In this task, use StreamTokenizer instead Scanner, otherwise will be TLE

Thank you for this tip. It helped me a lot.
I also didn't manage to implement bubble sort (as suggested in the problem tip), because I was getting TLE on test #12. I used a map with solved problem count as keys and team id lists as values instead.

Thanks for the tip. Seems unfair we have to jump through hoops with java/scala as they are slower on a single run.