SELECT * FROM c
LEFT JOIN a
INNER JOIN b
ON a.x = b.k
ON a.y = c.j

SELECT * FROM a
INNER JOIN b ON a.x = b.k
RIGHT JOIN c ON a.y = c.j

…becomes…

SELECT * FROM
c
LEFT JOIN (a INNER JOIN b ON a.x = b.k) ON a.y = c.j

Very good explanation.Dont you mind may be its a silly doubt but if you make this query as left outer join still you can filter this condition in your where clause right ???

Let me know if this condition will affect any where or will give me some error…..

Thanks and regards
kamesh shah

Reason was good that it forces me to see how practically it goes . i tried my hand to see it practical but for me it is working in reverse order

Table1

c1 c2
1 NULL
2 2
3 3
4 4
5 5
6 6

Table2

c1 c2
2 NULL
3 2
4 NULL
5 NULL
6 6
7 NULL

Table3

c1 c2
4 NULL
5 2
6 NULL
7 2
8 NULL
9 9

select table1.c1,table2.c1,table2.c2 from table1
RIGHT OUTER JOIN table2
on table1.c2 = table2.c2
RIGHT OUTER JOIN table3
on table2.c2=table3.c2

select table1.c1,table2.c1,table2.c2 from table3
LEFT OUTER JOIN table2
on table3.c2 = table2.c2
LEFT OUTER JOIN table1
on table1.c2 = table2.c2

check out this both queries it gives same result?? and i think at all i can complete my task without Right outer join ..

If somewhere i m wrong correct me..

Thanks and regards
kamesh shah

Would you be able to come up with similar example in AdventureWorks database?

2.dear Marc maybe it is possible to write every queries with left in diffrent ways but a complicated code is not a good code especialy when language giving you tools.

3. imagine you have 5 table joined to each other. it is useful to order tables in some way(2 parpose:1. optimization join. 2.readability of the query 3. achive correct data) so when you want use outer join in such condition, it depends which side is your table not what you prefer.

Symbolically, a J b RJ c => c LJ ( a J b )
And the parens are important.

In T-SQL:

SELECT * FROM a
INNER JOIN b ON a.x = b.k
RIGHT JOIN c ON a.y = c.j

can be rewritten without RIGHT JOIN as

SELECT *FROM c
LEFT JOIN (
SELECT * FROM a — disambiguate common columns in *
INNER JOIN b ON a.x = b.k
) d ON c.j = d.y

but the RIGHT JOIN is less noisy.

Product 1 income count – 10 date: Somedate
Product 2 income count – 5 date: TheSameDate
Product 3 income count – 3 date: TheSameDate

Product 1 income count – 5 date: Somedate+2 days
Product 2 income count – 7 date: Somedate+2 days

Product 1 income count – 6 date: Somedate+4 days
Product 2 income count – 7 date: Somedate+6 days
.
.
.

first three records have the one parent record in book table where I store information about wharehouse, where income happened
and some other additional important information.

Ok now I have orders from my clients.
Orders table and OrderDetails table (Parent Child)
Orders Table stores Order date,Customer ID, plus some other information and OrderDetails table stores information about products
and amount of them that customer has ordered.

and here is a task with solution:
I want to consider some of my orders. Lets say there are 10 products in this order and now I want to know if I have corresponding
amount of each product in exaclty my wharehouse to satisfy clients order. And here is my SQL Select clause.

SELECT OD.prOrdDetID, OD.ProdPPID, OD.OCount, ISNULL(SUM(S.RCount), 0) AS SRCount
FROM Book AS B INNER JOIN
Supplies AS S ON B.Book_id = S.Book_id AND B.RecStatus = 0 AND B.db = @Acc RIGHT OUTER JOIN
OrderDetails AS OD
WHERE (OD.OrderID = @OrderID)
GROUP BY OD.prOrdDetID, OD.ProdPPID, OD.OCount

prOrdDetID – Is order detail ID in OrderDetails Table
ProdPPID – is product ID
OCount – is amount of products that has been ordered by customer
SRCount – is summary of product income in my wharehouse.

As you can see, first I’m taking Book and Supplies table and then I’m joining to OrderDetails
if I’d take OrderDetails table first and LEFT OUTER JOIN Them to Supplies + Book tables I’d loose (B.db = @Acc) filter
whare db indicates wharehouse number (in which I’m want to know product amount)

SELECT PD.ProductID,PD.Name,PD.ProductNumber,PM.Name AS Product_Model,
PCH.StandardCost FROM Production.Product PD
LEFT OUTER JOIN Production.ProductModel PM
ON PD.ProductModelID = PM.ProductModelID
LEFT OUTER JOIN Production.ProductCostHistory PCH
ON PD.ProductID = PCH.ProductID

No programming task requires a join, but you can also write all of your queries using syntax like SELECT * FROM a, b, c, d WHERE (a.id = b.a_id OR b.a_id IS NULL) and still have perfectly valid, well-formed, and ANSI compliant SQL.

And even thinking about reading Left-to-Right, maybe Right Joins are used more often in cultures that read from Right-to-Left, such as many Middle Eastern cultures?

Still, I don’t use Right Joins all that often – I can’t even remember the last time…