Question

This one is probably a softball question for any DBA, but here’s my challenge. I have a table that looks like this:

id     parent_id    active
---    ---------    -------
1      5            y
2      6            y
3      6            y
4      6            y
5      7            y
6      8            y

The way the system I am working on operates, it should only have one active row per parent. Thus, it’d be ok if ID #2 and #3 were active = ‘n’.

I need to run a query that finds all rows that have duplicate parent_ids who are active and flip all but the highest ID to active = ‘y’.

Can this be done in a single query, or do I have to write a script for it? (Using Postgresql, btw)

Answer 1

ANSI style:

update table set
    active = 'n'
where
    id <> (select max(id) from table t1 where t1.parent_id = table.parent_id)

Postgres specific:

update t1 set
    active = 'n'
from
    table t1
    inner join (select max(id) as topId, parent_id from table group by parent_id) t2 on
        t1.id < t2.topId
        and t1.parent_id = t2.parent_id

The second one is probably a bit faster, since it’s not doing a correlated subquery for each row. Enjoy!