Re: delete cascade
From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Mon, 23 Apr 2007 22:13:32 GMT
Message-ID: <gcaXh.27427$PV3.286287_at_ursa-nb00s0.nbnet.nb.ca>
>
> ...
>
>
> I had an example like this in mind:
>
> V = (invoices where invoice# = 99) <AND> (items where invoice# = 99);
> DELETE V;
That's still a compound delete where someone spells out the delete, or are you suggesting that V is a pre-existing view?
Date: Mon, 23 Apr 2007 22:13:32 GMT
Message-ID: <gcaXh.27427$PV3.286287_at_ursa-nb00s0.nbnet.nb.ca>
paul c wrote:
>> paul c wrote:
>
> ...
>
>>>> I mention it because it would fully disclose the data to be deleted, >>>> which is the invoice and the associated items. >>> >>> >>> >>> I can see that it would fully enclose, ie., enclose in one relation, >>> but not sure how it would disclose any more fully than two selections >>> and a join. >>> >>> Not trying to be picky but I found your view suggestion to be >>> profound and want to make sure I understand this one. >> >> >> >> I didn't refer to two selects and a join so I find your comment >> confusing. I was discussing a single delete that deletes both just as >> an "on delete cascade" trigger would delete both.
>
> I had an example like this in mind:
>
> V = (invoices where invoice# = 99) <AND> (items where invoice# = 99);
> DELETE V;
That's still a compound delete where someone spells out the delete, or are you suggesting that V is a pre-existing view?
If pre-existing, I would think the following more likely:
V = (invoices <AND> items);
DELETE V where invoice# = 99;
But the above has the problem that only invoices with items will appear
in the view, which is why I suggested an RVA. Received on Tue Apr 24 2007 - 00:13:32 CEST