On 05/08/2018 05:39 AM, Philipp Kraus wrote:
Hello,

I have got a function with this definition:

CREATE OR REPLACE FUNCTION vectorize(refcursor)
     RETURNS SETOF refcursor
     LANGUAGE 'plpgsql'
     COST 100
     STABLE
     ROWS 1000
AS $BODY$

begin
        perform pivottable(
                '_vector',
                'select * from viewdata',
                array['ideate', 'name', 'description', 'latitude', 'longitude'],
                array['parametername'],
                array['parametervalue::text', 'parametervaluetext']
        );      
        open $1 scroll for select * from _vector;
        return next $1;
end

$BODY$;

The perform call creates a dynamic column pivot table, if I run manually

select pivottable(
        '_vector',
        'select * from viewdata',
        array['ideate', 'name', 'description', 'latitude', 'longitude'],
        array['parametername'],
        array['parametervalue::text', 'parametervaluetext']
);      
select * from _vector;

I get all the data in the output, so everything is fine.

My goal is now to encapsulate the two lines into a function, so I define
a stable function and based on the dynamic column set a cursor. I get in
pgadmin the column names back, but the rows are empty if I run:

select * from vectorize('myvec');
fetch all from myvec;

Can you explain me, which part is wrong?

I am going to say:

perform pivottable( ...

https://www.postgresql.org/docs/10/static/plpgsql-statements.html#PLPGSQL-STATEMENTS-SQL-NORESULT

"Sometimes it is useful to evaluate an expression or SELECT query but discard the result, for example when calling a function that has side-effects but no useful result value. To do this in PL/pgSQL, use the PERFORM statement:

PERFORM query;

This executes query and discards the result.  ..."
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^



Thanks

Phil






--
Adrian Klaver
adrian.kla...@aklaver.com

Reply via email to