Обсуждение: newbie question - delete before insert

Поиск
Список
Период
Сортировка

newbie question - delete before insert

От
Grant Mckenzie
Дата:
Hi,

a question that I imagine is a faq but have not been able to find much help.

How do people implement insert or upate ( otherwise known as upsert ) behaviour in postgres i.e. insert a row if it's
keydoes not exist in the database else update the existing row? 

I tried using an insert rule to delete any existing rows first then insert however this leads to infinitely recursive
rules( which postgres properly rejects. ) 

I'm guesisng that the most sensible approach would be a stored proc/function?

Cheers
G.




Re: newbie question - delete before insert

От
Thomas Kellerer
Дата:
Grant Mckenzie wrote on 20.11.2010 07:00:
> How do people implement insert or upate ( otherwise known as upsert )
> behaviour in postgres i.e. insert a row if it's key does not exist in
> the database else update the existing row?
>

You can simply send the UPDATE, if nothing was updated, it's safe to send the INSERT

Regards
Thomas


Re: newbie question - delete before insert

От
Ashish Karalkar
Дата:
On 11/20/2010 02:43 PM, Thomas Kellerer wrote:
Grant Mckenzie wrote on 20.11.2010 07:00:
How do people implement insert or upate ( otherwise known as upsert )
behaviour in postgres i.e. insert a row if it's key does not exist in
the database else update the existing row?


You can simply send the UPDATE, if nothing was updated, it's safe to send the INSERT

Regards
Thomas



something on the line of ....

http://www.postgresql.org/docs/8.4/interactive/plpgsql-control-structures.html

Example 38-2. Exceptions with UPDATE/INSERT

-- 


With Regards
Ashish Karalkar

Re: newbie question - delete before insert

От
Sim Zacks
Дата:
> How do people implement insert or upate ( otherwise known as upsert ) behaviour in postgres i.e. insert a row if it's
keydoes not exist in the database else update the existing row? 
>
> I tried using an insert rule to delete any existing rows first then insert however this leads to infinitely recursive
rules( which postgres properly rejects. ) 
>
> I'm guesisng that the most sensible approach would be a stored proc/function?
I've done it with an on insert trigger. The where clause contains the
values that make this row unique. If it finds another row with the same
fields, it doe san update instead of the insert:

CREATE OR REPLACE FUNCTION stock.trg_beforeinsertstock()
   RETURNS "trigger" AS
$BODY$
declare
     v_stockid int;
begin
     select stockid into v_stockid from stock where pnid=new.pnid
             and ownerid=new.ownerid and
coalesce(stocklocationid,-1)=coalesce(new.stocklocationid,-1)
             and coalesce(batchid,-1)=coalesce(new.batchid,-1);
     if v_stockid is not null then
         Update stock set stock=stock+new.stock where stockid=v_stockid;
         return null;
     else
         return new;
     end if;
end;
$BODY$
   LANGUAGE plpgsql VOLATILE;