pypdfocr_pdf.py 文件源码

def overlay_hocr_pages(self, dpi, hocr_filenames, orig_pdf_filename):

        logging.debug("Going to overlay following files onto %s" % orig_pdf_filename)
        # Sort the hocr_filenames into natural keys!
        hocr_filenames.sort(key=lambda x: self.natural_keys(x[0] ))
        logging.debug(hocr_filenames)

        pdf_dir, pdf_basename = os.path.split(orig_pdf_filename)
        basename = os.path.splitext(pdf_basename)[0]
        pdf_filename = os.path.join(pdf_dir, "%s_ocr.pdf" % (basename))

        text_pdf_filenames = []
        for img_filename, hocr_filename in hocr_filenames:
            text_pdf_filename = self.overlay_hocr_page(dpi, hocr_filename, img_filename)
            logging.info("Created temp OCR'ed pdf containing only the text as %s" % (text_pdf_filename))
            text_pdf_filenames.append(text_pdf_filename)

        # Now, concatenate this text_pdfs into one single file.
        # This is a hack to save memory/running time when we have to do the actual merge with a writer

        all_text_filename = os.path.join(pdf_dir, "%s_text.pdf" % (basename))
        merger = PdfFileMerger()
        for text_pdf_filename in text_pdf_filenames:
            merger.append(PdfFileReader(file(text_pdf_filename, 'rb')))
        merger.write(all_text_filename)
        merger.close()
    del merger


        writer = PdfFileWriter()
        orig = open(orig_pdf_filename, 'rb')
        text_file = open(all_text_filename, 'rb')

        for orig_pg, text_pg in zip(self.iter_pdf_page(orig), self.iter_pdf_page(text_file)):
            orig_pg = self._get_merged_single_page(orig_pg, text_pg)
            writer.addPage(orig_pg)

        with open(pdf_filename, 'wb') as f:
            # Flush out this page merge so we can close the text_file
            writer.write(f)

        orig.close()
        text_file.close()

        # Windows sometimes locks the temp text file for no reason, so we need to retry a few times to delete
        for fn in text_pdf_filenames:
            #os.remove(fn)
            Retry(partial(os.remove, fn), tries=10, pause=3).call_with_retry() 

        os.remove(all_text_filename)
        logging.info("Created OCR'ed pdf as %s" % (pdf_filename))

        return pdf_filename