I need to create a view from several tables. One of the columns in the view will have to be composed out of a number of rows from one of the table as a string with comma-separated values.
Here is a simplified example of what I want to do.
Customers:
CustomerId int
CustomerName VARCHAR(100)
Orders:
CustomerId int
OrderName VARCHAR(100)
There is a one-to-many relationship between Customer and Orders. So given this data
Customers
1 'John'
2 'Marry'
Orders
1 'New Hat'
1 'New Book'
1 'New Phone'
I want a view to be like this:
Name Orders
'John' New Hat, New Book, New Phone
'Marry' NULL
So that EVERYBODY shows up in the table, regardless of whether they have orders or not.
I have a stored procedure that i need to translate to this view, but it seems that you cant declare params and call stored procs within a view. Any suggestions on how to get this query into a view?
CREATE PROCEDURE getCustomerOrders(@customerId int)
AS
DECLARE @CustomerName varchar(100)
DECLARE @Orders varchar (5000)
SELECT @Orders=COALESCE(@Orders,'') + COALESCE(OrderName,'') + ','
FROM Orders WHERE CustomerId=@customerId
-- this has to be done separately in case orders returns NULL, so no customers are excluded
SELECT @CustomerName=CustomerName FROM Customers WHERE CustomerId=@customerId
SELECT @CustomerName as CustomerName, @Orders as Orders
With ISNULL(), you can only provide one alternate value but with COALESCE you can provide more than one e.g. if col1 IS NULL then take value from column2, if that is NULL then take the default value.
COALESCE() is literally shorthand for a CASE statement, they will perform identically. However, as podiluska mentioned, ISNULL() can be occasionally faster than a CASE statement, but it's likely to be a miniscule increase as these functions are very unlikely to bottleneck your procedure.
In a narrow case, using the built-in isnull function results in better performance than coalesce on columns that are not nullable. This pattern should generally be avoided, of course. On columns that are nullable, things can really go sideways in either case.
You can use coalesce anywhere, including the where clause, yes.
EDIT: Modified answer to include creation of view.
/* Set up sample data */
create table Customers (
CustomerId int,
CustomerName VARCHAR(100)
)
create table Orders (
CustomerId int,
OrderName VARCHAR(100)
)
insert into Customers
(CustomerId, CustomerName)
select 1, 'John' union all
select 2, 'Marry'
insert into Orders
(CustomerId, OrderName)
select 1, 'New Hat' union all
select 1, 'New Book' union all
select 1, 'New Phone'
go
/* Create the view */
create view OrderView as
select c.CustomerName, x.OrderNames
from Customers c
cross apply (select stuff((select ',' + OrderName from Orders o where o.CustomerId = c.CustomerId for xml path('')),1,1,'') as OrderNames) x
go
/* Demo the view */
select * from OrderView
go
/* Clean up after demo */
drop view OrderView
drop table Customers
drop table Orders
go
In SQL Server 2008, you can take advantage of some of the features added for XML to do this all in one query without using a stored proc:
SELECT CustomerName,
STUFF( -- "STUFF" deletes the leading ', '
( SELECT ', ' + OrderName
FROM Orders
WHERE CustomerId = Customers.CutomerId
-- This causes the sub-select to be returned as a concatenated string
FOR XML PATH('')
),
1, 2, '' )
AS Orders
FROM Customers
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With