protected function parseInternal($URL_string)
{
$URL_map = @parse_URL($URL_string);
if (!$URL_map || !$URL_map['host'] || !preg_match('/[a-zA-Z0-9._-]*[a-zA-Z0-9]$/', $URL_map['host'])) {
return false;
}
$URL_map = array_map('trim', $URL_map);
$this->parsedURL = $URL_string;
$this->scheme = isset($URL_map['scheme']) ? $URL_map['scheme'] : 'http';
$this->host = $URL_map['host'];
$this->port = isset($URL_map['port']) ? (int) $URL_map['port'] : $this->scheme == 'https' ? 443 : 80;
$this->path = isset($URL_map['path']) ? $URL_map['path'] : '';
if (isset($URL_map['query'])) {
parse_str($URL_map['query'], $this->query);
}
if (!$this->query) {
$this->query = array();
}
if ($this->path == '') {
$this->path = '/';
}
$this->path .= isset($URL_map['query']) ? "?{$URL_map['query']}" : '';
isset($URL_map['fragment']) and $this->path .= '#' . $URL_map['fragment'];
return true;
}
function validate_URL($URL)
{
$URL_components = parse_URL($URL);
if (!isset($URL_components['scheme']) || !isset($URL_components['host'])) {
return false;
}
//explode our host name on '.'s and reattach the last two with a dot in between... voila, a domain
$domain_components = explode('.', $URL_components['host']);
$domain = $domain_components[count($domain_components) - 2] . '.' . $domain_components[count($domain_components) - 1];
if (in_array($domain, $this->trusted_domains) && ($URL_components['scheme'] == 'http' || $URL_components['scheme'] == 'https')) {
return true;
} else {
return false;
}
}
/**
* Header the browser to the current location of the most recent page
* that occupied a given URL
*
* How it works:
*
* 1. Looks for the URL in the URL_history table.
*
* 2. If there is no URL, send a 404 header.
* If there are URLs, send a 301 header and a Location header to the
* location of the live page that most recent inhabited that URL.
*
* Important: Because it may attempt to header the client to a different URL,
* this method must be called before any output is started, or in the context
* of output buffering.
*
* @param string $request_uri a URL relative to the host root (e.g. /foo/bar/)
* @return NULL
*
* @todo modify to make multidomain safe
*/
function check_URL_history($request_uri)
{
$url_arr = parse_URL($request_uri);
// This catches links that might not have had a trailing slash
// pages always have a trailing slash in the db
$URL = '/' . trim_slashes($url_arr['path']) . '/';
$URL = str_replace('//', '/', $URL);
$query_string = !empty($url_arr['query']) ? '?' . $url_arr['query'] : '';
$query = 'SELECT * FROM URL_history WHERE url ="' . reason_sql_string_escape($URL) . '" ORDER BY timestamp DESC';
$results = db_query($query);
$num_results = mysql_num_rows($results);
if (mysql_num_rows($results) > 0) {
while ($row = mysql_fetch_array($results)) {
$page_id = $row['page_id'];
$page = new entity($page_id);
if (reason_is_entity($page, 'minisite_page') && $page->get_value('state') == 'Live' && ($redir = @reason_get_page_url($page))) {
if ($redir == $request_uri) {
//Could potentially update rewrites here, solving most times this happens, perhaps.
trigger_error("A page should exist here, but apparently does not at the moment. A web administrator may need to run URL updating on this site.");
} else {
header('Location: ' . $redir . $query_string, true, 301);
exit;
}
}
}
}
// if we have gotten this far and not found a URL lets send a 404
http_response_code(404);
}