From: Ahti Legonkov <lego@127.0.0.1>
Subject: Re: sql question:
Date: Sun, 14 Jul 2002 05:38:45 +0300
Lines: 35
Message-ID: <3D30E435.4070208@127.0.0.1>
References: <sM1Y8.89790$Bt1.4501578@bin5.nnrp.aus1.giganews.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:1.0.0) Gecko/20020530
To: pgsql-sql@postgresql.org

Chris Aiello wrote:
> hi all:
> 
> I'm trying to figure out SQL to do the following:
> I have an application that tracks SQL that is being sent to the database,
> and one of it's features is the ability to identify whether a query is an
> insert, update, delete, select, select with all rows returned, the query is
> the first in a user session....and many other criteria.   Because of the
> nature of SQL, i.e. many of the above could be true, the deisgners made each
> flag a 'bit'.  So an example is:
> 4 is a select
> 8 is insert
> 16 is update
> 32 is first query in session
> 64  is delete
> 128 is a cancelled query
> 256 is database cancelled query
> 
> 
> 
> Now the SQL that I have to find is 'which of these records is a delete?'
> The values could be 64, 96, 416, 445, 320 and many others.   All in all
> there are probably 20 possible values and the permutations are to lengthy to
> put in a 'like', so I need some kind of algorithm.   Does anyone have any
> ideas?

The algorithm is as simple as that:

if (value & 64 == 64) {
	// it is a delete
}

-- 
Ahti Legonkov