($this->read_timeout > 0)
socket_set_timeout($fp, $this->read_timeout);
$this->timed_out = false;
fwrite($fp,$headers.$body,strlen($headers.$body));
$this->_redirectaddr = false;
unset($this->headers);
while($currentHeader = fgets($fp,$this->_maxlinelen))
{
if ($this->read_timeout > 0 && $this->_check_timeout($fp))
{
$this->status=-100;
return false;
}
if($currentHeader == "\r\n")
break;
// if a header begins with Location: or URI:, set the redirect
if(preg_match("/^(Location:|URI:)/i",$currentHeader))
{
// get URL portion of the redirect
preg_match("/^(Location:|URI:)[ ]+(.*)/i",chop($currentHeader),$matches);
// look for :// in the Location header to see if hostname is included
if(!preg_match("|\:\/\/|",$matches[2]))
{
// no host in the path, so prepend
$this->_redirectaddr = $URI_PARTS["scheme"]."://".$this->host.":".$this->port;
// eliminate double slash
if(!preg_match("|^/|",$matches[2]))
$this->_redirectaddr .= "/".$matches[2];
else
$this->_redirectaddr .= $matches[2];
}
else
$this->_redirectaddr = $matches[2];
}
if(preg_match("|^HTTP/|",$currentHeader))
{
if(preg_match("|^HTTP/[^\s]*\s(.*?)\s|",$currentHeader, $status))
{
$this->status= $status[1];
}
$this->response_code = $currentHeader;
}
$this->headers[] = $currentHeader;
} $results = '';
do {
$_data = fread($fp, $this->maxlength);
if (strlen($_data) == 0) {
break;
}
$results .= $_data;
} while(true); if ($this->read_timeout > 0 && $this->_check_timeout($fp))
{
$this->status=-100;
return false;
}
// check if there is a a redirect meta tag
if(preg_match("'<meta[\s]*http-equiv[^>]*?content[\s]*=[\s]*[\"\']?\d+;[\s]*URL[\s]*=[\s]*([^\"\']*?)[\"\']?>'i",$results,$match)) {
$this->_redirectaddr = $this->_expandlinks($match[1],$URI);
} // have we hit our frame depth and is there frame src to fetch?
if(($this->_framedepth < $this->maxframes) && preg_match_all("'<frame\s+.*src[\s]*=[\'\"]?([^\'\"\>]+)'i",$results,$match))
{
$this->results[] = $results;
for($x=0; $x<count($match[1]); $x++)
$this->_frameurls[] = $this->_expandlinks($match[1][$x],$URI_PARTS["scheme"]."://".$this->host);
}
// have we already fetched framed content?
elseif(is_array($this->results))
$this->results[] = $results;
// no framed content
else
$this->results = $results;
return true;
}
Function: _httpsrequest
Purpose: go get the https data from the server using curl
Input: $url the url to fetch
$URI the full URI
$body body contents to send if any (POST)
Output:
\*======================================================================*/
function _httpsrequest($url,$URI,$http_method,$content_type="",$body="")
{
if($this->passcookies && $this->_redirectaddr)
$this->setcookies(); $headers = array();
$URI_PARTS = parse_url($URI);
if(empty($url))
$url = "/";
// GET ... header not needed for curl
//$headers[] = $http_method." ".$url." ".$this->_httpversion;
if(!empty($this->agent))
$headers[] = "User-Agent: ".$this->agent;
if(!empty($this->host))
if(!empty($this->port))
$headers[] = "Host: ".$this->host.":".$this->port;
else
$headers[] = "Host: ".$this->host;
if(!empty($this->accept))
$headers[] = "Accept: ".$this->accept;
if(!empty($this->referer))
$headers[] = "Referer: ".$this->referer;
if(!empty($this->cookies))
{
if(!is_array($this->cookies))
$this->cookies = (array)$this->cookies;
reset($this->cookies);
if ( count($this->cookies) > 0 ) {
$cookie_str = 'Cookie: ';
foreach ( $this->cookies as $cookieKey => $cookieVal ) {
$cookie_str .= $cookieKey."=".urlencode($cookieVal)."; ";
}
$headers[] = substr($cookie_str,0,-2);
}
}
if(!empty($this->rawheaders))
{
if(!is_array($this->rawheaders))
$this->rawheaders = (array)$this->rawheaders;
while(list($headerKey,$headerVal) = each($this->rawheaders))
$headers[] = $headerKey.": ".$headerVal;
}
if(!empty($content_type)) {
if ($content_type == "multipart/form-data")
$headers[] = "Content-type: $content_type; boundary=".$this->_mime_boundary;
else
$headers[] = "Content-type: $content_type";
}
if(!empty($body))
$headers[] = "Content-length: ".strlen($body);
if(!empty($this->user) || !empty($this->pass))
$headers[] = "Authorization: BASIC ".base64_encode($this->user.":".$this->pass);
for($curr_header = 0; $curr_header < count($headers); $curr_header++) {
$safer_header = strtr( $headers[$curr_header], "\"", " " );
$cmdline_params .= " -H \"".$safer_header."\"";
}
if(!empty($body))
$cmdline_params .= " -d \"$body\"";
if($this->read_timeout > 0)
$cmdline_params .= " -m ".$this->read_timeout;
$headerfile = tempnam($temp_dir, "sno"); exec($this->curl_path." -k -D \"$headerfile\"".$cmdline_params." \"".escapeshellcmd($URI)."\"",$results,$return);
if($return)
{
$this->error = "Error: cURL could not retrieve the document, error $return.";
return false;
}
$results = implode("\r\n",$results);
$result_headers = file("$headerfile");
$this->_redirectaddr = false;
unset($this->headers);
for($currentHeader = 0; $currentHeader < count($result_headers); $currentHeader++)
{
// if a header begins with Location: or URI:, set the redirect
if(preg_match("/^(Location: |URI: )/i",$result_headers[$currentHeader]))
{
// get URL portion of the redirect
preg_match("/^(Location: |URI:)\s+(.*)/",chop($result_headers[$currentHeader]),$matches);
// look for :// in the Location header to see if hostname is included
if(!preg_match("|\:\/\/|",$matches[2]))
{
// no host in the path, so prepend
$this->_redirectaddr = $URI_PARTS["scheme"]."://".$this->host.":".$this->port;
// eliminate double slash
if(!preg_match("|^/|",$matches[2]))
$this->_redirectaddr .= "/".$matches[2];
else
$this->_redirectaddr .= $matches[2];
}
else
$this->_redirectaddr = $matches[2];
}
if(preg_match("|^HTTP/|",$result_headers[$currentHeader]))
$this->response_code = $result_headers[$currentHeader]; $this->headers[] = $result_headers[$currentHeader];
} // check if there is a a redirect meta tag
if(preg_match("'<meta[\s]*http-equiv[^>]*?content[\s]*=[\s]*[\"\']?\d+;[\s]*URL[\s]*=[\s]*([^\"\']*?)[\"\']?>'i",$results,$match))
{
$this->_redirectaddr = $this->_expandlinks($match[1],$URI);
} // have we hit our frame depth and is there frame src to fetch?
if(($this->_framedepth < $this->maxframes) && preg_match_all("'<frame\s+.*src[\s]*=[\'\"]?([^\'\"\>]+)'i",$results,$match))
{
$this->results[] = $results;
for($x=0; $x<count($match[1]); $x++)
$this->_frameurls[] = $this->_expandlinks($match[1][$x],$URI_PARTS["scheme"]."://".$this->host);
}
// have we already fetched framed content?
elseif(is_array($this->results))
$this->results[] = $results;
// no framed content
else
$this->results = $results; unlink("$headerfile");
return true;
}/*======================================================================*\
Function: _connect
Purpose: make a socket connection
Input: $fp file pointer
\*======================================================================*/
function _connect(&$fp)
{
if(!empty($this->proxy_host) && !empty($this->proxy_port))
{
$this->_isproxy = true;
$host = $this->proxy_host;
$port = $this->proxy_port;
}
else
{
$host = $this->host;
$port = $this->port;
}
$this->status = 0;
if($fp = fsockopen(
$host,
$port,
$errno,
$errstr,
$this->_fp_timeout
))
{
// socket connection succeeded return true;
}
else
{
// socket connection failed
$this->status = $errno;
switch($errno)
{
case -3:
$this->error="socket creation failed (-3)";
case -4:
$this->error="dns lookup failure (-4)";
case -5:
$this->error="connection refused or timed out (-5)";
default:
$this->error="connection failed (".$errno.")";
}
return false;
}
}
/*======================================================================*\
Function: _disconnect
Purpose: disconnect a socket connection
Input: $fp file pointer
\*======================================================================*/
function _disconnect($fp)
{
return(fclose($fp));
}
/*======================================================================*\
Function: _prepare_post_body
Purpose: Prepare post body according to encoding type
Input: $formvars - form variables
$formfiles - form upload files
Output: post body
\*======================================================================*/
function _prepare_post_body($formvars, $formfiles)
{
settype($formvars, "array");
settype($formfiles, "array");
$postdata = ''; if (count($formvars) == 0 && count($formfiles) == 0)
return;
switch ($this->_submit_type) {
case "application/x-www-form-urlencoded":
reset($formvars);
while(list($key,$val) = each($formvars)) {
if (is_array($val) || is_object($val)) {
while (list($cur_key, $cur_val) = each($val)) {
$postdata .= urlencode($key)."[]=".urlencode($cur_val)."&";
}
} else
$postdata .= urlencode($key)."=".urlencode($val)."&";
}
break; case "multipart/form-data":
$this->_mime_boundary = "Snoopy".md5(uniqid(microtime()));
reset($formvars);
while(list($key,$val) = each($formvars)) {
if (is_array($val) || is_object($val)) {
while (list($cur_key, $cur_val) = each($val)) {
$postdata .= "--".$this->_mime_boundary."\r\n";
$postdata .= "Content-Disposition: form-data; name=\"$key\[\]\"\r\n\r\n";
$postdata .= "$cur_val\r\n";
}
} else {
$postdata .= "--".$this->_mime_boundary."\r\n";
$postdata .= "Content-Disposition: form-data; name=\"$key\"\r\n\r\n";
$postdata .= "$val\r\n";
}
}
reset($formfiles);
while (list($field_name, $file_names) = each($formfiles)) {
settype($file_names, "array");
while (list(, $file_name) = each($file_names)) {
if (!is_readable($file_name)) continue; $fp = fopen($file_name, "r");
$file_content = fread($fp, filesize($file_name));
fclose($fp);
$base_name = basename($file_name); $postdata .= "--".$this->_mime_boundary."\r\n";
$postdata .= "Content-Disposition: form-data; name=\"$field_name\"; filename=\"$base_name\"\r\n\r\n";
$postdata .= "$file_content\r\n";
}
}
$postdata .= "--".$this->_mime_boundary."--\r\n";
break;
} return $postdata;
}
}?>
// preg_match_all('/<div\s*id=\"article_content\"\s*class=\"article_content\">(.*)<\/div>/iS', $content, $matchs);
print_r($matchs);有两个关于文章内容的表达式帮忙看下，谢谢了。下面贴我的应用代码：<?PHP
//网络爬虫
//include_once('conn_db.php');
include_once('Snoopy.class.php');
//require("lib_splitword_full.php");$spider = new Snoopy();
//$sp = new SplitWord();$spider->begin_url="http://blog.csdn.net/tianlesoftware/article/details/6704642";
//$array_aurl=$spider->fetch_turl($spider->begin_url);
//print_r($array_aurl);$spider->fetch($spider->begin_url);
$_text=$spider->results;$_title=getTitle($_text);
echo $_title;$_details=getdetails($_text);
print_r($_details);
//echo get_magic_quotes_gpc();function getdetails($content){
$content=addslashes($content);
preg_match_all('/<div id=\"article_content\" class=\"article_content\">(.*)<\/div>/iSU',$content,$matchs);
// preg_match_all('/<div\s*id=\"article_content\"\s*class=\"article_content\">(.*)<\/div>/iS', $content, $matchs);
print_r($matchs);
}function getTitle($content){
preg_match('/<title>(.+)<\/title>/i',$content,$matchs);
return $matchs[1];
}
?>