UCS-2 encoding. Record base

Question

PHP

Friends. I don't even know how to ask the right question. Most likely only GURU will help.
In general, I'm digging an interesting engine, very similar to toster.ru. That's why I'm digging.
A friend asked me to make layout. But in the process I noticed one interesting thing.
If the question contains Latin characters, the question is split into keywords and written into the database. At the same time, search, related news, meta keywords process these values ​​and give a result based on them. But here's the problem, when the question is written in Cyrillic, no values ​​get into the database. Only NULL.
I am not very strong in php. As I understand it, everything happens in the file, which was uploaded here dropmefiles.com/LmhWD

Answer 1

a-z add.
And why is the encoding ucs-2 and not utf-8?