is there an algorithm that can spot similarities between two linked lists on the virtue of what an element is linked to? I'm working with my bank statements in CSV and want to work with them mechanically, and I don't really want to spend time checking manually what I have imported (to Wallet by Budget Bakers) and what I'm missing between imports

Sometimes I have transactions repeated like this that I truly did (they aren't errors of double spending) and as you can see, there is little to no information to differentiate a repeated transaction (because you are importing an CSV that carries something you already imported) from a truly repeated transaction (I bought something twice the same day)

@atheist I made a comment with a picture directly from my banks web page displaying what the data actually looks and how much information there is

EDIT: There is no time in that date, not even in the CSV 🤷

@catholic-emacs

You could write a small script that e.g. imports to access / SQLite or even do a pure Excel solution.

Read CSV line by line
Hash Data + Account Number + bank code

Does Hash Exist?
- Yes, Manual Intervention necessary
- No, add to known entries

Not perfect, but without an transaction ID you're really clueless.

@IntrusionCM I try something like that on my first try, but was too naive (code here https://gitlab.com/shackra/... and whatever solution I come to will be in there)

I'm looking to use EdgeDB just because I don't want to shot myself in the foot with SQL for something more involve as linked lists

@catholic-emacs

@atheist had a good idea with the balance after transaction, that would be a unique criteria... And could be calculated as starting point if missing.