Question

Looking on some source code I’ve inherited, there’s a snippet of code that calls a SQL stored procedure making an Update.

The stored procedure returns -1 in case something goes wrong:

IF @@Error <> 0
    BEGIN
        ROLLBACK TRAN
        SELECT -1
        RETURN
    END
COMMIT TRAN
SELECT 0

The C# code is something like this:

    System.Data.SqlClient.SqlDataReader myReader;
    try{
        SqlDbConnection.Open();
        SqlDbCommand.Connection = SqlDbConnection;
        SqlDbCommand.CommandType = System.Data.CommandType.StoredProcedure;
        SqlDbCommand.CommandText = "StoredProcedured_UpdateFoo";
        SqlDbCommand.Parameters.Clear();
        SqlDbCommand.Parameters.Add("@FooData", SqlDbType.DateTime); SqlDbCommand.Parameters["@FooData"].Value = System.DateTime.Now.ToString("yyyy-MM-dd");
        myReader = SqlDbCommand.ExecuteReader();   
        if (myReader.Read())
           {
             if (int.Parse(myReader.GetValue(0).ToString()) == -1) throw new ErrorDataTxRx("Error FOO ");
           }
    } finally {
      if (SqlDbConnection.State != ConnectionState.Closed){
           SqlDbConnection.Close();
      }

      if (myReader != null){              
          if (!myReader.IsClosed) myReader.Close();
      }
    }

I also see part of the same code, checking the same thing using System.Data.DataSet() with Fill method.
Is there a more elegant way to check if the returned value is -1?
Is it OK to use an ExecuteReader in this case?

Answer 1

Have you tried using ExecuteScalar? That’s designed for queries which return a single value:

Executes the query, and returns the
first column of the first row in the
result set returned by the query.
Additional columns or rows are
ignored