basic_merging.py

#!/usr/bin/env python
"""
Sample code that demonstrates merging together three PDFs into one, picking and choosing which pages appear in which order.
Selected pages can be added to the end of the output PDF being built, or inserted in the middle.
This example takes input from the command line.
"""
from __future__ import print_function

from os import pardir
from os.path import abspath, dirname, join
from sys import argv, path

from pypdf import PdfFileMerger, PdfFileReader

SAMPLE_CODE_ROOT = dirname(__file__)
SAMPLE_PDF_ROOT = join(SAMPLE_CODE_ROOT, "pdfsamples")

path.append(abspath(join(SAMPLE_CODE_ROOT, pardir)))


FLAG_HELP = {"-h", "--help"}
USAGE = """\
Merges three PDF documents input from the command line.

%(progname)s: <PDF 1> <PDF 2> <PDF 3> [output filename]
%(progname)s: [-h | --help]
""" % {
    "progname": argv[0]
}


def main():
    requiredPages = 3
    output = "PyPDF-Merging-Output.pdf"

    if set(argv) & FLAG_HELP:
        print(USAGE)
        exit(0)
    elif len(argv) < 4:
        print(USAGE)
        exit(1)
    else:
        files = [f.strip() for f in argv[1:4]]

        if len(argv) > 4:
            output = argv[4].strip()

    reader1 = PdfFileReader(files[0])
    merger = PdfFileMerger(open(output, "wb"))

    if reader1.numPages < requiredPages:
        print(
            "File 1 requires %d pages, but it has just %d"
            % (requiredPages, reader1.numPages)
        )
        exit(1)

    input1 = open(files[0], "rb")
    input2 = open(files[1], "rb")
    input3 = open(files[2], "rb")

    # Add the first 3 pages of input1 to output
    merger.append(fileobj=input1, pages=(0, 3))

    # Insert the first page of input2 into the output beginning after the
    # second page
    merger.merge(position=2, fileobj=input2, pages=(0, 1))

    # Append entire input3 document to the end of the output document
    merger.append(input3)

    merger.write()
    print("Output successfully written to", output)

    merger.close()


if __name__ == "__main__":
    main()