+/* To-Do
+https://developers.google.com/+/plugins/snippet/
+
+<meta itemprop="name" content="Toller Titel">
+<meta itemprop="description" content="Eine tolle Beschreibung">
+<meta itemprop="image" content="http://maple.libertreeproject.org/images/tree-icon.png">
+
+<body itemscope itemtype="http://schema.org/Product">
+ <h1 itemprop="name">Shiny Trinket</h1>
+ <img itemprop="image" src="{image-url}" />
+ <p itemprop="description">Shiny trinkets are shiny.</p>
+</body>
+*/
+
+if(!function_exists('deletenode')) {
+ function deletenode(&$doc, $node)
+ {
+ $xpath = new DomXPath($doc);
+ $list = $xpath->query("//".$node);
+ foreach ($list as $child)
+ $child->parentNode->removeChild($child);
+ }
+}
+
+function completeurl($url, $scheme) {
+ $urlarr = parse_url($url);
+
+ if (isset($urlarr["scheme"]))
+ return($url);
+
+ $schemearr = parse_url($scheme);
+
+ $complete = $schemearr["scheme"]."://".$schemearr["host"];
+
+ if (@$schemearr["port"] != "")
+ $complete .= ":".$schemearr["port"];
+
+ if(strpos($urlarr['path'],'/') !== 0)
+ $complete .= '/';
+
+ $complete .= $urlarr["path"];
+
+ if (@$urlarr["query"] != "")
+ $complete .= "?".$urlarr["query"];
+
+ if (@$urlarr["fragment"] != "")
+ $complete .= "#".$urlarr["fragment"];
+
+ return($complete);
+}
+
+function parseurl_getsiteinfo($url) {
+ $siteinfo = array();
+ $ch = curl_init();
+ curl_setopt($ch, CURLOPT_URL, $url);
+ curl_setopt($ch, CURLOPT_HEADER, 1);
+ curl_setopt($ch, CURLOPT_NOBODY, 0);
+ curl_setopt($ch, CURLOPT_TIMEOUT, 3);
+ curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
+ //curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1);
+ curl_setopt($ch,CURLOPT_USERAGENT,'Opera/9.64(Windows NT 5.1; U; de) Presto/2.1.1');
+
+ $header = curl_exec($ch);
+ $curl_info = @curl_getinfo($ch);
+ $http_code = $curl_info['http_code'];
+ curl_close($ch);
+
+ if ((($curl_info['http_code'] == "301") OR ($curl_info['http_code'] == "302"))
+ AND (($curl_info['redirect_url'] != "") OR ($curl_info['location'] != ""))) {
+ if ($curl_info['redirect_url'] != "")
+ $siteinfo = parseurl_getsiteinfo($curl_info['redirect_url']);
+ else
+ $siteinfo = parseurl_getsiteinfo($curl_info['location']);
+ return($siteinfo);
+ }
+
+ // Fetch the first mentioned charset. Can be in body or header
+ if (preg_match('/charset=(.*?)['."'".'"\s\n]/', $header, $matches))
+ $charset = trim(array_pop($matches));
+ else
+ $charset = "utf-8";
+
+ $pos = strpos($header, "\r\n\r\n");
+
+ if ($pos)
+ $body = trim(substr($header, $pos));
+ else
+ $body = $header;
+
+ $body = mb_convert_encoding($body, "UTF-8", $charset);
+ $body = mb_convert_encoding($body, 'HTML-ENTITIES', "UTF-8");
+
+ $doc = new DOMDocument();
+ @$doc->loadHTML($body);
+
+ deletenode($doc, 'style');
+ deletenode($doc, 'script');
+ deletenode($doc, 'option');
+ deletenode($doc, 'h1');
+ deletenode($doc, 'h2');
+ deletenode($doc, 'h3');
+ deletenode($doc, 'h4');
+ deletenode($doc, 'h5');
+ deletenode($doc, 'h6');
+ deletenode($doc, 'ol');
+ deletenode($doc, 'ul');
+
+ $xpath = new DomXPath($doc);
+
+ $list = $xpath->query("//meta[@content]");
+ foreach ($list as $node) {
+ $attr = array();
+ if ($node->attributes->length)
+ foreach ($node->attributes as $attribute)
+ $attr[$attribute->name] = $attribute->value;
+
+ if (@$attr["http-equiv"] == 'refresh') {
+ $path = $attr["content"];
+ $pathinfo = explode(";", $path);
+ $content = "";
+ foreach ($pathinfo AS $value) {
+ if (substr(strtolower($value), 0, 4) == "url=")
+ $content = substr($value, 4);
+ }
+ if ($content != "") {
+ $siteinfo = parseurl_getsiteinfo($content);
+ return($siteinfo);
+ }
+ }
+ }
+
+ //$list = $xpath->query("head/title");
+ $list = $xpath->query("//title");
+ foreach ($list as $node)
+ $siteinfo["title"] = html_entity_decode($node->nodeValue, ENT_QUOTES, "UTF-8");
+
+ //$list = $xpath->query("head/meta[@name]");
+ $list = $xpath->query("//meta[@name]");
+ foreach ($list as $node) {
+ $attr = array();
+ if ($node->attributes->length)
+ foreach ($node->attributes as $attribute)
+ $attr[$attribute->name] = $attribute->value;
+
+ $attr["content"] = html_entity_decode($attr["content"], ENT_QUOTES, "UTF-8");
+
+ switch (strtolower($attr["name"])) {
+ case "fulltitle":
+ $siteinfo["title"] = $attr["content"];
+ break;
+ case "description":
+ $siteinfo["text"] = $attr["content"];
+ break;
+ case "dc.title":
+ $siteinfo["title"] = $attr["content"];
+ break;
+ case "dc.description":
+ $siteinfo["text"] = $attr["content"];
+ break;
+ }
+ }