Newbie account matching

[ar]

Hi,

I had just switched to gnucash from many years of quicken. So, far it works
fine.

The only problem is account matching. I am not looking into something fancy
- the vast majority of our expenses is in less than 10 stores and less than
10 gas stations, everything else I can classify manually.

In addition to that there are quirks/bugs per reporting company.

I suppose I can easily write a script that will perform proper matching.
However, I am using it on win10 and there are no python bindings. Hence I
have to write a standalone pre-processor that will read downloaded file,
fix the quirks and add explicit account-id string.

My current plan is to generate .csv file and then import it. Should I
target .qif instead? What strings should I use as account-ids in .csv file?
Say if want something to be matched to 'Expenses:Auto:Gas' what string
should I use? Should I add any configuration for this to happen?

Thank you,

Aleksey