Re: REPOST: Trouble with SQL conversion
От | Richard Ellerbrock |
---|---|
Тема | Re: REPOST: Trouble with SQL conversion |
Дата | |
Msg-id | scb5b513.058@eskom.co.za обсуждение исходный текст |
Ответ на | REPOST: Trouble with SQL conversion ("Richard Ellerbrock" <richarde@eskom.co.za>) |
Список | pgsql-sql |
Ok, did that - now it works but returns no records. What I am after is to reduce the right hand table to one row matching the left table. There may be no, one or many matching records in the right table, but I want to group them on the key and find out the number of records in the right table at the same time. This works just fine in that other database :-) >>> Christopher Kings-Lynne <chriskl@familyhealth.com.au> 2002/04/11 03:28:32 >>> > SELECT base.baseindex, base.subnetsize, base.descrip, base.baseaddr, > base.admingrp, count(ipaddr.baseindex) AS cnt FROM base > LEFT JOIN ipaddr ON base.baseindex=ipaddr.baseindex WHERE > base.baseaddr > BETWEEN 2473473024 AND 2473473151 AND base.customer=1 > GROUP BY base.baseindex > ORDER BY base.baseaddr; > 0: ERROR: Attribute base.subnetsize must be GROUPed or used in an > aggregate function Do what it says. Add base.subnetsize to the GROUP BY. You'll probably need to add base.descrip, base.baseaddr and base.admingrp as well. Chris
В списке pgsql-sql по дате отправления: