This forum has been very helpful to me before, so I'm coming to you again.
OK, here's our situation. There's more to this story, but I'll simplify it as much as possible.
We have a large database with thousands of records that we update daily. Due to a server crash, the data and our backups were lost. We still had a copy of the database itself, just with no data in it.
Over a period of days, the IT folks were trying to retrieve all the company data from the server. Meanwhile, we still had to do our jobs. So during those days we were waiting, we used our database shell and started entering new records there. When we had to update records entered prior to the crash, we just put in the record number and just filled out the fields that changed, and left everything else blank, planning to reconcile it all later.
Now we have the old database back. So what to do?
With the brand new records that were created since the crash, that's easy. They all have new record numbers assigned to them. I can just copy and paste them into the old table that we were using.
With the records that had been created previously, that is more tricky, because now we essentially have duplicate record numbers with different data. We could have someone put the old and new data side by side, and manually copy and paste anything that needs to be updated, but that will take a long time, and that person would probably miss things. I was wondering if anyone knew of a way to merge tables, and whenever there is a conflict between identical records (defined by the record number), to use the most recent data and overwrite the old data?