Modeling relational data
Our application currently stores data in a single CSV file; a file like this is often called a flat file, because the data has been flattened to two dimensions. While this format works acceptably for our application and could be translated directly to a SQL table, a more accurate and useful data model requires more complexity. In this section, we're going to go through some concepts of data modeling that will help us convert our CSV data into effective relational tables.
Primary keys
Every table in a relational database should have something called a primary key. The primary key is a value, or set of values, that uniquely identifies a record in the table; as such, it should be a value or set of values that is unique and non-null for every row in a table. Other tables in the database can use this field to reference particular rows of the table. This is called a foreign key relationship.
How do we figure out what the primary key is for a set...
