the problem with writing your own parser is that it will have to be adjusted when we add new things to schema