All but one tests pass, test_input_file_not_a_pdf Not sure if PyPDF2 metadata generation will mangle the first page.