Does it work on scanned (non-“digital”) PDFs as well? That has been my biggest pain point so far. Haven’t tried Koreader yet.