Setup a bot on gotosocial

First, create a user

Because I closed registration this has to be done via the cli. Commands for my instance running on
# create account
/app/gotosocial --config-path /app/config.yaml admin account create --username some_username --email --password 'SOME_PASSWORD'

# activate account
/app/gotosocial --config-path /app/config.yaml admin account confirm --username some_username

Setup Account

Most settings of an account can be changed via the user panel in gotosocial. But not the bot=true setting.
For automation we need an app anyway, so lets create an app and use it to change user settings.

To get the accesstoken the easiest way is to use this site: Set the scope to write!
For the following commands we only need the access_token and not the client_id/client_secret.
The command to set bot=true, locked=false, display_name and note:
curl -X PATCH -H 'Authorization: Bearer ACCESS_TOKEN' -F 'bot=true' -F 'locked=false' -F 'display_name=test user' -F 'note=some note'

Send a toot

Three ways to send a status: curl, httpie or python-requests.

Please be aware of the visibility rules on your server!
Some servers don't want bots on the local timeline, so use unlisted instead of public.
My examples here are set to private. This is only to show how posting a status works.


curl -H 'Authorization: Bearer ACCESS_TOKEN' -F 'status=toot via curl (private)' -F 'visibility=private'


http POST 'Authorization: Bearer ACCESS_TOKEN' 'status=toot via httpie (private)' 'visibility=private'


import requests
        "Authorization": "Bearer ACCESS_TOKEN"
    json={"status": "toot via requests (private)", "visibility": "private"},

The same code should work with httpx: replace requests with httpx.

ING DiBa csv to ledger converter

Because the fints api of ing diba is down for a while now I switched to parse the csv export. Before I used fints and this worked perfectly for years.

My new solution is only converting from csv to ledger -- no additional magic.

The contents of the csv begins with a non-csv part and the first line of actual csv contains the header of the rows below.
One line of csv looks like this:
26.08.2022;26.08.2022;VISA ALDI SUED;Lastschrift;NR XXXX YYYY STUTTGART KAUFUMSATZ 24.08 ...;100,00;EUR;-12,34;EUR
And I want an output like this:
2022-08-26 FIXME
    Expenses:FIXME                             €12.34

After conversion I copy the entries into a ledger file per month and manually add the accounts I want this entries booked into.

First working solution:
import csv
import io
from datetime import datetime
from decimal import Decimal
from pathlib import Path

import click

def csv2ledger(filename):
    # find start of csv and read csv data
    start_of_csv = -1
    data = []
    with open(filename, "r", encoding="cp1252") as fp:
        for index, line in enumerate(fp.readlines()):
            if start_of_csv < 0 and line.startswith("Buchung;Valuta"):
                start_of_csv = index
            if start_of_csv > 0 and index >= start_of_csv:

    # add header; but reverse rows: newest at the end
    csv_data = io.StringIO(data[0] + "\n".join(reversed(data[1:])))
    for row in csv.DictReader(csv_data, delimiter=";"):
        yield row

def row2ledger(row):
    currency = "€" if row["Währung"] == "EUR" else row["Währung"]
    # convert amount from German to international numbers
    amount = Decimal(row["Betrag"].replace(".", "").replace(",", "."))
    # remove '.00'
    amount = int(amount) if amount == int(amount) else amount
    # reverse sign
    amount = f"{currency}{amount * -1}"
    dt = datetime.strptime(row["Valuta"], "%d.%m.%Y").date()
    details = " | ".join(
            row.get(i) or ""
            for i in ["Buchungstext", "Auftraggeber/Empfänger", "Verwendungszweck"]
    return (
        "# {}\n".format(details.strip())
        + f"{dt} FIXME\n"
        + f"    Expenses:FIXME                        {amount:>10}\n"
        "    Assets:Girokonto\n"

@click.option("-f", "--filename", type=str, required=True)
def main(filename):
    target = Path(filename).stem + ".ledger"
    with open(target, "w") as fp:
        for row in csv2ledger(filename):

if __name__ == "__main__":

A few things are ignored here, i.e. the "effective date" vs. "booking date". I chose to only use the effective date because this should be the same one I got from the fints-api before.

Disclaimer: This bookings are not used for anything official. They are used only for the personal bookkeeping of my finances.

Using feh to sort images

In the beginning of the year I wrote a tool to label images but I don't use it anymore. Nowadays I use feh for this.

Feh has customizable actions that are by default mapped to the keys 1 to 9. For example this command:

feh -ZFd --action1 "rm '%f'" * --action2 "cp '%f' label-x"

The key 1 deletes the current file and the key 2 copies the current file to the folder label-x. When the command is run feh moves to the next image which is perfect for fast processing of images. The options ZFd are: auto-zoom, fullscreen and show filename in the left-top corner.