@dakdad dompdf is focused on converting HTML, so the api is sparse; for that level of control you might try the bundled CPDF library