I uploaded a PDF to archive.org that auto-OCRs with plenty of mistakes. I have found no way of updating the entire stack of documents produced. I wonder if Project Gutenberg is similar