Return a default value on Select

  • Hi All

    I am not sure if this can be done, i am using a select statement from a table with a where clause, if there are no records on the table then it must return a default value, if there are then the value from the record.

    Create table #Select (id int,Name varchar(10))

    Insert into #Select values(1,'Test')

    Insert into #Select values(2,'Test1')

    Insert into #Select values(3,'Test2')

    Insert into #Select values(4,'Test3')

    If i choose select * from #Service where id=5 , it must return some constant value if there is no record on the table, (say the default value is 10,Notavail) else return the value .

    Can this be done, i have used isnull, nullif and colase, case on select, nothing worked, 🙂

    Thanks in advance 😀

  • This works, but is probably not ideal:

    Create table #Select (id int, Name varchar(10))

    Insert into

    #Select

    values

    (1, 'Test')

    Insert into

    #Select

    values

    (2, 'Test1')

    Insert into

    #Select

    values

    (3, 'Test2')

    Insert into

    #Select

    values

    (4, 'Test3')

    select

    *

    from

    #Select

    where

    id = 5

    UNION ALL

    SELECT

    10,

    'N/A'

    WHERE

    NOT EXISTS ( SELECT

    1

    FROM

    #Select AS S

    WHERE

    id = 5 )

    DROP TABLE #Select

    I'd recommend loading the defaults in the UI if no rows are returned.

  • This is similar to the above and still not ideal but if you have a nums table you could certainly join back to that and Coalesce. (for creation of nums go to

    http://www.sqlmag.com/Blog/index.cfm?action=blogindex&DepartmentID=1016

    select coalesce(s.ID, 10) as id, coalesce(s.name, 'NotAvailable') as Name

    from nums n left join #Select s on n.N=s.id

    where n=5

  • Thanks Jack

    But i am looking to create a view, with this select statement, hope i cant do with this can i???:)

  • Mike's solution is probably going to perform better because you only hit the main table 1 time so I'd go with that, which, once you replace the temp table with a real table, will allow you to create a view. Here is how you'd create a view with my code:

    Create table my_test (id int, Name varchar(10))

    Insert into

    my_test

    values

    (1, 'Test')

    Insert into

    my_test

    values

    (2, 'Test1')

    Insert into

    my_test

    values

    (3, 'Test2')

    Insert into

    my_test

    values

    (4, 'Test3')

    GO

    CREATE VIEW test

    AS

    select

    *

    from

    my_test

    where

    id = 5

    UNION ALL

    SELECT

    10,

    'N/A'

    WHERE

    NOT EXISTS ( SELECT

    1

    FROM

    my_test AS S

    WHERE

    id = 5 )

    Go

    SELECT * FROM test AS T

    DROP TABLE my_test

    Drop view test

  • Hi Mick

    Thanks , i don't have a separate table, this is for a Java application, i have to return a default value from a view if there is no values on the table, As jack said i am a bit concerned about reading the table 2 times since this application is multi threaded and the table which i will be Querying across has got around a million numbers 😀

  • Creating an auxiliary table of numbers is something I find useful on most servers for one reason or another (Itzik Ben-Gan frequently has information that utlizes nums and as far as I know he was the first to come up with this)

    This will create it for you (pulled from his post).

    CREATE TABLE dbo.Nums(n INT NOT NULL PRIMARY KEY)

    DECLARE @max-2 AS INT, @rc AS INT;

    SET @max-2 = 1000000

    SET @rc = 1

    INSERT INTO dbo.Nums(n) VALUES(1);

    WHILE @rc * 2 <= @max-2

    BEGIN

    INSERT INTO dbo.Nums(n) SELECT n + @rc FROM dbo.Nums;

    SET @rc = @rc * 2

    END

    INSERT INTO dbo.Nums(n)

    SELECT n + @rc FROM dbo.Nums WHERE n + @rc <= @max-2

    And the code that I have above will join back to it easily within a view or however you would like to do so in order to get the data you are looking for.

    Again, not ideal but should suffice for the need you describe here.

  • I would be tempted to check the rowcount of records returned by the SELECT statement. For example:

    Create table #Select (id int, Name varchar(10))

    Insert into #Select values (1, 'Test')

    Insert into #Select values(2, 'Test1')

    Insert into #Select values(3, 'Test2')

    Insert into #Select values(4, 'Test3')

    GO

    DECLARE @RowsReturn AS INT

    DECLARE @Name AS VARCHAR(15)

    SET @RowsReturn = 0

    SET @Name = 'Not available'

    select * from #Select WHERE id = 5

    IF @@ROWCOUNT = 0

    BEGIN

    SELECT @RowsReturn, @Name

    END

    -- Clean up for additional testing

    DROP TABLE #Select

    Return for id = 5

    (No column name)(No column name)

    0 Not available

    Return for id = 3

    idName

    3Test2

    If everything seems to be going well, you have obviously overlooked something.

    Ron

    Please help us, help you -before posting a question please read[/url]
    Before posting a performance problem please read[/url]

  • Hi Mick

    This would have been ideal if i am dealing with numbers, the problem is that the table has a primary key of phone number and as you said you i have to create a table that contains phone number and if there are any numbers added to that table then that has to be inserted into the num table as well, hope i understand this right 🙂

  • Thanks Bit Bucket

    This is great, but this is not working on a view, it gives null rows, rather than my default value 🙂

  • CrazyMan (2/3/2009)


    Hi Mick

    This would have been ideal if i am dealing with numbers, the problem is that the table has a primary key of phone number and as you said you i have to create a table that contains phone number and if there are any numbers added to that table then that has to be inserted into the num table as well, hope i understand this right 🙂

    This changes the requirement fairly significantly though in theory it should work for phone numbers as well if you changed the nums table to use a bigint column and change the max to be much higher (much, much, much higher actually depending on the longest phone number possibility in your db).

    CREATE TABLE dbo.Nums(n BIGINT NOT NULL PRIMARY KEY)

    DECLARE @max-2 AS BIGINT, @rc AS INT;

    SET @max-2 = 999999999999999

    That said, I am not sure I would want my auxiliary table of numbers to be that big but it is doable.

  • Here is another option that may work for you. It starts with using Union to add the default value to a result set, but only reads the primary table one time.

    select id,Name

    from

    (

    select *,Row_Number() OVER(Order By roworder) AS rownum

    from

    (

    select *,0 as roworder

    from #Select

    where id = 5

    union

    select 0,'Notavail',1

    ) as a

    ) as b

    where rownum = 1

    The subquery 'a' adds the default value Notavail to the result set, so 'a' returns either just the default value row or the default row and a row from the phone number table. The roworder column is used for the outer subquery (note that it assumes that the phone number column has a unique constraint). The outer subquery 'b' then adds a row number to the result set. The outermost query then selects the record who's row number is 1, which will be the record from the phone number table, if one exists, and the default value record if no match is found on the phone number table.

  • --CREATE VIEW viewname AS

    SELECT

    COALESCE(sel.id, defaults.id) AS Id,

    COALESCE(sel.name, defaults.name) AS Name

    FROM (

    SELECT 10 AS id, 'NotAvail' AS Name

    ) AS defaults

    LEFT OUTER JOIN #Select sel ON sel.id = 5

    SQL DBA,SQL Server MVP(07, 08, 09) "It's a dog-eat-dog world, and I'm wearing Milk-Bone underwear." "Norm", on "Cheers". Also from "Cheers", from "Carla": "You need to know 3 things about Tortelli men: Tortelli men draw women like flies; Tortelli men treat women like flies; Tortelli men's brains are in their flies".

Viewing 13 posts - 1 through 12 (of 12 total)

You must be logged in to reply to this topic. Login to reply