Dynamicallly writing those pesky Update Statements

Whenever you are loading a data warehouse, often one of the most time consuming tasks is writing the update statements that you need in order to properly handle changing records coming from your system. Further adding to development time, columns are then split into type 1 or type 2 attributes, and in my current shop, processes for deletes out of the source system as well. After you write a few of these, that loving feeling you got when you started starts to dissipate, and you start to feel bored with the same statements. So with that in mind, I wrote the below procedure that takes a table name and an update_type variable of U, V, or D and then generates the appropriate update, type 2 version, or delete statement you need. The procedure makes 2 assumptions, one in that you have a separate etl_working database to handle set based updates…