Library for working with the nYNAB private APIBudget/Transactions Read/Write/Migration tools
a python client for the new YNAB API
pip install pynYNAB or download the source, then call python setup.py install
Either code your own script that uses the pynYNAB api, or use the provided scripts, ofximport, YNAB4 migrate, csvimport
see appropriate README
The connection object needs email and password for the nYNAB account
Once you have created your nYnabClient object, all data should have already been synced up with YNAB’s servers. If you need the updated data from the server, call sync on the nYnabClient.
All the entity handling is done through the Budget and Catalog objects, they contain collections such as be_accounts, be_transactions, ce_user_settings, etc. Look into the budget/catalog for the schema.
In order to write some data to YNAB servers for your budget, you just need to modify those Budget/Catalog objects then call nYnabobject.push . all the subcollections like ce_budget_transactions support append, remove, etc, like normal lists.
push takes an expected_delta argument, this is to safeguard modifying too much of your data on the server by error. Examples: * if you add two matched transactions to be_transactions, then the expected delta is 2 * If you add a single payee, then the expected delta is 1
I’ve provided some tested methods e.g. add_account, add_transaction, in the nYnabClient class to add/delete accounts and transactions as examples. Some actions are not always simple (e.g., to add an account, you need to add a transfer payee, a starting balance transaction, otherwise the server refuses the request). You’re welcome to contribute new actions e.g. by catching the requests done in the UI with Fiddler to see what’s done.
Caution with add_transaction, it works even for large amount of transactions (tested up to 3000), but please don’t stress test the YNAB servers with it…
I’ve taken all precautionary steps so that this python Client can’t affect YNAB even if used too widely. * It honors requests by the server to throttle its requests > Same mechanisme that the JS app uses * It self-limits the requests to 5 request per second * It clearly identifies itself by User-Agent > Easy to lock it out if it causes trouble