Scan double sided document use single side ADF to PDF

scan.py

import subprocess
import tempfile
import argparse
import shutil
import os.path
import glob

from PyPDF2 import PdfFileMerger, PdfFileReader

DEVICE=""

if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='Batch Scan')
    parser.add_argument('-c', '--color', action='store_true')
    parser.add_argument('-d', '--double', action='store_true')
    parser.add_argument('-o', '--output', type=str, default='output.pdf')
    parser.add_argument('--no-ocr',  action='store_true')
    args = parser.parse_args()
    tmp = tempfile.mkdtemp()
    def scan_page(side='f'):
        cmd = ['scanimage', '-d', DEVICE, '--resolution', '300',
            '--source', 'ADF', '--batch=%s/%s%%d.pnm' % (tmp, side),
            '--batch-start=1000',
            '-y', '279.4',
            '--mode', 'Color' if args.color else 'Gray']
        subprocess.check_call(cmd)
    scan_page()
    if args.double:
        raw_input('Turn your pages')
        scan_page('b')
    for fn in glob.glob(os.path.join(tmp, '*.pnm')):
        if args.no_ocr:
            subprocess.check_call(['convert', fn,
                '-density', '300x300',
                '-quality', '75',
                '-compress', 'jpeg',
                fn[:-4]+'.pdf'])
        else:
            subprocess.check_call([
                'tesseract',
                fn, fn[:-4],
                '/usr/share/tessdata/configs/pdf'])
    pdfs = sorted(glob.glob(os.path.join(tmp, '*.pdf')))
    sheets = len(pdfs) / 2
    if args.double:
        sorted_pdfs = [x for t in zip(pdfs[sheets:], reversed(pdfs[:sheets])) for x in t]
    else:
        sorted_pdfs = pdfs
    merger = PdfFileMerger()
    for fn in sorted_pdfs:
        merger.append(PdfFileReader(file(fn, 'rb')))

    merger.write(args.output)
    shutil.rmtree(tmp)