* Functions used to protect against Cross-Site Request Forgery
* The security token has to base on at least one value that an attacker can't know - here it's the session ID and the private key.
* In this implementation, a security token is reusable (if the user submits a form, goes back and resubmits the form, maybe with small changes;
- * or if the security token is used for ajax-calls that happen several times), but only valid for a certain amout of time (3hours).
- * The "typename" seperates the security tokens of different types of forms. This could be relevant in the following case:
- * A security token is used to protekt a link from CSRF (e.g. the "delete this profile"-link).
+ * or if the security token is used for ajax-calls that happen several times), but only valid for a certain amount of time (3hours).
+ * The "typename" separates the security tokens of different types of forms. This could be relevant in the following case:
+ * A security token is used to protect a link from CSRF (e.g. the "delete this profile"-link).
* If the new page contains by any chance external elements, then the used security token is exposed by the referrer.
- * Actually, important actions should not be triggered by Links / GET-Requests at all, but somethimes they still are,
+ * Actually, important actions should not be triggered by Links / GET-Requests at all, but sometimes they still are,
* so this mechanism brings in some damage control (the attacker would be able to forge a request to a form of this type, but not to forms of other types).
*/
public static function getFormSecurityToken($typename = '')
$a = DI::app();
$timestamp = time();
- $sec_hash = hash('whirlpool', $a->user['guid'] . $a->user['prvkey'] . session_id() . $timestamp . $typename);
+ $sec_hash = hash('whirlpool', ($a->user['guid'] ?? '') . ($a->user['prvkey'] ?? '') . session_id() . $timestamp . $typename);
return $timestamp . '.' . $sec_hash;
}