This program contains some utility functions for pdf files, e.g. conversion from pdf to jpg, or vice versa.

pdf_utils.py

#=======================================================================#
# pdf_utils.py                                                          #
#=======================================================================#
# usage: pdf_utils.py [-h] [-i INPUT_DIR] [-o OUTPUT_DIR] [-m METHOD]
#
# This program contains some utility functions for pdf files.
#
# optional arguments:
#  -h, --help            show this help message and exit
#  -i INPUT_DIR, --input_dir INPUT_DIR
#                        Input directory for the files to be modified
#  -o OUTPUT_DIR, --output_dir OUTPUT_DIR
#                        Output directory for the files to be modified
#  -m METHOD, --method METHOD
#                        The method that is applied, e.g. for convert
#                        and merge c+m.
#=======================================================================#
#=======================================================================#
# Sample usage:                                                         #
#=======================================================================#
# python pdf_utils.py --input_dir docs --output_dir output --method p2j
# python pdf_utils.py --input_dir output --output_dir output --method p2j
#=======================================================================#

from pdf2image import *
import os
import time
from fpdf import FPDF
import argparse
import glob

arg_keywords = {
    "pdf2jpg": "p2j",
    "jpg2pdf": "j2p"
}


def convert_pdf(file, file_format, dpi):
    if file.endswith(".pdf"):
        file_name = os.path.basename(file).split(".")[0]
        output_file_dir = os.path.join(output_dir, file_name)
        file_path = os.path.join(input_dir, file)
        if not os.path.exists(output_file_dir):
            os.makedirs(output_file_dir)

        with tempfile.TemporaryDirectory() as path:
            images_from_path = convert_from_path(file_path, dpi=dpi, output_folder=path)
            size = len(images_from_path)
            i = 1
            for image in images_from_path:
                prefix = "> " + file + " is now being converted to " + file_format + " format..." + \
                         " [" + str(i) + "/" + str(size) + "]"
                printProgressBar(i, size, prefix=prefix + ' Progress:', suffix='Complete', length=25)
                image.save(os.path.join(output_file_dir, str(i) + file_format))
                i += 1


def create_pdf(file, image_dir):
    image_list = glob.glob(image_dir + '/*.jpg')
    image_list = (sorted(image_list))
    size = len(image_list)
    pdf = FPDF()
    i = 1
    for image in image_list:
        prefix = "> " + os.path.basename(file) + " is now being generated... " + \
                 " [" + str(i) + "/" + str(size) + "]"
        printProgressBar(i, size, prefix=prefix + ' Progress:', suffix='Complete', length=25)
        pdf.add_page()
        pdf.image(image, 0, 0, 210, 297)
        i += 1

    pdf.output(file, "F")
    print("Congrats! " + file.split("/")[-1] + " is created.")


def printProgressBar(iteration, total, prefix='', suffix='', decimals=1, length=100, fill='█'):
    percent = ("{0:." + str(decimals) + "f}").format(100 * (iteration / float(total)))
    filled_length = int(length * iteration // total)
    bar = fill * filled_length + '-' * (length - filled_length)
    print('\r%s |%s| %s%% %s' % (prefix, bar, percent, suffix), end='\r')
    # Print New Line on Complete
    if iteration == total:
        print()


def show_license():
    print("\n# ======================================================================================= #\n"
          "# PDF Utils - Converts pdf files into jpg format, or vice versa.                          #\n" +
          "# Copyright © 2018 - Berk Kaan Kuguoglu                                                   #\n" +
          "# ======================================================================================= #\n" +
          "# usage: pdf_utils.py [-h] [-i INPUT_DIR] [-o OUTPUT_DIR] [-m METHOD]                     #\n" +
          "# ======================================================================================= #\n" +
          "# optional arguments:                                                                     #\n" +
          "#  -h, --help            show this help message and exit                                  #\n" +
          "#  -i INPUT_DIR, --input_dir INPUT_DIR                                                    #\n" +
          "#                        Input directory for the files to be modified                     #\n" +
          "#  -o OUTPUT_DIR, --output_dir OUTPUT_DIR                                                 #\n" +
          "#                        Output directory for the files to be modified                    #\n" +
          "#  -m METHOD, --method METHOD                                                             #\n" +
          "#                        The method that is applied, e.g. p2j denotes 'pdf to jpg'.       #\n" +
          "# ======================================================================================= #\n")


if __name__ == '__main__':
    show_license()
    parser = argparse.ArgumentParser(description="This program contains some utility functions for pdf files.")
    parser.add_argument("-i", "--input_dir", help="Input directory for the files to be modified")
    parser.add_argument("-o", "--output_dir", help="Output directory for the files to be modified")
    parser.add_argument("-m", "--method", help="The method that is applied, e.g. for convert and merge c+m.")
    args = parser.parse_args()
    input_dir = args.input_dir
    output_dir = args.output_dir
    method = args.method

    if arg_keywords["pdf2jpg"] in method:
        for file in sorted(os.listdir(input_dir)):
            if file == ".DS_Store":
                continue
            start = time.time()
            convert_pdf(file=file, file_format=".jpg", dpi=300)
            finish = time.time()
            print("It took " + str(finish - start) + " seconds to convert " + file + " into .jpg format.")

    elif arg_keywords["jpg2pdf"] in method:
        for file in sorted(os.listdir(input_dir)):
            if file == ".DS_Store":
                continue
            start = time.time()
            image_dir = os.path.join(output_dir, file)
            pdf_name = file + ".pdf"
            pdf_path = os.path.join(image_dir, pdf_name)
            create_pdf(pdf_path, image_dir)
            finish = time.time()
            print("It took " + str(finish-start) + " seconds to merge the document " + pdf_name + ".")