extract_pdf2png.py

from os import path
import os
from io import BytesIO
from wand.image import Image
from PyPDF2 import PdfFileWriter, PdfFileReader


base_dir = path.dirname(__file__)
data_dir = path.join(base_dir, '..', 'data')
asset_dir = path.join(base_dir, '..', 'out')

filenames = os.listdir(data_dir)

for filename in filenames:
    name, ext = path.splitext(filename)

    result_dir = path.join(asset_dir, name)
    if not path.isdir(result_dir):
        os.mkdir(result_dir)

    filepath = path.join(data_dir, filename)

    original_pdf = PdfFileReader(open(filepath, 'rb'))

    for i in range(original_pdf.numPages):
        pdf_file = BytesIO()
        new_pdf = PdfFileWriter()
        new_pdf.addPage(original_pdf.getPage(i))
        new_pdf.write(pdf_file)
        pdf_file.seek(0)
        
        with Image(file=pdf_file, resolution=300) as image:
            image_path = path.join(result_dir, '{0:03d}.png'.format(i))
            image.save(filename=image_path)