Hi Andy.

Taking Dirk's suggestion a bit further, firstly create a FIFO file in Unix using
mknod or mkfifo, then start up an awk script reading from the FIFO, containing
something like:

BEGIN {
FS = "|"
}

{
if (condition)
print > "reload.unl"
else
print | "gzip > purge.unl.gz"
}

Use compress if you don't have gzip. Then in dbaccess run:

UNLOAD TO "fifo" SELECT * FROM tablename

--
Regards,
Doug Lawry
www.douglawry.webhop.org


"Dirk Gunsthövel" <dirk@guncon.de> wrote in message
news:cs12ja$elv$00$1@news.t-online.com...
> Hi :-)
>
> What about writing a SP to_keep(...) which returns 1 or 0
> based on the data of the row.
>
> then unload
> UNLOAD to somefile SELECT to_keep(...),* FROM thetable;
>
> Then use an easy awk script or something alike (on Windows systems
> FIND is your friend) to create two files based on the first column.
>
> Regards,
> Dirk
> --
> -- Dirk Gunsthoevel IT Systemanalyse phone: +49 (0)251 28446-0
> -- Hammer Str. 13 fax: +49 (0)251 28446-55
> -- D-48153 Muenster http://www.GunCon.de/
> -- "Toto, I don't think we're in Kansas anymore..."
>
> "Andy Kent" <andykent.bristol1095@virgin.net> schrieb im Newsbeitrag
> news:1105448438.466964.207560@f14g2000cwb.googlegr oups.com...
> > I am writing a purge routine on a large table where we need to keep a
> > copy of the purged rows. We also need to re-size the table extents. So
> > the best bet would be to produce two ASCII files. Because of the size
> > of the table I'd like to make as few passes through it as possible -
> > one would be good.
> >
> > In an ideal world I'd have a piece of code that did something like
> >
> > foreach row
> > if <to keep>
> > unload (append) to ascii file1
> > else
> > unload (append) to ascii file2
> > end foreach
> >
> > In an even more ideal world I'd do the whole thing in SPL. (HPL would
> > be even better but the criteria are rather verbose and may need to be
> > held in a memory-resident table).
> >
> > Given that we don't live in such Utopian conditions, what approach
> > would give the best compromise?
> >
> > Thanks
> >
> > Andy Kent