96
votes

So I have a table with an identity column as the primary key, so it is an integer. So, why does SCOPE_IDENTITY() always return a decimal value instead of an int to my C# application? This is really annoying since decimal values will not implicitly convert to integers in C#, which means I now have to rewrite a bunch of stuff and have a lot of helper methods because I use SQL Server and Postgres, which Postgres does return an integer for the equivalent function..

Why does SCOPE_IDENTITY() not just return a plain integer? Are there people out there that commonly use decimal/non-identity values for primary keys?

5

5 Answers

109
votes

In SQL Server, the IDENTITY property can be assigned to tinyint, smallint, int, bigint, decimal(p, 0), or numeric(p, 0) columns. Therefore the SCOPE_IDENTITY function has to return a data type that can encompass all of the above.

As previous answers have said, just cast it to int on the server before returning it, then ADO.NET will detect its type as you expect.

37
votes

Can't you just cast it before returning it from your query or stored proc (SPs alway return int anyway, but maybe you are using an output parameter)?

Like SELECT CAST(SCOPE_IDENTITY() AS INT) AS LAST_IDENTITY

And why it does this? Probably to be more flexible and handle larger numbers.

4
votes

Scope identity return value is decimal(38,0)

CAST it, use the OUTPUT clause, or assign to an output parameter rather than SELECT SCOPE_IDENTITY() to the client

3
votes

try using this and you'll get an integer back:

ExecuteScalar('insert...; select CONVERT(int,scope_identity())');
0
votes

This isn't an exact answer to the question as asked, but in the vein of the answers above, how about (in T-SQL - MS SQLServer):

select convert(bigint, SCOPE_IDENTITY())