Search table for rows with combination of N columns matching any from the provided list

I have a COFFEE_CONSUMPTION table with 5 columns in Postgres. It can grow up to e.g. 1 million rows and more.

EMPLOYEE_ID|ROOM_ID|DDAY      |COUNT|TYPE           |
-----------------------------------------------------
1          |1      |2023-02-16|1    |latte          |
1          |1      |2023-02-16|3    |espresso       |
2          |1      |2023-02-16|2    |latte          |
3          |2      |2023-02-16|1    |espresso       |
4          |2      |2023-02-17|3    |frappuccino    |
...        |...    |...       |...  |...            |

We need to search for records with EMPLOYEE_ID+DAY+TYPE combination matching one of those from the provided list.

Looking for an index-wise implementation I came up with such a solution:

1. Created 2 immutable functions

    CREATE OR REPLACE FUNCTION C_CONCAT(text, VARIADIC text[])
    RETURNS text
    LANGUAGE sql
    IMMUTABLE PARALLEL SAFE AS
        'SELECT array_to_string($2, $1)';
   
    CREATE FUNCTION C_TO_CHAR(date) RETURNS text AS
    $$
        select to_char($1, 'YYYY-MM-DD');
    $$
    LANGUAGE sql immutable;
   

2. Added index to the table:

    CREATE INDEX COFFEE_CONSUMPTION_IDX 
        ON COFFEE_CONSUMPTION (C_CONCAT('', EMPLOYEE_ID, C_TO_CHAR(DDAY), TYPE));
   

3. Wrote a query:

    SELECT * 
    FROM COFFEE_CONSUMPTION CC 
    WHERE C_CONCAT('', CC.EMPLOYEE_ID, C_TO_CHAR(CC.DDAY), CC.TYPE) IN ?
   

Any ideas on how to improve this and/or get rid of custom functions?

One of possible approaches would be to create a multi-column index, populate a temporary table with field combinations and inner join in select (as described here). But I believe this would be more verbose, more complex from client programming (e.g. JPA) perspective and equally performant.

UPD: the main problem turned out to be not the SQL query but JPA+Hibernate stack which doesn't allow querying by list of lists. So I decided to stick to the functions solution for now.

I don't see why you need that complication with a custom function:

create index on coffee_consumption (employee_id, dday, type);

That index should be used if you run:

select *
from coffee_consumption cc
where (cc.employee_id, cc.dday, cc.type) in (?,?,?) 

or

select *
from coffee_consumption cc
where cc.employee_id = ? 
      cc.dday = ? 
      cc.type = ?