Question

What we have:

3 MySQL DB tables: user, text, friend

user: username, password, email, etc.

text: username, text, date, etc.

friend: username, friend_username, etc.

Task:

Write an algorithm (in Java) to show 10 latest texts from your friends.

Ultimate target is to have running time within O(n log n).

DB tables can be modified (added new ones) as required.

Data volume: 200 000 users, ~50 text per user.

I would appreciate any ideas, examples, points, etc. Thank you!

(Not a homework. Pure problem, looking for performance improvement)

Answer 1

You didn’t say what database, so I’ll just make an assumption (Postgres syntax below):

select t.* from text t
    inner join friend f ON f.friend_username = t.username
    where f.username = 'myusername'
    order by t.date desc
    LIMIT 10