I have a Postgres table users (E_8_User) which contains User_id as a primary key, Boss as a foreign key to the same table (it is not nullable, if some user doesn't have boss, its Boss attribute = user_id). I need to get their bosses for all users in the table, so I'm trying to write CTE query:
WITH RECURSIVE herarchy_reports AS (
SELECT E_8_User.User_id, E_8_User.Boss, 1 as lvl, E_8_User.User_id as RootUserID
FROM E_8_User
WHERE E_8_User.User_id=E_8_User.Boss
UNION ALL
SELECT usr.User_id, usr.Boss, lvl+1 as lvl, rep.RootUserID
FROM herarchy_reports as rep JOIN E_8_User as usr ON
rep.user_id=usr.Boss
)
SELECT * FROM herarchy_reports ORDER BY RootUserID;
But it doesn't work: DB is constantly performing query. What am I doing wrong?
That's a typical recusive query:
with recursive cte as (
select u.user_id, u.boss, 1 as lvl from e_8_user u
union all
select u.user_id, c.boss, lvl + 1
from cte c
inner join e_8_user u on u.boss = c.user_id and u.user_id != c.user_id
)
select user_id, boss
from cte c
where lvl = (select max(c1.lvl) from cte c1 where c1.user_id = c.user_id)
order by user_id
In the recursive query, the trick is to stop recursing when a record is joined with itself (u.boss = c.user_id and u.user_id != c.user_id
).
Then, in the outer query, you want to select the record that has the greatest level for each user.
Assuming the following sample data:
user_id | boss ------: | ---: 1 | 1 2 | 1 3 | 2 4 | 3 5 | 2 6 | 6 7 | 6
The query produces:
user_id | boss ------: | ---: 1 | 1 2 | 1 3 | 1 4 | 1 5 | 1 6 | 6 7 | 6
In Postgres, we can simplify the outer query with distinct on
:
with recursive cte as (
select u.user_id, u.boss, 1 as lvl from e_8_user u
union all
select u.user_id, c.boss, lvl + 1
from cte c
inner join e_8_user u on u.boss = c.user_id and u.user_id != c.user_id
)
select distinct on (user_id) user_id, boss
from cte c
order by user_id, lvl desc