Обсуждение: Another problem with reverse engineering functions

Поиск
Список
Период
Сортировка

Another problem with reverse engineering functions

От
Kieran McCusker
Дата:
Hi

Sorry if this has been raised before but the following function will not 
be reverse engineered correctly - It will be changed into a setof return 
which will break anything expecting organisation_id.

CREATE OR REPLACE FUNCTION public.t() RETURNS TABLE (organisation_id 
integer) AS
$$ select 27;
$$ LANGUAGE 'sql';


In pgAdmin 10.2 it becomes :-

CREATE OR REPLACE FUNCTION public.t()  RETURNS SETOF integer AS
$BODY$ select 27;
$BODY$  LANGUAGE 'sql' VOLATILE  COST 100  ROWS 1000;

Many thanks

Kieran






Re: Another problem with reverse engineering functions

От
Guillaume Lelarge
Дата:
Le 07/04/2010 10:47, Kieran McCusker a écrit :
> [...]
> Sorry if this has been raised before but the following function will not
> be reverse engineered correctly - It will be changed into a setof return
> which will break anything expecting organisation_id.
> 

It's already fixed in the SVN version. See
http://code.pgadmin.org/trac/ticket/159 for details. It will be
available in 1.10.3.

Thanks for your report.


-- 
Guillaume.http://www.postgresqlfr.orghttp://dalibo.com