I've used BST but still cannot optimize it... similar code on C is accepted with 0.171s 300kb but Python is processing the code way longer and TLE #14. And no solutions on python so far. Is there something that can be done?