Re: if-clause to an exiting statement
От | Rob Sargent |
---|---|
Тема | Re: if-clause to an exiting statement |
Дата | |
Msg-id | 4D0B5FBE.6010708@gmail.com обсуждение исходный текст |
Ответ на | Re: if-clause to an exiting statement (Jasen Betts <jasen@xnet.co.nz>) |
Список | pgsql-general |
Jasen Betts wrote: > On 2010-12-07, Kobi Biton <kobi@comns.co.il> wrote: > >> hi i am a newbie to sql statments , I am running postgres 8.1 with >> application called opennms version 1.8.5 due to an application bug >> queries that I execute aginst the DB which returns raw-count=0 are being >> ignored and will not process a certain trigger I need to process. >> > > I think you want this: > > ORIGINAL QUERY > union > select DUMMY ROW DATA > where > not exists ( ORIGINAL QUERY ) > > > you need to return something to get a rowcount of 1 this is what the > dummy row data provides. the "where not exists" part blocks the dummy > row data when the main query returns something. > > Simple enough, but I suspect it runs the same query twice, so I hope it's not to expensive. I wonder what the app is/was doing with the vacuous single row or will do with the dummy data row? Seems the app/trigger simply needs to know the execution of the query was successful irrespective of the actual row count - or is that the bug with "1.8.5"?
В списке pgsql-general по дате отправления: