Question

I have a new project and just need some advice on how to do it efficiently and fast.

I have a table with about 100,000 rows with about 50 to 60 fields. It is indexed and has a primary key. The primay key in this case is a barcode.

I have to create a query that goes through each row and check for rules at each step.

For example the first rule is checking if digit 3 of the primary key is 0. So code wise it would be SUBSTRING(PkgBarcode, 3, 1) = 0

If this rule fails, I’d like to print out a message.
Also regardless if this rule fails or passes, I’d like to continue on because the rules are not dependent on each other, ie it might fail rule 2 but pass rule 3 which is okay.

I am seeking best design practices. Would it be efficient to just use select statements with print messages or some sort of loop?
How would you do this fast and efficiently?

Answer 1

You don’t want to do this “through each row” but rather with standard set theory. For example, you could add a BIT column for each test you need to run, and then run a query to update that bit based on the test:

UPDATE tblMyTable
SET [IsThirdDigitZero] = 0
WHERE SUBSTRING(PkgBarcode, 3, 1) <> 0

That makes it easy to pull a list of the ones that do not meet a certain test while avoiding a cursor.