Import Transaction ID

Hey

Do you have a concept of "import id" in this API? It would be super handy for duplicate checking when feeding data into the API. Right now, if we have 2 payments for the same amount, on the same day from the same payee to import, it's hard to know know if one of them is a duplicate or not.

10replies Oldest first
  • Oldest first
  • Newest first
  • Active threads
  • Popular
  • Scott Robertson We do have this concept on our side but it isn't currently exposed through the API.  The id we store is a concatenation of the amount and date along with an incremented value suffix if there is a dupe amount/date combination.  Is this something you just would want sent back on GET requests for transactions?  Or, are you wanting to be able to write it as well?

    Like
      • Scott Robertson
      • syncforynab.com Founder
      • scottrobertson
      • 2 yrs ago
      • Reported - view

      Brady at YNAB I think we would need to write too, as we would have no way of knowing what the incremented value suffix would be. 

      In the context we need this, we have 1 transaction and that is it, so we are unable to know how many others there should be on that date. (It's a webhook POST from the bank being forwarded to YNAB).

      Like
    • Scott Robertson There are a couple of hurdles on our side for this because we use this ID for Direct Import and File Based Import to prevent dupes.  But, I am thinking we might be able to support specifying this ID when creating transactions through the API but not letting this value be changed.  If you sent an ID we already stored, we could not create the transaction.  We could then also send this back in GET requests.  Then, from your side you could send up deterministic IDs (probably from the bank feed) to prevent dupes.  In the meantime, you might be able to use the memo field as a stopgap solution.

      Like
  • That would work nicely.

     

    When we used to use the reverse engineered library, we actually used the check_id field for this if i remember correctly. But exposing that in the API would be odd. I think doing it your way is much cleaner and more consistent. 

    Like 1
  • Scott Robertson We're working on supporting an import_id and we'll probably ship initial support for it next Wednesday.

    Like 1
      • Scott Robertson
      • syncforynab.com Founder
      • scottrobertson
      • 2 yrs ago
      • Reported - view

      Brady at YNAB Perfect. That will allow much better bulk imports on my end!

      Like
  • I would also need something like that.

     

    My app fetches the transactions from various bank accounts and doesn't hold any state. I currently rely on the .csv import feature that prevents duplicates from being imported (automated through selenium webdriver). When I change my tool to communicate with the ynab api, each time it creates those transactions again and again.

     

    With a transaction id flag I could generate some hash of the transactions (in case the bank doesn't provide any unique id themselves)

    Like
  • import_id support has been added! Please see the update here: https://support.youneedabudget.com/t/k9p014/updates-released-2018-02-14

    Like 2
      • Scott Robertson
      • syncforynab.com Founder
      • scottrobertson
      • 2 yrs ago
      • Reported - view

      Brady at YNAB Awesome work Brady. Thanks for all your hard work on this!

      Like
      • Scott Robertson
      • syncforynab.com Founder
      • scottrobertson
      • 2 yrs ago
      • 1
      • Reported - view

      Brady at YNAB Just implemented this. Works perfectly! 

      Like 1
Like1 Follow
  • Status Completed
  • 1 Likes
  • 2 yrs agoLast active
  • 10Replies
  • 1451Views
  • 3 Following