ABN AMRO Bank in the Netherlands lets you export your bank statements to plain text (.TAB) files. This importer (beancount-abnamro) helps you convert them to beancount transactions that you can append to your ledger.
This module isn't available on PyPI yet so you'll have to install in manually. A sample statement and configuration file is included in the repository.
Assuming you already have beancount installed and have configured it to use this importer, you should be able to run bean-extract.
bean-extract config.py TXT231209180908.TAB
(Don't forget to change TXT231209180908.TAB to the filename of the statement you are trying to import)
Under the lib/ folder of this repository, you'll find a file called vendor_map.py. This file contains a dictionary that maps the name of the vendor to the account. You'll very likely want to edit this file to use your own expense accounts and modify vendor names.
vendor_map.py
vendor_map = {
"KLM": "Expenses:Travel:Transport",
"Spotify AB by Adyen": "Expenses:Subscriptions:Spotify",
"Tokyo Ramen Iki": "Expenses:Eatout",
}
You could also leave the dictionary bank if you prefer to balance your books by hand. I personally prefer to map the most common transactions automatically and then use fava to look deeper.
You'll also see a file called bank_utils.py. It tries to parse transaction descriptions that ABN AMRO uses to a sensible, beancount friendly payee and narration. Each year I come across certain transaction descriptions (let's say around 1 in 1000) that defy all existing parsing logic. You can customize the unparsable_name and unparsable_description methods to handle those. By default these methods simply return whatever you pass to them.