From: aaron_watters at my-dejanews.com (aaron_watters at my-dejanews.com)
Date: Thu, 08 Apr 1999 15:15:30 GMT
Subject: GadFly - MemoryError
References: <Pine.SOL2.3.96.SK.990403182611.18369K-100000@sun.med.ru> <370B5E83.C796BFF3@palladion.com>
Message-ID: <7eih6b$ck2$1@nnrp1.dejanews.com>
Content-Length: 1636
X-UID: 144                                                  

It would be interesting to try this, but my guess is that this would be
slower than the first query since equalities are optimized and "in" is not. 
I hope oleg is using gadfly 1.0 too (not beta 0.2 or whatever).  -- Aaron
Watters

In article <370B5E83.C796BFF3 at palladion.com>,
  Tres Seaver <tseaver at palladion.com> wrote:
> Oleg Broytmann wrote:
> >
> > Hello!
> >
> >    I tried to add yeat another database backend to my project "Bookmarks
> > database". My database contains now about 3000 URLs, not too much, I think.
> > I subclass by BookmarksParser to parse bookmarks.html into gadfly database
> > and got a database of 500 Kbytes - very small database, I hope.
> >    Then I tried to find duplicates (there are duplicates). I ran the query:
> >
> > SELECT b1.rec_no, b2.rec_no, b1.URL
> >    FROM bookmarks b1, bookmarks b2
> > WHERE b1.URL = b2.URL
> > AND   b1.rec_no < b2.rec_no
>
> How many duplicates are there?  Something like
>
>   SELECT URL FROM bookmarks GROUP BY URL HAVING COUNT(*) > 1
>
> will produce the URL's with duplicates;  you could then do
>
>   SELECT rec_no, URL FROM bookmarks
>     WHERE URL IN
>       (SELECT URL FROM bookmarks GROUP BY URL HAVING COUNT(*) > 1)
>
> or create a temp table first with the results of the subquery, then join it in a
> separate query.
> --
> =========================================================
> Tres Seaver         tseaver at palladion.com    713-523-6582
> Palladion Software  http://www.palladion.com
>

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own