A Dynamic SQL query from lobodava is:
declare @sql nvarchar(4000) =
N';with cteColumnts (ORDINAL_POSITION, COLUMN_NAME) as
(
select ORDINAL_POSITION, COLUMN_NAME
from INFORMATION_SCHEMA.COLUMNS
where TABLE_NAME = N'''+ @tableName + ''' and COLUMN_NAME like ''' + @columnLikeFilter + '''
),
cteValues (ColumnName, SumValue) as
(
SELECT ColumnName, SumValue
FROM
(SELECT ' + @sumColumns + '
FROM dbo.' + @tableName + ') p
UNPIVOT
(SumValue FOR ColumnName IN
(' + @columns + ')
)AS unpvt
)
select row_number() over(order by ORDINAL_POSITION) as ID, ColumnName, SumValue
from cteColumnts c inner join cteValues v on COLUMN_NAME = ColumnName
order by ORDINAL_POSITION'
exec sp_executesql @sql
--OR
exec (@sql)
Why did lobodava pick exec sp_executesql @sql
and not exec(@sql)
So what is the difference here?
Is it better to use sp_executesql on recursive dynamic queries
?
In other post they say sp_executesql
is more likely to promote query plan reuse...
So it helps in these kind of queries?
Because EXEC sp_executesql
will cache the query plan -- EXEC
will not. For more info, and a very good read, see:
Caching a query means that the logistics to the query are temporarily stored, and make running the query later on faster for it.