I wonder if there's a possibility, using langsec(language based security) soomething similar to build a generic tool to solve this stuff once and for all ?