/**
* Check if this URL is a redirect and return redir info.
*
* Most code should call File_redirection::where instead, to check if we
* already know that redirection and avoid extra hits to the web.
*
* The URL is hit and any redirects are followed, up to 10 levels or until
* a protected URL is reached.
*
* @param string $in_url
* @return mixed one of:
* string - target URL, if this is a direct link or can't be followed
* array - redirect info if this is an *unknown* redirect:
* associative array with the following elements:
* code: HTTP status code
* redirects: count of redirects followed
* url: URL string of final target
* type (optional): MIME type from Content-Type header
* size (optional): byte size from Content-Length header
* time (optional): timestamp from Last-Modified header
*/
public function lookupWhere($short_url, $redirs = 10, $protected = false)
{
if ($redirs < 0) {
return false;
}
if (strpos($short_url, '://') === false) {
return $short_url;
}
try {
$request = self::_commonHttp($short_url, $redirs);
// Don't include body in output
$request->setMethod(HTTP_Request2::METHOD_HEAD);
$response = $request->send();
if (405 == $response->getStatus() || 204 == $response->getStatus()) {
// HTTP 405 Unsupported Method
// Server doesn't support HEAD method? Can this really happen?
// We'll try again as a GET and ignore the response data.
//
// HTTP 204 No Content
// YFrog sends 204 responses back for our HEAD checks, which
// seems like it may be a logic error in their servers. If
// we get a 204 back, re-run it as a GET... if there's really
// no content it'll be cheap. :)
$request = self::_commonHttp($short_url, $redirs);
$response = $request->send();
}
} catch (Exception $e) {
// Invalid URL or failure to reach server
common_log(LOG_ERR, "Error while following redirects for {$short_url}: " . $e->getMessage());
return $short_url;
}
if ($response->getRedirectCount() && File::isProtected($response->getUrl())) {
// Bump back up the redirect chain until we find a non-protected URL
return self::lookupWhere($short_url, $response->getRedirectCount() - 1, true);
}
$ret = array('code' => $response->getStatus(), 'redirects' => $response->getRedirectCount(), 'url' => $response->getUrl());
$type = $response->getHeader('Content-Type');
if ($type) {
$ret['type'] = $type;
}
if ($protected) {
$ret['protected'] = true;
}
$size = $response->getHeader('Content-Length');
// @fixme bytes?
if ($size) {
$ret['size'] = $size;
}
$time = $response->getHeader('Last-Modified');
if ($time) {
$ret['time'] = strtotime($time);
}
return $ret;
}
function _redirectWhere_imp($short_url, $redirs = 10, $protected = false)
{
if ($redirs < 0) {
return false;
}
// let's see if we know this...
$a = File::staticGet('url', $short_url);
if (!empty($a)) {
// this is a direct link to $a->url
return $a->url;
} else {
$b = File_redirection::staticGet('url', $short_url);
if (!empty($b)) {
// this is a redirect to $b->file_id
$a = File::staticGet('id', $b->file_id);
return $a->url;
}
}
$curlh = File_redirection::_commonCurl($short_url, $redirs);
// Don't include body in output
curl_setopt($curlh, CURLOPT_NOBODY, true);
curl_exec($curlh);
$info = curl_getinfo($curlh);
curl_close($curlh);
if (405 == $info['http_code']) {
$curlh = File_redirection::_commonCurl($short_url, $redirs);
curl_exec($curlh);
$info = curl_getinfo($curlh);
curl_close($curlh);
}
if (!empty($info['redirect_count']) && File::isProtected($info['url'])) {
return File_redirection::_redirectWhere_imp($short_url, $info['redirect_count'] - 1, true);
}
$ret = array('code' => $info['http_code'], 'redirects' => $info['redirect_count'], 'url' => $info['url']);
if (!empty($info['content_type'])) {
$ret['type'] = $info['content_type'];
}
if ($protected) {
$ret['protected'] = true;
}
if (!empty($info['download_content_length'])) {
$ret['size'] = $info['download_content_length'];
}
if (isset($info['filetime']) && $info['filetime'] > 0) {
$ret['time'] = $info['filetime'];
}
return $ret;
}
