How many times have you come across unfriendly T-SQL such as “IF( @statusID = 1 )” wishing instead for a more meaningful encounter such as “IF( @statusID = ActiveCustomer)”? And if you think that is bad, try searching for all logic pertaining to active customers! Right from CS-101, littering code with non-obvious literal constant values is strongly discouraged which is why it is really surprising that even after the release of SQL Server 2008, there is no native support for constants in T-SQL. While we are on the subject of missing language constructs, another addition that I would really like to see in T-SQL is support for an Oracle-like “CREATE OR REPLACE” to obviate awkward “IF EXISTS…DROP…CREATE” sequences not to mention the associated chain of security processes to reapply access control.
In fairness, a CLR-based mechanism was introduced (which I will cover in a later post) to support constants in T-SQL with SQL Server 2005. However, that requires deploying .NET assemblies making the implementation disjointed at best. In any case, for all T-SQL developers out there, here is a work-around for implementing constants entirely in T-SQL!
First, because support for a “CREATE OR REPLACE” syntax is missing, create a helper procedure as follows (this is not entirely necessary, merely for convenience):
Next, create pseudo-constants to your heart’s content as follows:
And there you have it! Never again be faced with the frustration of coming face-to-face with: “IF( @statusID = 1 )”