I think I had something, not a parser but replacer, can't find now, it helped me when I was building xmls for pdf