Beancount getting started

nugget · October 31, 2019, 11:21am

Hey all,
i finally switched to beancount. i must admit it was not too easy to get stuff running, especially since i started using linux (ubuntu WSL) just at the same time.

now, anybody here has some scripts already made to import postfinance account statements?

would be really nice @hedgehog maybe, if you are still around?

thanks!

Gesk · November 1, 2019, 8:06pm

I cannot help you with postfinance but I would be interested if anybody has scripts for IBKR or revolut. I have some scripts for UBS and cembra credit card. Basically you can easily tune the csv script to work with any csv list, even different ones at the same time. I just had some troubles with the original file encoding depending on the provider.

MeowMeowFuzzyface · November 1, 2019, 11:34pm

Webgui looks awesome ( https://beancount.github.io/fava/ )
Demo: https://fava.pythonanywhere.com/example-with-budgets/income_statement/

I also found another ledger/beancount tool, but this one has a postfinance and interactive brokers parser: https://github.com/sboehler/beans

MeowMeowFuzzyface · November 2, 2019, 12:11am

I found some importer for beancount:

DKB

ING-DiBa

Commerzbank

N26

Tino · November 2, 2019, 8:21am

I’m using beancount for a while and I’m really happy!!
I started with imports, etc…and finished doing passing all to CSV, from there to Numbers or Excel and… adding unique identifiers after each field to copy and paste everything to a power text editor, kind of Visual Code where I format in beancount format.

It’s long and I don’t have time to explain the complete method but is something like that: You have:

10.10.2019. “Buy in COOP” 123 CHF

I put. In excel columns like:

10.10.2019 * ###_### “Buy in COOP” —- Expenses:Groceries 123 CHF

Copy and paste to Visual Code and find all the “special” fields to make a global “enter” and spaces…etc… you edit a line but you are editing the thousand of lines that have your file… I don’t know if is clear.

Takes time, I know, but is much more quicker than finding an importer

MeowMeowFuzzyface · November 2, 2019, 10:37am

You can use this https://sboehler.github.io/beans/ , it provides parsing for:

us.interactivebrokers ( Interactive Brokers )
ch.postfinance ( Postfinance )
ch.supercardplus ( Supercard Plus )

"The importing functionality of beans can of course also be used if you use another plain-text accounting tool. beancount’s syntax is largely compatible with beans. "

MeowMeowFuzzyface · November 3, 2019, 4:31pm

Do we have a beancount expert or python coder here?
I’m new to beancount and need an importer for revolut and cashback creditcards.
Some example files:
https://www.dropbox.com/sh/yrpjw2tvmo8uszk/AAB00B6XAoe2UGWeMOGiGCGwa?dl=0

I pay you a beer or coffee

MeowMeowFuzzyface · November 4, 2019, 5:52pm

@nugget
I posted a parser for postfinance ( https://sboehler.github.io/beans/ ) , works perfect with beancount

With the csv you can parse it and use it with beancount.
You can edit the “examples.rules” for the matching:
…
description =~ “Mintos Marketplace” -> Assets:DE:Mintos:Investments
description =~ “Wohnungsmiete.*” -> Expenses:Home:Rent
…
And parse your csv file with beans:
./beans import -i ch.postfinance -c example.rules -a Assets:CH:Postfinance:Shared postfinance20191102.csv >> myfile.beans

nugget · November 5, 2019, 6:51am

ahahah i never scrolled down that page to realize the export-button. thanks for pointing out!

thanks for posting the command too, i’m quit new to unix and it always takes ages to figure out the next command

nugget · November 16, 2019, 1:00am

… anyone looked at beancount.io and furious.ca? seems to me the author removes some content and moves towards a paid solution. what u think?

Gesk · November 16, 2019, 8:19am

I’m not even sure it’s from the same author:

Beancount.io is not affiliated with beancount · GitHub but uses Fava under MIT license

The source code is still available from the bitbucket repo but the doc on furius.ca seems to be gone, weird.

nugget · November 18, 2019, 5:32pm

yes exactly, i find alot of dead links recently

nugget · November 19, 2019, 6:09pm

Anyone got the

`beancount.price`

stock quote fetching running?

[edit] yes! yahoo works, now working on google:

bean-price -e USD:yahoo/VTI
2019-11-19 price VTI 158.73 USD
bean-price -e USD:google/NYSEARCA:VTI
ERROR : Could not fetch for job: DatedPrice(base=‘NYSEARCA:VTI’, quote=‘USD’, date=None, sources=[PriceSource(module=<module ‘google’ (namespace)>, symbol=‘NYSEARCA:VTI’, invert=False)])

Gesk · November 19, 2019, 6:21pm

I always put both Google and yahoo, I’m not sure if Google is still working.

nugget · November 19, 2019, 6:23pm

could you share your explicit line of code? thanks

Tino · November 20, 2019, 9:46am

Google API is not maintained/available for a while…
Yahoo or Alphavantage are your best alternatives.
I’ve been using Yahoo with Beancount for months without any problem.

Some other ideas:

Tino · November 20, 2019, 11:25am

An advice for people starting with beancount:

Install VS Code, is a free ASCII editor available in all the platforms. Is very well done.
Install a extension named: VSCode-Beancount and follow the instructions about how to organise the files!!!

https://marketplace.visualstudio.com/items?itemName=Lencerf.beancount

You’ll love beancount after that

Tino.

nugget · November 20, 2019, 11:36am

I think I got exactly to this solution - by some chance
no idea on how cruel it is if I started doing stuff with nano (i am new to ubuntu…)

thanks for mentioning

MeowMeowFuzzyface · November 20, 2019, 6:03pm

I got help from the beancounts maillinglist for revolut import.

The code for the importer:

from dateutil.parser import parse
from io import StringIO
from beancount.ingest import importer
from beancount.core import data
from beancount.core import amount
from beancount.core.number import D
from beancount.ingest.importers.mixins import identifier
import csv

  class Importer(identifier.IdentifyMixin,
  importer.ImporterProtocol):
      """An importer for Revolut CSV files."""
  
      def __init__(self, regexps, account, currency):
          identifier.IdentifyMixin.__init__(self, matchers=[
              ('filename', regexps)
          ])
          self.account = account
          self.currency = currency
  
      def name(self):
          return super().name() + self.account
  
      def file_account(self, file):
          return self.account
  
      def extract(self, file, existing_entries):
          entries = []
  
          with StringIO(file.contents()) as csvfile:
              reader = csv.DictReader(csvfile, ['Date',
  'Reference','PaidOut', 'PaidIn', 'ExchangeOut', 'ExchangeIn',
  'Balance', 'Category', 'Notes'], delimiter=';',
  skipinitialspace=True)
              next(reader)
              for row in reader:
                  metakv = {
                      'category': row['Category'].strip(),
                  }
                  exchangeIn = row['ExchangeIn'].strip()
                  exchangeOut = row['ExchangeOut'].strip()
                  if exchangeIn and exchangeOut:
                      metakv['originalIn'] = exchangeIn
                      metakv['originalOut'] = exchangeOut
                  elif exchangeIn:
                      metakv['original'] = exchangeIn
                  elif exchangeOut:
                      metakv['original'] = exchangeOut
  
                  meta = data.new_metadata(file.name, 0, metakv)
                  entry = data.Transaction(
                      meta,
                      parse(row['Date'].strip()).date(),
                      '*',
                      '',
                      (row['Reference'].strip() + ' ' +
  row['Notes'].strip()).strip(),
                      data.EMPTY_SET,
                      data.EMPTY_SET,
                      [
                          data.Posting(self.account,
  amount.Amount(D(row['PaidIn'].strip()) -
  D(row['PaidOut'].strip()), self.currency), None, None, None,
  None),
                      ]
                  )
                  entries.append(entry)
          return entries`

config.py:

`from dateutil.parser import parse
from io import StringIO
from beancount.ingest import importer
from beancount.core import data
from beancount.core import amount
from beancount.core.number import D
from beancount.ingest.importers.mixins import identifier
import csv

  class Importer(identifier.IdentifyMixin,
  importer.ImporterProtocol):
      """An importer for Revolut CSV files."""
  
      def __init__(self, regexps, account, currency):
          identifier.IdentifyMixin.__init__(self, matchers=[
              ('filename', regexps)
          ])
          self.account = account
          self.currency = currency
  
      def name(self):
          return super().name() + self.account
  
      def file_account(self, file):
          return self.account
  
      def extract(self, file, existing_entries):
          entries = []
  
          with StringIO(file.contents()) as csvfile:
              reader = csv.DictReader(csvfile, ['Date',
  'Reference','PaidOut', 'PaidIn', 'ExchangeOut', 'ExchangeIn',
  'Balance', 'Category', 'Notes'], delimiter=';',
  skipinitialspace=True)
              next(reader)
              for row in reader:
                  metakv = {
                      'category': row['Category'].strip(),
                  }
                  exchangeIn = row['ExchangeIn'].strip()
                  exchangeOut = row['ExchangeOut'].strip()
                  if exchangeIn and exchangeOut:
                      metakv['originalIn'] = exchangeIn
                      metakv['originalOut'] = exchangeOut
                  elif exchangeIn:
                      metakv['original'] = exchangeIn
                  elif exchangeOut:
                      metakv['original'] = exchangeOut
  
                  meta = data.new_metadata(file.name, 0, metakv)
                  entry = data.Transaction(
                      meta,
                      parse(row['Date'].strip()).date(),
                      '*',
                      '',
                      (row['Reference'].strip() + ' ' +
  row['Notes'].strip()).strip(),
                      data.EMPTY_SET,
                      data.EMPTY_SET,
                      [
                          data.Posting(self.account,
  amount.Amount(D(row['PaidIn'].strip()) -
  D(row['PaidOut'].strip()), self.currency), None, None, None,
  None),
                      ]
                  )
                  entries.append(entry)
          return entries`

Run with command:
bean-extract config.py Revolut-CHF.myname.csv

Ed_Waadt · April 8, 2020, 3:17pm

Hi!

I know my way well around Linux but use a Windows PC for personal stuff. Have some VMs with Linux though where, among other things, I run Python scripts to update Gnucash’s stock prices database.

I’m looking into transitioning to beancount. My main goal is to be able to edit the bean files easily from my Windows PCs. I’m guessing that’s not a problem per se.

But I also don’t want to waste time setting stuff up. Now that you all seem to have gotten the gist of things and knowing that I know absolutely nothing about beancount, how would you recomend I start? Docker images on Windows? Any tutorial?

I can also spin up a VM just for this, but seems inconvenient. If you find it easier, please let me know as well.

Thanks, guys!