Re: PostgreSQL equivalent to Oracles ANYDATASET

[Dirk Mika <Dirk.Mika@xxxxxxxxxxxxx>]

Hi,

 

That looks promising. I'll take a closer look at that.

 

Regards

Dirk

 

--
Dirk Mika
Software Developer

mika:timing GmbH
Strundepark - Kürtener Str. 11b
51465 Bergisch Gladbach
Germany

fon +49 2202 2401-1197
dirk.mika@xxxxxxxxxxxxx
www.mikatiming.de

AG Köln HRB 47509 * WEEE-Reg.-Nr. DE 90029884
Geschäftsführer: Harald Mika, Jörg Mika

Von: Pavel Stehule <pavel.stehule@xxxxxxxxx>
Datum: Freitag, 13. November 2020 um 18:05
An: Dirk Mika <Dirk.Mika@xxxxxxxxxxxxx>
Cc: "pgsql-general@xxxxxxxxxxxxxxxxxxxx" <pgsql-general@xxxxxxxxxxxxxxxxxxxx>
Betreff: Re: PostgreSQL equivalent to Oracles ANYDATASET

 

Hi

 

pá 13. 11. 2020 v 17:57 odesílatel Dirk Mika <Dirk.Mika@xxxxxxxxxxxxx> napsal:

Hello,

 

we’re currently using Oracle and have a few functions in some packages, which are defined e.g. as follows:

 

   FUNCTION get_results(p_id_series series.id_series%TYPE)

      RETURN ANYDATASET

      PIPELINED USING t_series_get_results;

 

It can be called like:

 

SELECT * FROM TABLE(series_pkg.get_results(1));

 

The purpose of this function is to provide a DATASET, which has different columns in the result depending on the passed parameter.

 

Is there any way to achieve something similar in PostreSQL?

 

Currently there is not this possibility

 

Theoretically you can use a function, that returns refcursor, and outside function you can read this cursor

 

http://okbob.blogspot.com/2008/08/using-cursors-for-generating-cross.html

 

Regards

 

Pavel

 

BR

Dirk

--
Dirk Mika
Software Developer

mika:timing GmbH
Strundepark - Kürtener Str. 11b
51465 Bergisch Gladbach
Germany

fon +49 2202 2401-1197
dirk.mika@xxxxxxxxxxxxx
www.mikatiming.de

AG Köln HRB 47509 * WEEE-Reg.-Nr. DE 90029884
Geschäftsführer: Harald Mika, Jörg Mika