Re: SQL delete question

Yeah, have seen that - uses embedded SQL in RPG to do its work - nice solution for, as he says, light work. There is no output file capability in one statement, although you could do it using the 3-jump method someone else gave you. The QMQRY method is very cool, IMO - <vbg>

Vern

-------------- Original message ----------------------
From: Alan Shore <AlanShore@xxxxxxxx>

Thanks for your reply Vern
Actually, I am using a command called RUNSQL where select can be used, and
the SQL can be used in CL
See the following web site
http://www.mcpressonline.com/sql/using-dynamic-sql-in-cl-part-1running-action-qu
eries.html
written by Michael Sansoterra




Alan Shore

NBTY, Inc
(631) 244-2000 ext. 5019
AShore@xxxxxxxx
"If you're going through Hell, keep going" - Winston Churchill
rpg400-l-bounces@xxxxxxxxxxxx wrote on 01/18/2008 01:59:16 PM:

Allan

There is a way, using QM queries, but it is a little scary.

Just put whatever the desired end result is into a SELECT statement
in a QMQRY source or one of the generic statement runners out there
- not RUNSQLSTM, which cannot do SELECT statements. Make the output
be *OUTFILE and make the outfile be the same as the file you are
running the SELECT over.

SELECT DISTINCT * FROM ITSLIB/YOURFILE

STRQMQRY QMQRY(YOURLIB/YOURQMQRY)
OUTPUT(*OUTFILE)
OUTFILE(ITSLIB/YOURFILE)
OUTMBR(*FIRST *REPLACE)

The scary thing is, you could replace the entire contents with a
single summary record, so backup - copy - backup - copy - all that.
An alternative is to make the outfile some temporary copy and then
CPYF back in to the original.

HTH
Vern

-------------- Original message ----------------------
From: Alan Shore <AlanShore@xxxxxxxx>

Afternoon all
I'm pretty sure that this question has been asked before.
I have done a search to no avail, so here is my question
I have a file that contains numerous fields on each record.
To delete duplicate records I have used
DELETE FROM FILEONE A WHERE RRN(A)
IN (SELECT RRN(B) FROM FILEONE B,
FILEONE C WHERE
B.Fielda = C.Fielda and
B.Fieldb = C.Fieldb and
B/Fieldc = C.Fieldc and
etc. etc. and
RRN(B) < RRN(C)
where etc.etc. is the list of ALL field names on the record.
My question is
Is there a different way of specifying this WITHOUT listing ALL the

field

names. If more fields are inserted into the record, and the value in

these

fields are such that the records that were duplicate are no longer
duplicate, I have to make sure that these new fields are inserted into

the

SQL statement

Thanks in advance



Alan Shore

NBTY, Inc
(631) 244-2000 ext. 5019
AShore@xxxxxxxx
"If you're going through Hell, keep going" - Winston Churchill