Source Code to Page Forward

<?
/*

* CHANGELOG:
* 01/15/05: PF v1.1-Initial Public Release.
* 10/05/05: PF v1.2-Fixed erratic '302 Moved' errors appearing when certain sites were viewed with the proxy.
* 10/25/05: PF v1.3-Improved CSS2 support; sites like msn.com and urbandictionary.com now work.
*
* README:
* This program is a modified, simplified version of Simple Browser Proxy (located at http://sbp.sf.net.)
* The author of this program felt SBP had a lot of potential but had a few flaws that needed ironing out.
* PF is a PHP program that will let you surf the web through a makeshift proxy (through the web server the
* program runs on) to bypass things like internet filters.
*
* INSTALLATION AND USAGE:
* Change the first two uncommented lines (the ones that start with dollar signs) below this information if
* you need to. Upload this file to a web server running PHP. Make sure the program has proper execute permissions.
* The program is used by going to http://your.webserver.com/PageForward.php?url=x in a browser where x is the
* URL of a site that you wish to view through the proxy. Any links on the original page will be edited and
* 'redirected' through the proxy; things typed in the address bar of your browser will not (unless they appear
* after the 'url='. PageForward will still work if you change the name of the file to something besides
* PageForward.php.
*
* KNOWN BUGS:
* PageForward still has some trouble with CSS as well as forms (anything with text fields and a 'submit' button)
* depending on how some sites are coded--some sites will display/work fine, and others will not. Fixes for these
* issues are planned in a future release.
*
* OTHER INFO:
* This program is released under the GNU Lesser GPL. It is a modified version of Simple Browser Proxy
* (located at http://sbp.sf.net) and I'd like to thank the original author for writing and freely distributing
* SBP and its source code.
*/

//**BEGIN USER CONFIG**
//Page to display by default (if no URL is supplied)
$default_url = "http://www.google.com";
//Tag to prepend page titles
$title_tag = "PF--";
//**END USER CONFIG**

$start_time = microtime();

//Finds the nth position of a string within a string (stolen from http://us3.php.net/strings)
function strnpos($haystack, $needle, $occurance, $pos = 0) {

for ($i = 1; $i <= $occurance; $i++) {
$pos = strpos($haystack, $needle, $pos) + 1;
}
return $pos - 1;
}

//URL parser that works better than PHP's built-in function
function parseURL($url)
{
//protocol(1), auth user(2), auth password(3), hostname(4), path(5), filename(6), file extension(7) and query(8) //
$pattern = "/^(?:(http[s]?):\/\/(?:(.*):(.*)@)?([^\/]+))?((?:[\/])?(?:[^\.]*?)?(?:[\/])?)?(?:([^\/^\.]+)\.([^\?]+))?(?:\?(.+))?$/i";
preg_match($pattern, $url, $matches);

$URI_PARTS["scheme"] = $matches[1];
$URI_PARTS["host"] = $matches[4];
$URI_PARTS["path"] = $matches[5];

return $URI_PARTS;
}

//Turns any local URLs into fully qualified URLs
function completeURLs($HTML, $url)
{
$URI_PARTS = parseURL($url);
$path = trim($URI_PARTS["path"], "/");
$host_url = trim($URI_PARTS["host"], "/");

//$host = $URI_PARTS["scheme"]."://".trim($URI_PARTS["host"], "/")."/".$path; //ORIGINAL
$host = $URI_PARTS["scheme"]."://".$host_url."/".$path."/";
$host_no_path = $URI_PARTS["scheme"]."://".$host_url."/";

//make sure the host doesn't end in '//'
$host = rtrim($host, '/')."/";

//replace '//' with 'http://'
$pattern = "#(?<=\"|'|=)\/\/#"; //the '|=' is experimental as it's probably not necessary
$HTML = preg_replace($pattern, "http://", $HTML);

//matches [src|href|background|action]="/ because in the following pattern the '/' shouldn't stay
$pattern = "#(src|href|background|action)(=\"|='|=(?!'|\"))\/#i";
$HTML = preg_replace($pattern, "\$1\$2".$host_no_path, $HTML);

$pattern = "#(href|src|background|action)(=\"|=(?!'|\")|=')(?!http|ftp|https|\"|'|javascript:|mailto:)#i";
$HTML = preg_replace($pattern, "\$1\$2".$host, $HTML);

//TODO: need to be able to clean off the crap after the action.
$pattern = "#action=(.*?)>#is";
$replace = "action=".$_SERVER['PHP_SELF']."><input type=\"hidden\" name=\"original_url\" value=\"\$1\">";
$HTML = preg_replace($pattern, $replace, $HTML);

//mathces '/[any assortment of chars or nums]/../'
$pattern = "#\/(\w*?)\/\.\.\/(.*?)>#ims";
$replace = "/\$2>";
$HTML = preg_replace($pattern, $replace, $HTML);

//matches '/./'
$pattern = "#\/\.\/(.*?)>#ims";
$replace = "/\$1>";
$HTML = preg_replace($pattern, $replace, $HTML);

//Handle CSS2 imports of CSS files (EXAMPLE: <style type="text/css" media="screen">@import /themes/blue/blue.css";</style>)
if (strpos($HTML, "import url(\"http") == false && (strpos($HTML, "import \"http") == false) && strpos($HTML, "import url(\"www") == false && (strpos($HTML, "import \"www") == false)) {
$pattern = "#import .(.*?).;#ims";
$mainurl = substr($host, 0, strnpos($host, "/", 3));
$replace = "import '".$mainurl."\$1';";
$HTML = preg_replace($pattern, $replace, $HTML);
}
return $HTML;
}

//Redirects link targets through this proxy
function proxyURLs($HTML)
{
$edited_tag = "PF"; //used to check if the link has already been modified by the proxy

//BASE tag needs to be removed for sites like yahoo.com
//OR make the proxy insert the FULL URL to itself
$pattern = "#\<base(.*?)\>#ims";
$replacement = "<!-- <base\$1> -->"; //comment it out for now//
$HTML = preg_replace($pattern, $replacement, $HTML);

//edit <link tags so that 'edited="$edit_tag" ' is just before 'href'
$pattern = "#\<link(.*?)(\shref=)#ims";
$HTML = preg_replace($pattern, "<link\$1 edited=\"".$edited_tag."\"\$2", $HTML);

//matches everything with a </a> after it on the same line....fails to match when that is on another line.
$pattern = "#(?<!edited=\"".$edited_tag."\"\s)(href='|href=\"|href=(?!'|\"))(?=(.+)\</a\>)(?!mailto:|http://ftp|ftp|javascript:|'|\")#ims";
$HTML = preg_replace($pattern, "edited=\"".$edited_tag."\" \$1".$_SERVER['PHP_SELF'].'?url=', $HTML);

return $HTML;
}

//Calculates the differences in microtime captures
function microtime_diff($a, $b)
{

list($a_dec, $a_sec) = explode(" ", $a);
list($b_dec, $b_sec) = explode(" ", $b);

return $b_sec - $a_sec + $b_dec - $a_dec;
}

//Get ready to display data...

$url = $_GET['url'];
if(empty($url)) $url = $default_url;

//Check the URL for protocol, etc....
if(substr($url, 0, 7) != "http://") //didn't start with 'http://'...we have a problem.
{
$url = "http://".$url;
}

//Checks if there was a form redirected to this proxy.
if(!empty($_POST['original_url']))
{
$form_submission = true;
}
else if(!empty($_GET['original_url']))
{
//have to strip off any unwanted stuff from original_url
$url = explode(" ", $_GET['original_url']);
$url = $url[0];

$form_submission = false;
$url = urldecode($url)."?".str_replace("original_url=".urlencode($_GET['original_url'])."&", "", $_SERVER['QUERY_STRING']);
}
if(!$form_submission) //OK, no redirected form so go ahead and fetch a page.
{

$ch = curl_init($url);
curl_setopt($ch, CURLOPT_HEADER, false);
curl_setopt($ch, CURLOPT_FOLLOWLOCATION, true);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
$page = curl_exec($ch);
curl_close($ch);
$HTML = $page;
//$HTML = preg_replace("#\<TITLE\>#i", "<TITLE>".$title_tag, $HTML, 1);
$HTML = preg_replace("#\<(title|TITLE)\>#", "<\$1>".$title_tag, $HTML, 1);
$HTML = completeURLs($HTML, $url); //Complete local links so that they are fully qualified URIs
$HTML = proxyURLs($HTML);  //Complete links so that they pass through this proxy
print_r($HTML); //Output the page using print_r so that frames at least partially are written
flush();

//Calculate execution time and add HTML comment with that info
$duration = microtime_diff($start_time, microtime());
$duration = sprintf("%0.3f", $duration);
echo ("\n<!-- PageForward took $duration seconds to render the page.-->");
}
?>