You can ignore them. You can delete them. You can rewrite them with SendKey. I don’t think there is anything stopping you ? There is no way to recover the original primary key from just the digest though, so unless its stored somewhere or you can rewrite all your data it is lost.
How are you querying? Can you give us a snippet? Maybe this is something I’m not understanding thats baked into the language?
Here is the case:
I don’t have the ids of the documents (this is some userIds).
My code in JAVA.
If I need to get a specific document for a user - I can get id with getByIds - client.get(null, keys).
But I need to update all the documents (again - I don’t have the list of ids).
So I try to do scanAll and loop over the result - but we have documents without an id
So when I try to save the updated document - it failed “Id must not be null!”
How can I save a ducument that I just fetch but there is no PK.
Can you show us a code snippet with the scan all? Surely you can detect the record in the scancallback, determine if it has the Id key, and handle it differently?
scanCallback(Key key, Record record) - provides the key object - which you can use to update the record. It has the digest - so even if you don’t know the user key, you can still update that record.