Path: utzoo!utgpu!jarvis.csri.toronto.edu!torsqnt!tmsoft!masnet!canremote!steve.jackson
From: steve.jackson@canremote.uucp (STEVE JACKSON)
Newsgroups: comp.databases
Subject: Re: SQL query for duplicates
Message-ID: <89122504054787@masnet.uucp>
Date: 24 Dec 89 04:30:00 GMT
Organization: Canada Remote Systems Limited, Mississauga, ON, Canada
Lines: 22

>I >hope I am not reduced to listing all rows and visually identifying
the >duplicates.  There are about 1000 rows.  I would appreciate any
>>suggestions.
How about:
=======================================================================
REM The select for update is required for updates and deletes based
REM on rowid.
select * from table
for update of column
/
REM All rows having duplicate column keys will be deleted except for
REM one in each group having the minimum rowid
delete from table x
where rowid > 
(select min(rowid)
from table
where column = x.column)
/
# copied from Oracle Technical Bulletin, June 1988
sjackson
---
 * Via ProDoor 3.1R