user1634341
user1634341

Reputation: 23

Sybase *= to Ansi Standard with 2 different outer tables for same inner table

I am trying to migrate some legacy procedural code. I am having trouble figuring out the ANSI standard syntax to produce the same results.

Below is one of the many combinations I have tried. What is the inner table for the second join, is it the output from the first join or is it the source table.

Please help I have a lot of code to change.

Original SQL Statement

select * from  
JT1 a, JT2 b, JT3 c  
where a.ID *= b.ID   
  and c.JOB *= b.JOB  

My Conversion

select *   
from JT1 a  
 left outer join JT2 b   
 on a.ID = b.ID  
 right outer join JT3 c  
 on c.JOB = b.JOB  

Below is the SQL table definitions and sample data.

Create table JT1 (  
 ID   int(4)   not null,  
 NAME char(20) not null)  


Create table JT2 ( 
  ID  int(4)   not null, 
  JOB char(20) not null)  


Create table JT3 ( 
  JOB  char(20) not null, 
  DUTY char(20) not null)  

INSERT INTO dbo.JT1 VALUES(10, "Saunders")  
INSERT INTO dbo.JT1 VALUES(20, "Pernal")  
INSERT INTO dbo.JT1 VALUES(30, "Marenghi")  
INSERT INTO dbo.JT2 VALUES(20, "Sales")  
INSERT INTO dbo.JT2 VALUES(30, "Clerk")   
INSERT INTO dbo.JT2 VALUES(30, "Mgr")  
INSERT INTO dbo.JT2 VALUES(40, "Sales")  
INSERT INTO dbo.JT2 VALUES(50, "Mgr")  
INSERT INTO dbo.JT3 VALUES("Mgr","Evaluate")  
INSERT INTO dbo.JT3 VALUES("Mgr","Reports")  
INSERT INTO dbo.JT3 VALUES("Mgr","Meeting")  
INSERT INTO dbo.JT3 VALUES("Clerk","Stocking")  
INSERT INTO dbo.JT3 VALUES("Clerk","Customer Request")  

Upvotes: 2

Views: 2161

Answers (4)

how to create query...

Sample table : Transact

Entry    Date       PartyIdno  Catage    Credit/Debit  Amount
------   ---------- ---------- --------- ------------  --------
     1   02-01-2016          1   Receipt       C          8,200
     1   02-01-2016          5   Payment       D          8,200
     2   14-02-2016          1     Sales       D         11,200 
     2   14-02-2016          4     Sales       C          6,500
     2   14-02-2016          2     Sales       C          4,700

Output
------

Entry  Date       PartyIdno                     Debit   Credit   Balance
-----  ------     ----------                    ------  -------  -----------                   
                             SubId  SubAmount
                             -----  ---------
  1   02-01-2016      1                                  8,200    8,200 Cr
      02-01-2016               5     8,200 Dr
  2   14-02-2016      1                         11,200            3,000 Dr
      14-02-2016               4     6,500 Cr
      14-02-2016               2     4,700 Cr

SQL Statement :

GLOBALSSS.Open_DB_Connection()

Dim cmd As New Sql Command("Select T1.ENTRY,COALESCE(T1.PARTYIDNO,T2.PARTYIDNO) AS Colour FROM TRANSACT T1 FULL OUTER JOIN (SELECT PARTYIDNO  FROM TRANSACT WHERE ENTRY=1) T2  ON 1 = 0 WHERE T1.PARTYIDNO=1 OR T2.PARTYIDNO-1  ", Connection)

Dim da As New SqlDataAdapter(cmd)

Dim dt As New DataTable

da.Fill(dt)

DataGridView1.DataSource = dt

GLOBALSSS.Close_DB_Connection()

Upvotes: 0

HLGEM
HLGEM

Reputation: 96552

OK it took me awhile but try this:

select   a.ID,  a.NAME, b.ID,   b.JOB,  a.JOB,  a.DUTY    
from (Select * from #jt1    
      cross join #jt3  ) a 
left outer join #jt2 b    
  on a.ID = b.ID    and a.job = b.job

The problem with using that left join operator mulitple times is that you really had a hidden cross join in there. This should get the right results, As to whether the results have been incorrect all along due to developers not undersatnding waht they were doing, only you can tell.

Upvotes: 3

Dave
Dave

Reputation: 4597

*= is equivalent to left [outer] join

=* is equivalent to right [outer] join

Upvotes: 1

Robert
Robert

Reputation: 25753

An original query is equivalent to:

select * 
from JT1 a
left join JT2 b on a.ID = b.ID
left join JT3 c on c.JOB = b.JOB

Upvotes: 2

Related Questions