Store URL

Cuma numpang nempel aja, jangan terlalu dipikirkan. semoga amal, ilmu yang dibagi menjadi berkah.. amin
#!/usr/bin/perl
# $Rev$
# by chudy_fernandez@yahoo.com
# Youtube updates at http://wiki.squid-cache.org/ConfigExamples/DynamicContent/YouTube/Discussion


$|=1;
while (<>) {
    @X = split;
#       $X[1] =~ s/&sig=.*//;
        $x = $X[0] . " ";
        $_ = $X[1];
        $u = $X[1];

                        #photos-X.ak.fbcdn.net where X a-z
if (m/^http:\/\/photos-[a-z]?(.ak.fbcdn.net.*)/) {
        print $x . "http://photos" . $1  . "\n";

            #http://hphotos-ash2.fbcdn.net/290503_1735121757952_1835833812_1146280_4168472_o.jpg
} elsif (m/^http:\/\/photos-ash[1-5]?(.fbcdn.net.*)/) {
        print $x . "http://ash.photos" . $1  . "\n";
   
                       #photos.ak.fbcdn.net where X i(0-9)
                        #a5.sphotos.ak.fbcdn.ne
} elsif (m/^http:\/\/[a-z][0-9]?.(photos.ak.fbcdn.net.*)/) {
        print $x . "http://cdn.photos" . $1  . "\n";

            #BLOGSPOT
} elsif (m/^http:\/\/[1-4].bp.(blogspot.com.*)/) {
        print $x . "http://blog-cdn." . $1  . "\n";

                         #AVAST
} elsif (m/^http:\/\/download[0-9]{3}.(avast.com.*)/) {
          print $x . "http://avast-cdn." . $1  .  "\n";

                         #KAV
} elsif (m/^http:\/\/dnl-[0-9]{2}.(geo.kaspersky.com.*)/) {
          print $x . "http://kav-cdn." . $1  .  "\n";

                         #AVG
} elsif (m/^http:\/\/update.avg.com/) {
          print $x . "http://avg-cdn." . $1  .  "\n";

                        #maps.google.com
} elsif (m/^http:\/\/(cbk|mt|khm|mlt|tbn)[0-9]?(.google\.co(m|\.uk|\.id).*)/) {
        print $x . "http://" . $1  . $2 . "\n";

                        #gstatic and/or wikimapia
} elsif (m/^http:\/\/([a-z])[0-9]?(\.gstatic\.com.*|\.wikimapia\.org.*)/) {
        print $x . "http://" . $1  . $2 . "\n";

                        # compatibility for old cached get_video?video_id
} elsif (m/^http:\/\/([0-9.]{4}|.*\.youtube\.com|.*\.googlevideo\.com|.*\.video\.google\.com|.*\.tube.tachyon\.net\.id\:99).*?(videoplayback\?id=.*?|video_id=.*?)\&(.*?)/) {
        $z = $2; $z =~ s/video_id=/get_video?video_id=/;
        print $x . "http://video-srv.youtube.com.SQUIDINTERNAL/" . $z . "\n";
       
} elsif (m/^http:\/\/([0-9.]{4}|.*\.youtube\.com|.*\.googlevideo\.com|.*\.video\.google\.com)\/videoplayback\?(.*)/) {
        $p_str = $2;
        $tag = "";
        $alg = "";
        $id = "";
        $range = "";
        if ($p_str =~ m/(itag=[0-9]*)/){$tag = "&".$1}
        if ($p_str =~ m/(algorithm=[a-z]*\-[a-z]*)/){$alg = "&".$1}
        if ($p_str =~ m/(id=[a-zA-Z0-9]*)/){$id = "&".$1}
        if ($p_str =~ m/(range=[0-9\-]*)/){$range = "&".$1; $range =~ s/-//; $range =~ s/range=//; }
        print $x . "http://video-srv.youtube.com.SQUIDINTERNAL/" . $tag . "&" . $alg . "&" . $id . "&" . $range . "\n";

} elsif (m/^http:\/\/www\.google-analytics\.com\/__utm\.gif\?.*/) {
        print $x . "http://www.google-analytics.com/__utm.gif\n";

                        #Cache High Latency Ads
} elsif (m/^http:\/\/([a-z0-9.]*)(\.doubleclick\.net|\.quantserve\.com|\.googlesyndication\.com|yieldmanager|cpxinteractive)(.*)/) {
        $y = $3;$z = $2;
        for ($y) {
        s/pixel;.*/pixel/;
        s/activity;.*/activity/;
        s/(imgad[^&]*).*/\1/;
        s/;ord=[?0-9]*//;
        s/;&timestamp=[0-9]*//;
        s/[&?]correlator=[0-9]*//;
        s/&cookie=[^&]*//;
        s/&ga_hid=[^&]*//;
        s/&ga_vid=[^&]*//;
        s/&ga_sid=[^&]*//;
        # s/&prev_slotnames=[^&]*//
        # s/&u_his=[^&]*//;
        s/&dt=[^&]*//;
        s/&dtd=[^&]*//;
        s/&lmt=[^&]*//;
        s/(&alternate_ad_url=http%3A%2F%2F[^(%2F)]*)[^&]*/\1/;
        s/(&url=http%3A%2F%2F[^(%2F)]*)[^&]*/\1/;
        s/(&ref=http%3A%2F%2F[^(%2F)]*)[^&]*/\1/;
        s/(&cookie=http%3A%2F%2F[^(%2F)]*)[^&]*/\1/;
        s/[;&?]ord=[?0-9]*//;
        s/[;&]mpvid=[^&;]*//;
        s/&xpc=[^&]*//;
        # yieldmanager
        s/\?clickTag=[^&]*//;
        s/&u=[^&]*//;
        s/&slotname=[^&]*//;
        s/&page_slots=[^&]*//;
        }
        print $x . "http://" . $1 . $2 . $y . "\n";

                        #cache high latency ads
} elsif (m/^http:\/\/(.*?)\/(ads)\?(.*?)/) {
        print $x . "http://" . $1 . "/" . $2  . "\n";

                        #cdn, varialble 1st path
} elsif (($u =~ /filehippo|mediafire/) && (m/^http:\/\/(.*?)\.(.*?)\/(.*?)\/(.*)\.([a-zA-Z0-9]{2,3})(\?.*)?/)) {
        @y = ($1,$2,$4,$5);
        $y[0] =~ s/[a-z]+([0-9]+)?/cdn./;
        print $x . "http://" . $y[0] . $y[1] . "/" . $y[2] . "." . $y[3] . "\n";

} elsif (($u =~ /maxporn/) && (m/^http:\/\/([^\/]*?)\/(.*?)\/([^\/]*?)(\?.*)?$/)) {
        print $x . "http://" . $1 . "/SQUIDINTERNAL/" . $3 . "\n";
       
                        #domain/path/.*/path/filename
} elsif (($u =~ /fucktube/) && (m/^http:\/\/(.*?)(\.[^\.\-]*?[^\/]*\/[^\/]*)\/(.*)\/([^\/]*)\/([^\/\?\&]*)\.([^\/\?\&]{3,4})(\?.*?)$/)) {
        @y = ($1,$2,$4,$5,$6);
        $y[0] =~ s/(([a-zA-Z]+[0-9]+(-[a-zA-Z])?$)|([^\.]*cdn[^\.]*)|([^\.]*cache[^\.]*))/cdn/;
        print $x . "http://" . $y[0] . $y[1] . "/" . $y[2] . "/" . $y[3] . "." . $y[4] . "\n";

#                        #4shared audio/video preview
} elsif (($u =~ /4shared/) && (m/^http:\/\/(.*?)\.(.*?)\/(.*?)\/(dlink__2Fdownload_2F([^\/-]+))([a-zA-Z0-9-]+)\/([^\/\?\&]*\.[^\/\?\&]{2,3})(\?.*)?$/)) {
        @y = ($1,$2,$3,$4,$7);
        $y[0] =~ s/[a-z]+([0-9]+)?/cdn./;
        print $x . "http://" . $y[0] . $y[1] . "/" . $y[2] . "/" . $y[3] . "/" . $y[4] . "\n";

                        #indowebster
} elsif (m/^http:\/\/(.*?)(\.jkt\.3d\.x\.indowebster.com)\/(.*?)\/([^\/\?\&]*)\.([^\/\?\&]{2,4})(\?.*?)$/) {
        @y = ($1,$2,$4,$5);
        $y[0] =~ s/([a-z][0-9][a-z]dlod[\d]{3})|((cache|cdn)[-\d]*)|([a-zA-Z]+-?[0-9]+(-[a-zA-Z]*)?)/cdn/;
        print $x . "http://" . $y[0] . $y[1] . "/" . $y[2] . "." . $y[3] . "\n";

                        #like porn hub variables url and center part of the path, filename etention 3 or 4 with or without ? at the end
} elsif (($u =~ /tube8|pornhub|xvideos/) && (m/^http:\/\/(([A-Za-z]+[0-9-.]+)*?(\.[a-z]*)?)\.([a-z]*[0-9]?\.[^\/]{3}\/[a-z]*)(.*?)((\/[a-z]*)?(\/[^\/]*){4}\.[^\/\?]{3,4})(\?.*)?$/)) {
        print $x . "http://cdn." . $4 . $6 . "\n";

                        #for yimg.com video
} elsif (m/^http:\/\/(.*yimg.com)\/\/(.*)\/([^\/\?\&]*\/[^\/\?\&]*\.[^\/\?\&]{3,4})(\?.*)?$/) {
        print $x . "http://cdn.yimg.com//" . $3 . "\n";
       
                        #for yimg.com doubled
} elsif (m/^http:\/\/(.*?)\.yimg\.com\/(.*?)\.yimg\.com\/(.*?)\?(.*)/) {
        print $x . "http://cdn.yimg.com/"  . $3 . "\n";

                        #for yimg.com with &sig=
} elsif (m/^http:\/\/([^\.]*)\.yimg\.com\/(.*)/) {
        @y = ($1,$2);
        $y[0] =~ s/[a-z]+([0-9]+)?/cdn/;
        $y[1] =~ s/&sig=.*//;
        print $x . "http://" . $y[0] . ".yimg.com/"  . $y[1] . "\n";
                       
                        #youjizz. We use only domain and filename
} elsif (($u =~ /media[0-9]{1,5}\.youjizz/) && (m/^http:\/\/(.*?)(\.[^\.\-]*?\.[^\/]*)\/(.*)\/([^\/\?\&]*)\.([^\/\?\&]{3,4})(\?.*?)$/)) {
        @y = ($1,$2,$4,$5);
        $y[0] =~ s/(([a-zA-Z]+[0-9]+(-[a-zA-Z])?$)|([^\.]*cdn[^\.]*)|([^\.]*cache[^\.]*))/cdn/;
        print $x . "http://" . $y[0] . $y[1] . "/" . $y[2] . "." . $y[3] . "\n";

                        #general purpose for cdn servers. add above your specific servers.
} elsif (m/^http:\/\/([0-9.]*?)\/\/(.*?)\.(.*)\?(.*?)/) {
        print $x . "http://squid-cdn-url//" . $2  . "." . $3 . "\n";

                        #generic http://variable.domain.com/path/filename."ex", "ext" or "exte"
} elsif (m/^http:\/\/(.*?)(\.[^\.\-]*?\..*?)\/([^\?\&\=]*)\.([\w\d]{2,4})\??.*$/) {
        @y = ($1,$2,$3,$4);
        $y[0] =~ s/([a-z][0-9][a-z]dlod[\d]{3})|((cache|cdn)[-\d]*)|([a-zA-Z]+-?[0-9]+(-[a-zA-Z]*)?)/cdn/;
        print $x . "storeurl://" . $y[0] . $y[1] . "/" . $y[2] . "." . $y[3] . "\n";

                        # all that ends with ;
} elsif (m/^http:\/\/(.*?)\/(.*?)\;(.*)/) {
        print $x . "http://" . $1 . "/" . $2  . "\n";

} else {
        print $x . $_ . "\n";
}
}


sumber dari sini