financeager 0.22

command line tool for organizing finances

FINANCEAGER

An application (possibly running as Flask webservice) that helps you administering your daily expenses and earnings. Interact via the command line interface.

The financeager backend holds databases (internally referred to as 'periods'). A period contains entries of a certain year.

Quickstart

Who is this for?

You might be someone who wants to organize finances with a simple software because you're tired of Excel and the like. And you like the command line. And Python.

NOTE

The project is actively developed. Expect things to break - e.g. the command line interface, the REST API definitions, ... - before version 1.0.0 is released.

Installation

From PyPI package

pip install financeager

Using pipx

If you're using Python >= 3.6 and have pipx installed, install financeager into an isolated environment via

pipx install financeager

From source (master branch)

Clone the repo

git clone https://github.com/pylipp/financeager
cd financeager

Create a virtual environment

python -m venv --python=$(which python3) .venv
source .venv/bin/activate

Install

pip install --upgrade --editable .

Alternatively, you can omit the second and third step and install financeager to ~/.local with (requires pip3)

pip3 install . --user

Usage

You can use financeager as a client-server or a serverless application (default). The user interacts via the command line interface (CLI).

Serverless mode

The user request invoked from the CLI is passed to the backend which opens the appropriate database, processes the request, closes the database and returns a response. All communication happens within a single process, hence the label 'serverless'. The databases are stored in ~/.local/share/financeager.

Client-server mode

To run financeager as client-server application, start the flask webservice by

export FLASK_APP=financeager/fflask.py
flask run  # --help for more info

This does not store data persistently! Specify the environment variable FINANCEAGER_DATA_DIR.

For production use, you should wrap app = fflask.create_app(data_dir=...) in a WSGI or FCGI (see examples/ directory).

To communicate with the webservice, the financeager configuration has to be adjusted. Create and open the file ~/.config/financeager/config. If you're on the machine that runs the webservice, put the lines

[SERVICE]
name = flask

If you're on an actual remote 'client' machine, put

[SERVICE]
name = flask

[SERVICE:FLASK]
host = https://foo.pythonanywhere.com
timeout = 10
username = foouser
password = S3cr3t

This specifies the timeout for HTTP requests and username/password for basic auth, if required by the server.

In any case, you're all set up! See the next section about the available client CLI commands and options.

Command line client

usage: financeager [-h] [-V] {add,get,remove,update,copy,list,periods} ...

optional arguments:
  -h, --help            show this help message and exit
  -V, --version         display version info and exit

subcommands:
  {add,get,remove,update,copy,list,periods}
                        list of available subcommands
    add                 add an entry to the database
    get                 show information about single entry
    remove              remove an entry from the database
    update              update one or more fields of an database entry
    copy                copy an entry from one period to another
    list                list all entries in the period database
    periods             list all period databases

On the client side, financeager provides the following commands to interact with the backend: add, update, remove, get, list, periods, copy.

Add earnings (no/positive sign) and expenses (negative sign) to the database:

> financeager add burgers -19.99 --category Restaurants
> financeager add lottery 123.45 --date 03-14

Category and date can be optionally specified. They default to None and the current day's date, resp. financeager will try to derive the entry category from the database if not specified. If several matches are found, the default category is used.

Add recurrent entries using the -t recurrent flag (t for table name) and specify the frequency (yearly, half-yearly, quarterly, bi-monthly, monthly, weekly, daily) with the -f flag and optionally start and end date with the -s and -e flags, resp.

> financeager add rent -500 -t recurrent -f monthly -s 01-01 -c rent

If not specified, the start date defaults to the current date and the end date to the last day of the database's year.

Did you make a mistake when adding a new entry? Update one or more fields by calling the update command with the entry's ID and the respective corrected fields:

> financeager update 1 --name "McKing Burgers" --value -18.59

Remove an entry by specifying its ID (visible in the output of the list command). This removes the burgers entry:

> financeager remove 1

This would remove the recurrent rent entries (ID is also 1 because standard and recurrent entries are stored in separate tables):

> financeager remove 1 --table-name recurrent

Show a side-by-side overview of earnings and expenses (filter by date/category/name/value by passing the --filters option, e.g. --filters category=food to show entries in the categories food)

> financeager list

               Earnings               |                Expenses
Name               Value    Date  ID  | Name               Value    Date  ID
Unspecified          123.45           | Rent                1500.00
  Lottery            123.45 03-14   2 |   Rent January       500.00 01-01   1
                                      |   Rent February      500.00 02-01   1
                                      |   Rent March         500.00 03-01   1
=============================================================================
Total                123.45           | Total               1500.00

The aforementioned financeager commands operate on the default database (named by the current year, e.g. 2017) unless another period is specified by the --period flag.

> financeager add xmas-gifts -42 --date 12-23 --period 2016

Copy an entry from one database to another by specifying entry ID and source/destination period:

> financeager copy 1 --source 2017 --destination 2018

Detailed information is available from

> financeager --help
> financeager <subcommand> --help

You can turn on printing debug messages to the terminal using the --verbose option, e.g.

> financeager list --verbose

You can find a log of interactions at ~/.local/share/financeager/log (on both the client machine and the server).

More on configuration

Besides specifying the backend to communicate with, you can also configure frontend options: the name of the default category (assigned when omitting the category option when e.g. adding an entry) and the date format (string that datetime.strptime understands; note the double percent). The defaults are:

[FRONTEND]
default_category = unspecified
date_format = %%m-%%d

The financeager command line client tries to read the configuration from ~/.config/financeager/config. You can specify a custom path by passing it along with the -C/--config command line option.

More Goodies

• financeager will store requests if the server is not reachable (the timeout is configurable). The offline backup is restored the next time a connection is established. This feature is only available when running financeager with flask.

Expansion

Want to use a different database? Should be straightforward by deriving from Period and implementing the _entry() methods. Modify the Server class accordingly to use the new period type.

Architecture

The following diagram sketches the relationship between financeager's modules. See the module docstrings for more information.

+--------+   +-----------+   +---------+
| config |-->|    cli    |<->| offline |
+--------+   +-----------+   +---------+
                 ¦   Λ
                 V   ¦
+-------------------------------------+
|             communication           |
+-------------------------------------+
                                           +---------+     +---------+
  [pre-processing]      [formatting]  <--  | listing | <-- | entries |
                                           +---------+     +---------+
        ¦                     Λ
        V                     ¦

+--------------+   |   +--------------+
| httprequests |   |   |              |     FRONTEND
+--------------+   |   |              |
================   |   |              |    ==========
+--------------+   |   | localserver  |
|    fflask    |   |   |              |     BACKEND
+--------------+   |   |              |
|  resources   |   |   |              |
+--------------+   |   +--------------+

        ¦                     Λ
        V                     ¦
+-------------------------------------+
|                server               |
+-------------------------------------+
        ¦                     Λ
        V                     ¦
+-------------------------------------+
|                period               |
+-------------------------------------+

Known bugs

• see issues
• Please. Report. Them.

Contributing

Always welcome! For installation, follow the steps above (section 'From source') and install development dependencies via

pip install --upgrade --editable .[develop]
pre-commit install
gitlint install-hook

You're all set for hacking! Please adhere to test-driven development, if possible: When adding a feature, or fixing a bug, try to construct a test first, and subsequently adapt the implementation. Run the tests from the root directory via

python setup.py test

If you added a non-cosmetic change (i.e. a change in functionality, e.g. a bug fix or a new feature), please update Changelog.md accordingly as well. Check this README whether the content is still up to date.

Releasing

1. Tag the latest commit on master by incrementing the current version accordingly (scheme v0.X).
2. Run make release.

financeager features

Ideas

• experiment with urwid for building TUI or remi for HTML-based GUI
• support querying of standard/recurrent table with list
• return element data as response to add/copy/update request
• support passing multiple elements IDs to update/remove/copy/get (maybe together with asynchronous HTTP requests)
• extended period names (something along 2018-personal)
• support list at date other than today

Discarded ideas

• select from multiple options if possible (e.g. when searching or deleting an entry): breaks the concept of having a single request-response action. Instead, the user is expected to know which element he wants to delete (by using the element ID) and can give a precise command

Roadmap for release of version 1.0

This requires some restructuring of the software architecture. Motivation and goals are outlined below.

Status quo

• module functionalities and responsibilities particularly overlap
• also apparent in test code: no clear distinction between integration and unit tests

Goals

• three separated top modules: core, backend, client
• responsibilities:
  1. core:
     • constants
     • configuration (maybe move to client)
     • exceptions
  2. backend:
     • interfaces (localserver, fflask)
     • REST API (resources)
     • database management (server, period)
  3. client
     • CLI
     • communication pre-/post-processing
     • HTTP requests
     • response formatting (entries, listing)
• consistent, modular test structure
• pave way for terminal user interface

TODOs

• remove TinyDB usage from model and entries
• remove entries import from period
• more fine-grained error-handling in period (distinguish between errors during validation and about non-existing elements)
• integration test of cli module
• move data dir to ~/.local/share/financeager
• install pre-commit framework
• use logging module instead of print
• use marshmallow package for keyword validation/serialization in period and resources
• have return codes in cli.run
• introduce verbose cli option
• add loggers to config and offline modules
• add loggers to resources and server
• avoid test code interfering with actual file system content
• test offline feature with 'none' backend
• rename 'model' to 'listing'
• clean up test_communication

PERSONAL NOTE

This is a 'sandbox' project of mine. I'm exploring and experimenting with databases, data models, server applications (Pyro4 and flask), frontends (command line, Qt-based GUI), software architecture, programming best practices (cough) and general Python development.

Feel free to browse the project and give feedback (comments, issues, pull requests).