hi all,
I am currently working on a database with lots of weird values in it. It was originally a spreadsheet, and I need to make it a database (and its a pretty bad spreadsheet too)
Here is a the UNF representation
UNF (Date, Client, Address, Contact, Helicopter, Pilot, Charter_Type, Cost_Per_Hour, Passengers, Leg, ETA, ETD, Origin, Destination, ATA, ATD, Engine_Time, Total_Engine_Time, Special_Requirements)
The Leg field has a value of 1, 2, 3 etc, but this may be duplicated as there can be multiple Clients. I am thinking of using the Client and the Leg as an initial composite PK.
Basically I work it as following (PKs are in bold)
1NF
CHARTER (Date, Client, Address, Contact, Helicopter_Number, Pilot, Charter_Type, Cost_Per_Hour, Passengers, Leg, ETA, ETD, Origin, Destination, ATA, ATD, Engine_Time, Total_Engine_Time, Special_Requirements)
So to cut it down, I have four relations at the end
Client (Client, Address, Contact)
Leg(Leg, Date, ETA, ETD, Origin, Destination, ATA, ATD, Engine_Time)
Helicopter(Helicopter_Number, Cost_Per_Hour)
Charter (Passengers, Total_Engine_Time, Special_Requirements, Charter_Type, Pilot)
Thing is, I have no PK for the Charter relation, should I just invent one??
If somebody could clear this up, I would appreciated it, and I apologise for being daft! I have read a number of guides on Normalisation, but nearly all assume there is one attribute that is obviously the PK...