Pynomina/ledger

see also https://github.com/WolfgangFahl/pynomina/issues/3

Motivation

In the past decades the author used different personal accounting tools:

• Quicken
• Microsoft Money
• Lexware Finanzmanager
• BankingZV

The pain the conversion between those tools created was finally big enough to do something about it.

Goals

• use a computer and human-readable ledger format that is ready to survive decades
• convert from and to the formats of the tool of choice
• allow for simple sanity checks and reports
• allow for systematic tidy up
• allow for integration into a larger organizational knowledge graph

Hub & Spoke Conversion

The pyNomina tool follows a Hub and Spoke model for conversion between different personal accounting file formats. The Ledger Book (YAML/JSON) format acts as the hub, with each supported format serving as a spoke. This setup simplifies conversions by allowing data to be transformed from any spoke to the hub and then to any other spoke format.

https://diagrams.bitplan.com/render/png/0xa0be5aae.png

Supported Formats

Format	Type	Description	Wikidata Entry
Ledger Book YAML/JSON	Hub	Main format of pyNomina for converting between formats.	Ledger Book
Beancount	Spoke	A plaintext accounting format.	Beancount
GnuCash XML	Spoke	An XML-based format used by GnuCash.	GnuCash
Microsoft Money	Spoke	Zip File exported with mny_export script using mdb-tools	Microsoft Money
Finanzmanager Deluxe (QIF)	Spoke	A variant of QIF used by Finanzmanager Deluxe.	Finanzmanager Deluxe
Quicken Interchange Format	Spoke	Quicken Interchange Format (QIF)	Quicken
pyNomina Banking ZV YAML	Spoke	A format for exporting banking data in YAML or JSON.	Banking ZV

Ledger Book

The pynomina Ledger Book model consists of four main classes:

• Book
• Account
• Transaction
• Split

These are the necessary classes which work together to represent a comprehensive financial ledger records.

Class Structure

https://diagrams.bitplan.com/render/png/0xfec2cab6.png

Book

The Book class represents the main container for all financial data. It includes:

• Basic information: name, owner, creation date, and source URL
• Collections of accounts and transactions
• Methods for managing accounts and transactions, including:
  • get_stats(): Retrieves statistics about the book
  • filter(): Filters transactions by date range
  • create_account(): Creates a new account
  • add_account(): Adds an account to the book
  • lookup_account(): Finds an account by ID

Account

The Account class represents a hierarchy of individual financial accounts within the ledger. It includes:

• account_id: Unique identifier for the account
• name: Human-readable account name
• account_type: Type of account (e.g., EXPENSE, INCOME)
• description: Optional account description
• currency: Account currency (default: EUR)
• parent_account_id: Optional parent account for hierarchical structure

Transaction

The Transaction class represents individual financial transactions. It includes:

• isodate: Date of the transaction in International Standards Organization date format yyyy-mm-dd
• description: Description of the transaction
• splits: List of Split objects representing the movement of money
• payee: Optional payee information
• memo: Optional additional notes
• total_amount(): Method to calculate the total transaction amount

Split

The Split class represents the individual components of a transaction, showing how money moves between accounts. It includes:

• amount: The amount of money involved in the split
• account_id: The account associated with this part of the transaction
• memo: Optional notes for this split
• reconciled: Boolean indicating if the split has been reconciled

Example Data

Here's an example of how the ledger model is used in practice:

Yaml Format

owner: Wolfgang Fahl
url: https://github.com/WolfgangFahl/pynomina/blob/main/nomina_examples/expenses2024.yaml
since: 2024-10-06
accounts:
  Expenses:
    account_id: Expenses
    name: Expenses
    account_type: EXPENSE
    description: 'General Expenses'
    currency: EUR
  Expenses:Food:
    account_id: Expenses:Food
    name: Dining
    account_type: EXPENSE
    description: 'Expenses for Food'
    currency: EUR
    parent_account_id: Expenses
  Cash in wallet:
    account_id: Wallet
    name: Cash in Wallet
    account_type: EXPENSE
    description: ''
    currency: EUR
transactions:
  Bakery2024-10-06_0900_1:
    isodate: '2024-10-06'
    description: Bread
    splits:
      - amount: -3.50
        account_id: Cash in Wallet
      - amount: 3.50
        account_id: Expenses:Dining
        memo: Fresh sourdough bread
  Bakery2024-10-06_0900_2:
    isodate: '2024-10-06'
    description: Buns for Breakfast
    splits:
      - amount: -2.40
        account_id: Cash in Wallet
      - amount: 2.40
        account_id: Expenses:Dining
        memo: 4 whole grain buns

Implementation

The ledger model is implemented in Python, utilizing dataclasses and type hinting for clear and maintainable code. The `@lod_storable` decorator is used to enable easy serialization and deserialization of the data in YAML and/or JSON and other formats. The core idea is that the records should be readily available an tabular "list of dicts (lod)" format.