$complete = $schemearr["scheme"]."://".$schemearr["host"];
- if ($schemearr["port"] != "")
+ if (@$schemearr["port"] != "")
$complete .= ":".$schemearr["port"];
if(strpos($urlarr['path'],'/') !== 0)
$complete .= $urlarr["path"];
- if ($urlarr["query"] != "")
+ if (@$urlarr["query"] != "")
$complete .= "?".$urlarr["query"];
- if ($urlarr["fragment"] != "")
+ if (@$urlarr["fragment"] != "")
$complete .= "#".$urlarr["fragment"];
return($complete);
}
-function parseurl_getsiteinfo($url) {
+function parseurl_getsiteinfo($url, $no_guessing = false) {
$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_USERAGENT,'Opera/9.64(Windows NT 5.1; U; de) Presto/2.1.1');
+ //curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1);
+ curl_setopt($ch,CURLOPT_USERAGENT,' Mozilla/5.0 (Macintosh; Intel Mac OS X 10.8; rv:24.0) Gecko/20100101 Firefox/24.0');
$header = curl_exec($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);
+ }
+
+ require_once("include/oembed.php");
+
+ $oembed_data = oembed_fetch_url($url);
+
+ if ($oembed_data->type == "link") {
+ if (isset($oembed_data->title))
+ $siteinfo["title"] = $oembed_data->title;
+ if (isset($oembed_data->description))
+ $siteinfo["text"] = $oembed_data->description;
+ if (isset($oembed_data->thumbnail_url))
+ $siteinfo["image"] = $oembed_data->thumbnail_url;
+ }
+
// Fetch the first mentioned charset. Can be in body or header
if (preg_match('/charset=(.*?)['."'".'"\s\n]/', $header, $matches))
$charset = trim(array_pop($matches));
$xpath = new DomXPath($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 $node)
}
}
- if ($siteinfo["image"] == "") {
+ if ((@$siteinfo["image"] == "") AND !$no_guessing) {
$list = $xpath->query("//img[@src]");
foreach ($list as $node) {
$attr = array();
"height"=>$photodata[1]);
}
- if ($siteinfo["text"] == "") {
+ if ((@$siteinfo["text"] == "") AND (@$siteinfo["title"] != "") AND !$no_guessing) {
$text = "";
$list = $xpath->query("//div[@class='article']");
$str_tags = '';
$textmode = false;
- if(local_user() && intval(get_pconfig(local_user(),'system','plaintext')))
- $textmode = true;
if(local_user() && (! feature_enabled(local_user(),'richtext')))
$textmode = true;