|
Scott
If the columns are integers, that's what it's
doing: adding them.
This will also be slow. Perhaps try
this:
edit all from salestrans t1 where
t1.plantno not in (select t2.plantno from job t2 where t2.custno = t1.custno and
t2.jobno = t2.jobno)
----- Original Message -----
From: Scott Stanfield
Sent: Wednesday, March 02, 2005 3:45 PM
Subject: [RBASE-L] - Finding Records in a table that violate the
one-to-many relationship Maybe
someone can help me with this situation. I issued the following
command:
R>edit all from salestrans where
plantno+custid+jobno not in (select plantno+custid+jobno from
job)
The
key to the table job is the plant number, customer number, and the job number
which make it unique.
Any
records in the salestrans must have a corresponding match in the job table
(many-to-one relationship).
What I
wanted to do is check the salestrans to make sure a record in the job table had
not been deleted, leaving an unmatched record(s) in the salestrans table.
This command has been sucessful in finding such strays, but it failed to find a
problem this time.
I
thought by saying plantno+custid+jobno was a way to use a multi-column key in a
command. Maybe it is literally adding the values?
|
- [RBASE-L] - Finding Records in a table that violate the o... Scott Stanfield
- [RBASE-L] - Re: Finding Records in a table that viol... benpetersen
- David M. Blocker

