The docs for PDF::Extract seem to indicate that it just pulls pages out of an existing PDF document and creates new PDF documents out of these subsets.