Skip to content

Instantly share code, notes, and snippets.

@gologius
Last active April 14, 2019 09:29
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save gologius/3c6a207f2a0b56242f636fb7ab3bb472 to your computer and use it in GitHub Desktop.
Save gologius/3c6a207f2a0b56242f636fb7ab3bb472 to your computer and use it in GitHub Desktop.
PDFをZIPに変換します。PDFファイル名を引数として複数指定でき、1PDFにつき1ZIPファイル生成します。
# -*- coding: utf-8 -*-
"""
Created on Sun Apr 14 12:26:28 2019
@author: gologius
"""
import datetime
import os
import pdf2image
import sys
import shutil
import tempfile
def check_file(filepath):
print("check file: "+ filepath)
root, ext = os.path.splitext(filepath)
if ext != '.pdf' or filepath=='':
return False
return True
def pdf2zip(filepath):
print('begin pdf to img')
with tempfile.TemporaryDirectory() as tmppath:
#なぜかoutput_foloderを指定して、一回ppmファイルとして出力しないと、PNGファイルとして出力できない
pages = pdf2image.convert_from_path(filepath, dpi=200, output_folder=tmppath)
basepath = os.path.dirname(filepath)
dirpath = basepath + "\\" + datetime.datetime.now().strftime("%Y%m%d_%H%M%S")
os.mkdir(dirpath)
print("make directory: " + dirpath)
for idx, page in enumerate(pages):
filename = '{0:04d}'.format(idx) + ".jpg"
page.save(dirpath + "\\" +filename, 'JPEG')
print(filename + ' ' + str(idx) + "/" + str(len(pages)))
print('end pdf to img')
#圧縮
print('begin img to zip')
shutil.make_archive(dirpath, 'zip', root_dir=dirpath)
shutil.rmtree(dirpath) #不要フォルダを削除
print('end img to zip')
def main():
if len(sys.argv) == 1:
print('引数にファイルパスが指定されていません')
return
for i,path in enumerate(sys.argv):
if i==0:
continue #0番目は引数ではないのでスキップする
result = check_file(path)
if result == False:
print('不正なファイルです')
continue
pdf2zip(path)
return
main()
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment