mdebray/tvl-depot
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
tvl-depot/monzo_ynab
History
William Carroll aba0cd7b9d Gracefully shutdown server 
Listen for SIGINT and SIGTERM signals and write the current state to the
key-value store before shutting down.
2020-02-10 10:06:40 +00:00
..
monzo Support serde for Monzo and YNAB transaction structs 2020-02-07 21:33:08 +00:00
ynab Support serde for Monzo and YNAB transaction structs 2020-02-07 21:33:08 +00:00
.envrc Support YNAB personal-access-token 2020-02-07 21:30:24 +00:00
job.nix Create gopkgs directory for golang libs 2020-02-10 10:06:40 +00:00
main.go Create gopkgs directory for golang libs 2020-02-10 10:06:40 +00:00
README.md Support OAuth 2.0 login flow for Monzo API 2020-02-05 23:33:23 +00:00
requests.txt Create server for managing Monzo credentials 2020-02-10 10:06:40 +00:00
shell.nix Nixify tokens.go 2020-02-10 10:06:40 +00:00
tokens.go Gracefully shutdown server 2020-02-10 10:06:40 +00:00
tokens.nix Nixify tokens.go 2020-02-10 10:06:40 +00:00

README.md

monzo_ynab

Exporting Monzo transactions to my YouNeedABudget.com (i.e. YNAB) account. YNAB unfortunately doesn't currently offer an Monzo integration. As a workaround and a practical excuse to learn Go, I decided to write one myself.

This job is going to run N times per 24 hours. Monzo offers webhooks for reacting to certain types of events. I don't expect I'll need realtime data for my YNAB integration. That may change, however, so it's worth noting.

Installation

Like many other packages in this repository, monzo_ynab is packaged using Nix. To install and use, you have two options:

You can install using nix-build and then run the resulting ./result/bin/monzo_ynab.

> nix-build . && ./result/bin/monzo_ynab

Or you can install using nix-env if you'd like to create the monzo_ynab symlink.

> nix-env -f ~/briefcase/monzo_ynab -i

Deployment

While this project is currently not deployed, my plan is to host it on Google Cloud and run it as a Cloud Run application. What I don't yet know is whether or not this is feasible or a good idea. One complication that I foresee is that the OAuth 2.0 login flow requires a web browser until the access token and refresh tokens are acquired. I'm unsure how to workaround this at the moment.

For more information about the general packaging and deployment strategies I'm currently using, refer to the deployments writeup.