pdf 重排

pdf.py

#!/usr/bin/env python
import sys
try:
    from PyPDF2 import PdfFileReader, PdfFileWriter
except ImportError:
    from pyPdf import PdfFileReader, PdfFileWriter
import os

def concatenate(input_files, output_file):
    input_streams = []
    output_stream = open(output_file, 'wb')
    try:
        for input_file in input_files:
            input_streams.append(open(input_file, 'rb'))
        writer = PdfFileWriter()
        for reader in map(PdfFileReader, input_streams):
            for n in range(reader.getNumPages()):
                writer.addPage(reader.getPage(n))
        writer.write(output_stream)
    finally:
        for f in input_streams:
            f.close()
        output_stream.close()


def reverse(input_file, output_file):
    with open(input_file, 'rb') as fin, open(output_file, 'wb') as fout:
        reader = PdfFileReader(fin)
        writer = PdfFileWriter()
        for i in range(reader.getNumPages()-1, -1, -1):
            writer.addPage(reader.getPage(i))
        writer.write(fout)


def cross(fileA, fileB, output_file):
    with open(fileA, 'rb') as fA, open(fileB, 'rb') as fB, open(output_file, 'wb') as fout:
        reader_A = PdfFileReader(fA)
        reader_B = PdfFileReader(fB)
        if reader_A.getNumPages() != reader_B.getNumPages():
            raise ValueError("Pages should be equal, now {} != {}".format(reader_A.getNumPages(), reader_B.getNumPages()))
        
        writer = PdfFileWriter()
        for i in range(reader_A.getNumPages()):
            writer.addPage(reader_A.getPage(i))
            writer.addPage(reader_B.getPage(i))
        writer.write(fout)

def double_scan_pdf_merge(front_side_pdfs, reverse_side_pdfs, output_file):
    """ 双面扫描的PDF进行合并

        Param:
            input_file_tuples: tuple, or list of tuple. format (front_side_pdf, reverse_side_pdf). 
                If the scan is split into sections, passed in order as list. 
            output_file: the output file
    """
    if type(front_side_pdfs) == str:
        front_side_pdfs = [front_side_pdfs]
    if type(reverse_side_pdfs) == str:
        reverse_side_pdfs = [reverse_side_pdfs]    
    concatenate(front_side_pdfs, 'tmp1.pdf')
    concatenate(reverse_side_pdfs, 'tmp2.pdf')
    reverse('tmp2.pdf', 'tmp3.pdf')
    cross('tmp1.pdf', 'tmp3.pdf', output_file)
    os.remove('tmp1.pdf')
    os.remove('tmp2.pdf')
    os.remove('tmp3.pdf')


if __name__ == '__main__':
    double_scan_pdf_merge('a.pdf', 'b.pdf', 'output.pdf')