Bank PDF Scraper in Python

• tested with MX-Linux 19 (Debian10-Derivat), python 3.7.3
• Ubuntu 18.04, python 3.6.9

required packages

• python3-selenium
• chromedriver
• chromium

init / design

• currently supports only comdirect, but extendable for other banks
• bank-specific config and logic via /configs/bank.py
  • see configs/comdirect.py
• bank-specific config imports user-specific configuration via /configs/bank_config.py
  • see sample config configs/comdirect_config.py.sample

  login = '012345678'
  password = '123456'
  
  downloads = '/home/user/Downloads'
  file_head = '/home/user/comdirect_head.txt'
  archive = False
  close = False
  
  

usage

• call it in terminal via

  # non-interactive without the -i
  python3 -i scraper.py
  

• opens browser, inserts login and password

• 2 Factor authentication is needed for comdirect - use photoTAN in browser

• different usage modes

• closes browser if bank-config states close = True

usage modes

default

• all PDFs in inbox will be downloaded, except links containing "Termingebundenes" which sometimes cause bugs due to html-links
• you can stop the script at any time via ctrl + d

archive

• you can download the whole archive through the configuration

  archive = True
  

delta

• usually you just want to download all new files, therefore the scraper needs to store the last downloaded file
• this is done through the configuration

  file_head = '/home/user/comdirect_head.txt'
  

• that file will contain the latest doc-id as headline and it's name as second line and will be used as reference
  • all files will be downloaded until that file is reached
  • afterwards the file will be updated and the previous file is stored as backup (and will be overridden on next succesful run)

interactive

• not needed for delta mode

tests

python3 -m doctest -v configs/comdirect.py