Mailing-List: contact derby-dev-help@db.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Derby Development" <derby-dev@db.apache.org>
Received-SPF: pass (hermes.apache.org: local policy)
Date: Wed, 05 Jan 2005 15:10:26 +0530
From: Amit Handa <Amit.Handa@Sun.COM>
Subject: Re: Position of the ResultSet after updateRow?
In-reply-to: <41DBB248.5080205@Sun.COM>
To: Derby Development <derby-dev@db.apache.org>
Reply-to: Amit.Handa@Sun.COM
Message-id: <41DBB60A.4000509@Sun.COM>
MIME-version: 1.0
Content-type: text/plain; charset=us-ascii
Content-transfer-encoding: 7BIT
User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.4) Gecko/20040414
References: <41DABA6C.7F334B5C@Remulak.Net> <41DAE0A6.3030808@debrunners.com>
 <41DBB248.5080205@Sun.COM>

Amit Handa wrote:
> First the Spec does not mandate that
> a ResultSet should see it's own updates or inserts.
> (but the changes should be reflected to the database irrespective)
> 
> Secondly if the DBMS or JDBC drivers
> support this(through  DatabaseMetaData.ownXXXAreVisible() ),
> it should.
> 
> 
>>>My question is about ResultSet position after updateRow. An updateRow
>>>can make the row not qualify for the original select query.
> 

Oops...re phrasing below para

 It should qualify irrespective of whether it
 satisfies the ResultSet query from which it was populated or not
 or the data should not be allowed to enter the ResultSet.

> I agree with Dan it will be part of ResultSet
> even though subsequent executions may not fetch that row.
> 
> thanks,
> Amit
> 
> Daniel John Debrunner wrote:
> 
>>-----BEGIN PGP SIGNED MESSAGE-----
>>Hash: SHA1
>>
>>Mamta Satoor wrote:
>>
>>
>>
>>>Hi,
>>>
>>>Some of you might recall the discussion on where the ResultSet should be
>>>positioned after deleteRow and the conclusion that it will be placed right
>>>before the next row in the ResultSet.
>>>
>>>My question is about ResultSet position after updateRow. An updateRow
>>>can make the row not qualify for the original select query.
>>
>>
>>Can anyone from the Sun JDBC team comment on how updateRow() and
>>insertRow() are intended to behave when they modify or create rows that
>>would not qualify according to the SQL query that created the ResultSet?
>>
> 
> 
>>Reading between the lines in the JDBC book seems to imply that the
>>updated or inserted row are always part of the ResultSet, even though a
>>subsequent execution would not return that row.
>>I can see this approach makes sense for going through a list of items in
>>an application. Say all items with a price <= 10.0, and modifying the
>>price of some to be > 10.0, as a user I might expect to see the newly
>>modified row remain in my view, rather than just disappear.
>>
>>Dan.
>>
>>-----BEGIN PGP SIGNATURE-----
>>Version: GnuPG v1.2.5 (MingW32)
>>Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>>
>>iD8DBQFB2uCmIv0S4qsbfuQRAtagAJwOkXd7mu0nLteBa26X4tqGY324RgCdE8ld
>>BXWcUymBmC0cK0Yh9dwPE84=
>>=YSr2
>>-----END PGP SIGNATURE-----
>>
> 
> 
> 
> 
> 
> 


-- 
Regards,

Amit Handa,
JDBC Engineering, Java Web Services.
Sun Microsystems Inc.