Question

I have two tables with a one-to-one relationship. Table1 has a composite primary key consisting of about 4 columns. Table2’s foreign key is set to Table1’s primary key.

When I try the following UPDATE clause, I am getting an error:

UPDATE Table2
SET column1 = fakeTable.c1
FROM Table2 INNER JOIN
    (
        SELECT Table1.primaryKey
        , (Table1.column3 + Table1.column4) AS c1
        FROM Table1
    ) AS c1
ON Table2.foreignKey = fakeTable.primaryKey

Am I not allowed to reference keys as if they are columns?

Answer 1

No, you need to list all the fields individually. But you can avoid the sub-query that you have…

UPDATE
  Table2
SET
  column1 = Table1.column3 + Table1.column4
FROM
  Table2
INNER JOIN
  Table1
    ON  Table2.foreignKey1 = Table1.primaryKey1
    AND Table2.foreignKey2 = Table1.primaryKey2
    AND Table2.foreignKey3 = Table1.primaryKey3
    AND Table2.foreignKey4 = Table1.primaryKey4

EDIT

Response to comment:
– I thought the whole point of keys was to avoid having to concatenate columns!

Keys aren’t a time saving device, they’re data integrity devices.

A primary key is a unique identifier. I can be a composite or not, but the important thing is that it is unique and not nullable.

A foreign key is also a data integrity device. It ensure that if data refers to something in another table, it actually must exist in that other table.