Hello guys,
I'm a problem with dynamic sql.
I am trying to write a generic function that is able to read and update a table based on some data coming from e previous record.
Here the example
_sqlStr=format('select *
from %1$s.%2$s
where (' || array_to_string(_activeRec.pk_columns_list, ',') || ') in (select ' ||
'row($1[''' || array_to_string(_activeRec.pk_columns_list, '''],$1[''') || ''']))'
, _activeRec.name_of_schema, _activeRec.main_table);
execute _sqlStr using oldRec into _rec;
from %1$s.%2$s
where (' || array_to_string(_activeRec.pk_columns_list, ',') || ') in (select ' ||
'row($1[''' || array_to_string(_activeRec.pk_columns_list, '''],$1[''') || ''']))'
, _activeRec.name_of_schema, _activeRec.main_table);
execute _sqlStr using oldRec into _rec;
My problem is oldRec is a type record, so the substitution performed by execute fails, because it can't recognize the field if the variable is record and not a specific composite record type.
I suppose this is a recurrent question, but I can't find a solution...
Domenico L.
