@tordans you only care about this for the CSV export, right? If so, I'll change the export to automatically unpack the state key from the ID, that will require far fewer changes :)

@k-nut yes, only CSV. But I can also continue extracting them myself if this is not useful for anyone else...

Ok, I've changed the export script to also unpack the state and add it as the last column. You can see a preview here: https://jedeschule.codefor.de/csv-data/jedeschule-data-2026-04-20.csv Does this work for you? If so, I'll keep the change permanently. Always happy to make changes to make it easier for others to consume the data.

Does this work for you? If so, I'll keep the change permanently. Always happy to make changes to make it easier for others to consume the data.

@k-nut yes, that works as well.

Personally I think it is a bit weird do have csv and API with different results, but maybe that is not so important ATM given the current usage and use cases.

I will close this PR now given the different approach.

@k-nut one thing about the name: state makes it look like this key is part of the address. However, there might be legitimate cases where a school is operated outside the state boundary, so it has a different address.

Ah, that's why you called it state_key 💡 How about we call it source_state or something like that in the export?

How about we call it source_state or something like that in the export?

Yes, sounds good. Could also be something related to the scraper scraper_key.