seands/gnucashxml

Python library to read GNU Cash XML files

GNU Cash XML Library

gnucashxml is a Python library to parse GNU Cash XML files. This allows writing reporting utilities that do not rely on the GNU Cash libraries themselves, or require the main program to run at all. Tested with GNU Cash 2.4.10.

The library supports extracting the account tree, including all transactions and splits. It does not support scheduled transactions, price tables, and likely none but the most basic commodities. In particular, writing of XML files is not supported.

Usage

The interface is intended to allow quickly writing reports using Python. It reuses as many Python data structures as possible. Whenever dates or times are used, the standard library datetime is used. All account and transaction balances are represented as the standard Decimal type.

The three main concepts in GNU Cash are accounts, transactions, and splits. A transaction consists of a number of splits that specify from which account or to which account commodities are transferred by this transaction. All splits within a transaction together are balanced.

The main classes provided by gnucashxml mirror these concepts. A Book is the main class containing everything else. A Commodity is what is stored in an account, for example, Euros or Dollars. An Account is part of a tree structure and contains splits. Splits again are part of Transactions.

These classes all have a slots member, which is a simple dictionary for extra information. GNU Cash information such as "hidden" are recorded here.

Example

import gnucashxml

book = gnucashxml.from_filename("test.gnucash")

income_total = 0
expense_total = 0
for account, subaccounts, splits in book.walk():
    if account.actype == 'INCOME':
        income_total += sum(split.value for split in account.splits)
    elif account.actype == 'EXPENSE':
        expense_total += sum(split.value for split in account.splits)

print "Total income : {:9.2f}".format(income_total * -1)
print "Total expense: {:9.2f}".format(expense_total)