Feature: duplicate transaction detection for CSV imports

[mgirouard]

Describe the bug
After a few weeks of using Maybe, I noticed that I find myself deleting duplicate transactions often. I regularly export CSVs from my banks and import them. It's easier for me to grab the last 30 days of activity and import them without having to grab specific dates, but when I do this I always wind up with duplicates.

To Reproduce
Steps to reproduce the behavior:

1. Go to any account
2. Click on the ellipsis (...) menu and select Import transactions
3. Follow the normal import flow (adjust fields, adjust data, confirm transactions)
4. Navigate back to the account and review the latest transactions

(Perhaps a quicker repro would be to create a temporary account and import transactions twice)

Expected behavior
Duplicate transactions should be filtered out and not be imported.

Screenshots / Recordings
n/a

Additional context
Perhaps this could be done by hashing the transaction data -- exact matches would be a high probability of being duplicates (same description, same amount, same date, etc).

There will almost certainly be edge cases though -- eg: if I intentionally make the same purchase twice. One of the screens during import might be a good place to call this out and give me a checkbox to import it in or not.

[NGUYENDUCCHUNG21]

This text is italicized

[zachgoll]

@mgirouard currently, we do not have any support for detecting duplicates, but I agree that the feature would be useful.

I agree with you that hashing the raw transaction data is probably the best option for detecting matches.

What I'm not 100% sure about is the UI/UX for this (cc @justinfar)

My guess is that we could provide some sort of duplicate filtering / removal in our "clean" step?

[mgirouard]

@zachgoll Yeah, this makes the most sense to me. Thanks for considering :)

[vanhalt]

I was thinking about this. But by hashing the transaction we may be discarding duplicated charges:

10/07/2024,Apple Store,-36
10/07/2024,Apple Store,-36

[zachgoll]

@vanhalt yeah, that's why I'm thinking we'll need a UI/UX that simply surfaces duplicates to the user and then lets them decide whether to remove them or not.