How to classify natural languages written in other forms of characters?

Question

Background

I would like to classify all the three phrases as Chinese, \'zh\' using fastText.

["Ni hao!", \'你好!\', \'ni hao!\']