Tough Problem, need help

[Matt Creely]

I have a very strange database with a very strange problem.

Consider 4 tables:

Table1:
----------------
Table1ID INT PK
Table2ID INT FK
Table3ID INT FK
OrderNo VARCHAR(50)

Table2
----------------
Table2ID INT PK
Table4ID INT FK

Table3
----------------
Table3ID INT PK
Table2ID INT FK

Table4
----------------
Table4ID INT PK
OrderTotal VARCHAR(50)
With Data:

Table1:
------------
1 1 NULL 90001
2 2 NULL 90002
3 NULL 1 90003
4 NULL 2 90004

Table2:
------------
1 1
2 1

Table3:
------------
1 1
2 2

Table4:
------------
1 500
2 1000
Table1 can have either a Table2ID OR a Table3ID but not both.
This is the query I'm attempting:
---------------------------

SELECT dbo.Table1.OrderNo, dbo.Table4.OrderTotal
FROM dbo.Table1 LEFT OUTER JOIN
dbo.Table4 INNER JOIN
dbo.Table2 ON dbo.Table4.Table4ID =
dbo.Table2.Table4ID INNER JOIN
dbo.Table3 ON dbo.Table2.Table2ID =
dbo.Table3.Table3ID ON dbo.Table1.Table2ID = dbo.Table2.Table2ID AND
dbo.Table1.Table3ID = dbo.Table3.Table3ID
Which gives me:
---------------------------
90001 NULL
90002 NULL
90003 NULL
90004 NULL

When I really want:
----------------------------
90001 500
90002 500
90003 500
90003 1000 (NOT 500)
I don't know how to do this. Are any of you sql guru's up to the
challenge?

Thanks in advance

-Matt

p.s. sql to recreate tables included

CREATE TABLE [dbo].[Table1] (
[Table1ID] [int] IDENTITY (1, 1) NOT NULL ,
[OrderNo] [varchar] (50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL ,
[Table2ID] [int] NULL ,
[Table3ID] [int] NULL
) ON [PRIMARY]
GO

CREATE TABLE [dbo].[Table2] (
[Table2ID] [int] IDENTITY (1, 1) NOT NULL ,
[Table4ID] [int] NULL
) ON [PRIMARY]
GO

CREATE TABLE [dbo].[Table3] (
[Table3ID] [int] IDENTITY (1, 1) NOT NULL ,
[Table2ID] [char] (10) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
) ON [PRIMARY]
GO

CREATE TABLE [dbo].[Table4] (
[Table4ID] [int] IDENTITY (1, 1) NOT NULL ,
[OrderTotal] [varchar] (50) COLLATE SQL_Latin1_General_CP1_CI_AS NULL
) ON [PRIMARY]
GO

ALTER TABLE [dbo].[Table1] WITH NOCHECK ADD
CONSTRAINT [PK_Table1] PRIMARY KEY CLUSTERED
(
[Table1ID]
) ON [PRIMARY]
GO

ALTER TABLE [dbo].[Table2] WITH NOCHECK ADD
CONSTRAINT [PK_Table2] PRIMARY KEY CLUSTERED
(
[Table2ID]
) ON [PRIMARY]
GO

ALTER TABLE [dbo].[Table3] WITH NOCHECK ADD
CONSTRAINT [PK_Table3] PRIMARY KEY CLUSTERED
(
[Table3ID]
) ON [PRIMARY]
GO

ALTER TABLE [dbo].[Table4] WITH NOCHECK ADD
CONSTRAINT [PK_Table4] PRIMARY KEY CLUSTERED
(
[Table4ID]
) ON [PRIMARY]
GO

ALTER TABLE [dbo].[Table1] ADD
CONSTRAINT [FK_Table1_Table2] FOREIGN KEY
(
[Table2ID]
) REFERENCES [dbo].[Table2] (
[Table2ID]
),
CONSTRAINT [FK_Table1_Table3] FOREIGN KEY
(
[Table3ID]
) REFERENCES [dbo].[Table3] (
[Table3ID]
)
GO

ALTER TABLE [dbo].[Table3] ADD
CONSTRAINT [FK_Table3_Table2] FOREIGN KEY
(
[Table3ID]
) REFERENCES [dbo].[Table2] (
[Table2ID]
)
GO

[Anith Sen]

Do:

SELECT Table2.Table3ID, Table4.OrderTotal
FROM Table1
LEFT OUTER JOIN Table2
ON COALESCE(Table1.Table2ID, 1) = Table2.Table2ID
INNER JOIN table4
ON Table2.Table2ID = Table4.Table4ID ;

--
- Anith
( Please reply to newsgroups only )

[Matt Creely]

Thanks for your help. However, your query only returns:

(changed Table2.Table3ID to Table1.OrderNo)
-------------------------------------------
90001 500
90003 500
90004 500

Instead of:
-------------------------------------------
90001 500
90002 500
90003 500
90003 1000 (NOT 500)

But I'm getting closer at least. Any other ideas?


"Anith Sen" <an***@bizdatasolutions.com> wrote in message news:<b0*********************@bgtnsc04-news.ops.worldnet.att.net>...

Do:

SELECT Table2.Table3ID, Table4.OrderTotal
FROM Table1
LEFT OUTER JOIN Table2
ON COALESCE(Table1.Table2ID, 1) = Table2.Table2ID
INNER JOIN table4
ON Table2.Table2ID = Table4.Table4ID ;

[Steve Kass]

Matt,

Is this what you want?

select Table1ID, OrderTotal
from Table1, Table2, Table4
where Table1.Table3ID is null
and Table1.Table2ID = Table2.Table2ID
and Table2.Table4ID = Table4.Table4ID
union all
select Table1ID, OrderTotal
from Table1, Table3, Table4
where Table1.Table2ID is null
and Table1.Table3ID = Table3.Table3ID
and Table3.Table4ID = Table4.Table4ID

-- Steve Kass
-- Drew University
-- Ref: 53DDD827-B99C-4787-9D20-46F3D86EA51B

Matt Creely wrote:

Thanks for your help. However, your query only returns:

(changed Table2.Table3ID to Table1.OrderNo)
-------------------------------------------
90001 500
90003 500
90004 500

Instead of:
-------------------------------------------
90001 500
90002 500
90003 500
90003 1000 (NOT 500)

But I'm getting closer at least. Any other ideas?


"Anith Sen" <an***@bizdatasolutions.com> wrote in message news:<b0*********************@bgtnsc04-news.ops.worldnet.att.net>...

Do:

SELECT Table2.Table3ID, Table4.OrderTotal
FROM Table1
LEFT OUTER JOIN Table2
ON COALESCE(Table1.Table2ID, 1) = Table2.Table2ID
INNER JOIN table4
ON Table2.Table2ID = Table4.Table4ID ;

[Matt Creely]

Actually, I DID include the Create Statements, and although I didn't
include the INSERT statements (I will next time) I did include the
data that should have existed in the tables :P

Erland Sommarskog <so****@algonet.se> wrote in message news:<Xn**********************@127.0.0.1>...

Matt Creely (cr*****@hotmail.com) writes:

Thanks for your help. However, your query only returns:

(changed Table2.Table3ID to Table1.OrderNo)
-------------------------------------------
90001 500
90003 500
90004 500

Instead of:
-------------------------------------------
90001 500
90002 500
90003 500
90003 1000 (NOT 500)

But I'm getting closer at least. Any other ideas?

Yes. To wit the standard suggestion. Include in your posting the
following:

o CREATE TABLE statments for the involved tables.
o INSERT statements for the sample data.
o The desired output. (Which you did include, but I repeat it for
completeness sake.)

Since you did not provide this, it was not possible for Anith to post a
tested solution.

[Guinness Mann]

In article <25**************************@posting.google.com >, cr33d09
@hotmail.com says...

Actually, I DID include the Create Statements, and although I didn't
include the INSERT statements (I will next time) I did include the
data that should have existed in the tables :P

You don't understand the way these guys work here, Matt. :-)

They sit at their computer and read the newsgroup with an SQL Query
Analyzer (or something like it) window open. When they come to one of
these questions they cut and paste the table creation statements into QA
and execute them. They then execute the data insertion statements and
voila, they are working on live data. Listing the contents of a table
is less than useless for them.

-- Rick

[Erland Sommarskog]

Matt Creely (cr*****@hotmail.com) writes:

Actually, I DID include the Create Statements, and although I didn't
include the INSERT statements (I will next time) I did include the
data that should have existed in the tables :P

I reviewed the thread, and as an extra precaution I also checked Google,
in case my newsserver would have missed a posting of yours, but I could
find no CREATE TABLE statements. There were outlines of the table, but
that is not the same.

As Guinness Mann points out, the deal is that with the SQL statements
prepared, it's an easy affair to cut and paste the statements into
Query Analyzer for test.

You may think that I am lazy, because I don't want to type CREATE TABLE
statements, but, well, it's always more rewarding to help someone who
also makes an effort himself.

--
Erland Sommarskog, SQL Server MVP, so****@algonet.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinf...2000/books.asp