tagirijus/ledger-parse

Simple parser for ledger-cli journal files

Description

A very simple ledger-parser for ledger-cli journal files. It probably does not parse everything and it is probably buggy. Handle with care!

Installation

Simply import this module either by putting the ledger-parse.py file into the folder of your python script, or into the pythons modules path (maybe something like /usr/lib/python2.7). Then just use import ledgerparse to import the module.

A good workflow for me is: generating a symbolic link to my python2.7 lib path so that I can import the module even if it's not in the actual pythons script path, while using always the latest version of the script. You can make such a symbolic link by entering the ledger-parse folder and firing this command: sudo ln -s $PWD/ledgerparse.py /usr/lib/python2.7/ledgerparse.py.

Usage

Load a ledger formatted journal into a variable - get the content as a simple string. With this string you use the ledgerparse.string_to_ledger() function to generate a list of ledgerparse.ledger_transaction objects. Example:

ìmport ledgerparse

# get the content of the file into a variable
f = open('ledger.journal')
JOURNAL_STRING = f.read()
f.close()

# convert the string into a list with transaction-objects
JOURNAL = ledgerparse.string_to_ledger(JOURNAL_STRING)

Then you have all the transaction entries from the ledger journal in a list and can access them for example like this:

# print the first transaction - this will print a string
print JOURNAL[0]

# get the payee of the first transaction
print JOURNAL[0].payee

# print the accounts with their amount of the first transaction
for account in JOURNAL[0].accounts:
    print account

These variables are accessable:

# the transaction
JOURNAL[0].date         # a datetime object
JOURNAL[0].aux_date     # a datetime object
JOURNAL[0].state        # string, e.g. '!' or '*'
JOURNAL[0].code         # string
JOURNAL[0].payee        # string
JOURNAL[0].comments     # list of strings
JOURNAL[0].accounts     # list of ledgerparse.ledger_account objects
JOURNAL[0].original     # original string of the transaction

# the account
JOUNRAL[0].accounts[0].name         # string
JOUNRAL[0].accounts[0].commodity    # string
JOUNRAL[0].accounts[0].amount       # ledgerparse.Money object
JOUNRAL[0].accounts[0].comments     # list of strings

Why not the native ledger module?

Simple: teh native ledger module which comes with the ledger-cli programm gets all the ledger journal trasactions as postings. A posting can have the connected transaction payee etc. ... but it does not really parse the transactions. I wanted a parser which is closer to the journal like you are probably write: a list of transactions.

IMPORTANT: This module is NOT for calculating like ledger is (at least not YET).

Example

If you would like to sort a non-sorted ledger-journal, you could use this script like this:

import ledgerparse

f = open('ledger.journal')
JOURNAL_STRING = f.read()
f.close()

SORTED_JOURNAL_STRING = '\n\n'.join([x.get_original() for x in sorted(ledgerparse.string_to_ledger(JOURNAL_STRING), key=lambda y: y.date)])

f = open('ledger_sorted.journal', 'w')
f.write(SORTED_JOURNAL_STRING)
f.close()

ledger-parse's Issues

provide API reference

Please flesh out what the interface of your library looks like, i.e. what it expects and what it provides. It's hard to distinquish from the code what it is trying to do.