Hi Chahit
Firebase is more like a NoSQL database than SQL. It's more oriented toward realtime and flexibility, than atomicity and fixed structure.
Otherwise, you can indeed think about workaround such as:
- query all documents and filter unique field values in your program (if not too many entries)
- denormalize the data: each time you store a new entry, you would set `/degree/{value}/count` to current value + 1, and decrement on delete. once you've done that, you can just query all `/degree` children with count > 0 which keys would be a distinct list of your main dataset degree values.
Hope that helps
--
Aurelien