Postgresql – How to define a parameter that will be used as an IN argument inside a function

functionsparameterpostgresqlpostgresql-10

I want to make a function that receives two parameters. The first one represents the salaries from a group of employees, and second one, the codes from a group of departments. Both, P_IN_SALARIES and P_IN_DEPARTMENTS_CODES parameters, will be used as
arguments in an IN function of a query, just as demonstrated in the code below:

CREATE OR REPLACE FUNCTION public.get_employees_id(P_IN_SALARIES WHICH_TYPE_COMES_HERE, P_IN_DEPARTMENTS_CODES WHICH_TYPE_COMES_HERE)
 RETURNS text
 LANGUAGE plpgsql
AS $function$
declare
    v_employees_ids text;
begin

    select STRING_AGG(employee.id || '', ',') into v_employees_ids
    from employee 
    inner join departament on department.id = employee.departament_id
    where employee.salary in (P_IN_SALARIES)
    and department.code in (P_IN_DEPARTMENTS_CODES);

    RETURN v_employees_ids;
END;
$function$

What is the type of a IN parameter in a SQL statement?
Is there a generic one that I might use in order to allow a kind of portability on an occasional database exchange? (e.g. to Oracle)

Best Answer

Use an array and the ANY operator:

CREATE FUNCTION public.get_employees_id(P_IN_SALARIES int[], P_IN_DEPARTMENTS_CODES text[])
...
    select STRING_AGG(employee.id || '', ',') into v_employees_ids
    from employee 
    inner join departament on department.id = employee.departament_id
    where employee.salary = any (P_IN_SALARIES)
    and department.code = any (P_IN_DEPARTMENTS_CODES);
...

When calling the function you can use any array notation you like, I prefer array[...]:

select get_employees_id(array[100,200], array['code1', 'code2'])
Related Question