detect natural document boundaries inside one big pdf · splits into separate files · runs locally
splits BEFORE every page where any text item matches · default catches insurance staples
tokens · {base} {n} {start} {end} {title}