X-Original-To: pgsql-php-postgresql.org@localhost.postgresql.org Received: from localhost (unknown [64.117.224.130]) by svr1.postgresql.org (Postfix) with ESMTP id 1ED03D1B9C3 for ; Thu, 21 Aug 2003 16:51:58 -0300 (ADT) Received: from svr1.postgresql.org ([64.117.224.193]) by localhost (neptune.hub.org [64.117.224.130]) (amavisd-new, port 10024) with ESMTP id 26393-04 for ; Thu, 21 Aug 2003 16:51:27 -0300 (ADT) Received: from mail.ez3pl.com (unknown [66.136.75.131]) by svr1.postgresql.org (Postfix) with ESMTP id B5C81D1B9B1 for ; Thu, 21 Aug 2003 16:51:10 -0300 (ADT) Received: from speedfc.com (djackson.espeedfulfillment.com [172.25.96.11]) by mail.ez3pl.com (8.11.2/8.11.2) with ESMTP id h7LJoKx11150; Thu, 21 Aug 2003 14:50:20 -0500 Message-ID: <3F452280.4090906@speedfc.com> Date: Thu, 21 Aug 2003 14:50:24 -0500 From: DeJuan Jackson User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030624 X-Accept-Language: en-us, en MIME-Version: 1.0 To: "Dave [Hawk-Systems]" Cc: pgsql-php@postgresql.org Subject: Re: select to detect overlapping timeframes References: In-Reply-To: Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-new at postgresql.org X-Archive-Number: 200308/94 X-Sequence-Number: 2271 (NOTE: I don't use outlook [Mozilla on a Linux box], but I'm gonna top post and fully quote) I'm not certain what you want but this might be an alternative to what you are trying to do (It might not perform better). SELECT DISTINCT t1.* FROM log t1, log t2 WHERE t1.date_time BETWEEN $dategt AND $datelt AND t2.date_time BETWEEN $dategt AND $datelt AND t1.name = '$username' AND t2.name = '$username' AND t1.date_time BETWEEN t2.date_time AND t2.date_time + t2.session_time ORDER BY t1.name ASC /* I don't see why you need to do this since you are looking at one name anyway */, t1.date_time DESC Tweak as needed for indexes and such. name='$username' AND \ date_time>$dategt AND date_time<$datelt Dave [Hawk-Systems] wrote: >looking for a better way to query and cehck for overlapping timeframes. > >am selecting records for display that has 3 fields of importance; > date_time (unix timestamp) > session_time (in seconds) > user_id (text field) > >Of all the sessions recorded, we are looking for a count of duplicate sessions >for a given month. currently we are doing this with two seperate calls, but it >is taking forever to process; > ># first query is to gather sessions that took place within a given month >$query = "SELECT * FROM logs WHERE name='$username' AND \ > date_time>$dategt AND date_time<$datelt \ > ORDER BY date_time DESC"; >$result = pg_exec($database,$query); >$numrows=pg_numrows($result); >for($count=0;$count<$numrows;$count++){ > $row = pg_fetch_array($result,$count); ># as we go through the hits, check for duplicates within that timeframe as well > $DUPquery = "SELECT session_id FROM logs WHERE name='$username' AND \ > date_time>$sessionstart AND \ > (date_time - session_time)<$datetime \ > ORDER BY name ASC, date_time DESC"; > $DUPresult=pg_exec($database,$DUPquery); > $DUPcount=pg_numrows($DUPresult); > > >Obviously this is extremely processor intensive, not to mention it seems sloppy. >Any better recommendations, either on altering the above code, or another way to >get the information we require? > >Dave > > > >---------------------------(end of broadcast)--------------------------- >TIP 8: explain analyze is your friend > >