waveFrontSet / django-ledger

A small ledger app in Django

This is a small Django app to compute account balances based on a history of transactions.

Too much boilerplate?

To ease the inspection and review, we've put the main code changes of the challenge solution into the PR #1.

Getting up and running

docker compose up -d --build will fire up three containers:

• One permanent container each for the database and a webserver serving the django app and
• One temporary container to apply migrations.

Because the provided Dockerfile uses manage.py as its entrypoint, you may use the following command to fire up the interactive django shell to interact with models:

docker compose run web shell

Running linters and formatters

We are using ruff both for linting and formatting purposes. The relevant commands are inside a pre-commit hook that you may install and run on-demand as follows.

• Install pre-commit.
• Run pre-commit install
• Run pre-commit run -a

Running tests

Because docker compose builds the development target of the Dockerfile, pytest and all necessary dev dependencies are installed. You may run tests in the already running web container via

docker compose exec web pytest

Models

We've implemented the ledger using two models: One Party model identifying the sender or recipient of a financial transaction and Transaction for the transaction data.

Noteworthy comments and thoughts:

• It is possible that the sender and the recipient are the same party. This has no impact on the computed balances, but it might pollute the database. We may prevent this by implementing custom model validation
• We assume that the only request patterns are for a single party and any given date. If no date is given, we take the current date (on the server) as the default.
• It is possible for transactions to lie in the future.

Bonus: ingestcsv command

To make it easier to ingest data and experiment with it, I've created a small custom command to ingest a csv file into the app. As an example, if you take

2015-01-16,john,mary,125.00
2015-01-17,john,supermarket,20.00
2015-01-17,mary,insurance,100.00

and put it into a file data.csv, you may ingest it by running

docker compose run web ingestcsv data.csv

Afterwards, you may fire up the interactive shell and experiment with the data:

❯ docker compose run web shell
>>> from financial_transactions.models import Party, Transaction
>>> Party.objects.all()
<QuerySet [<Party: john>, <Party: mary>, <Party: supermarket>, <Party: insurance>]>
>>> john = Party.objects.get(name="john")
>>> john.balance()
-14500