Detection accuracy

[vamseekm]

中国功夫是一门博大精深的武学艺术 , 中国功夫app , 介绍中国功夫的分类、特点、器材、门派等与中国功夫有关的内容！让广大读者能够更完整的了解中国功夫的精华！

If run the above snippet through your detection tool I am getting "en" as the answer. This is due to a three letter word in the snippet ("app"). Is it possible to fix this issue ?

[vamseekm]

Sorry my mistake, it was encoding issue. Thanks for the great tool.

[saffsd]

No worries. Indeed, my own check verifies that zh is correctly detected

中国功夫是一门博大精深的武学艺术 , 中国功夫app , 介绍中国功夫的分类、特点、器材、门派等与中国功夫有关的内容！让广大读者能够更完整的了解中国功夫的精华！
('zh', -1414.5709274662972)

Could you provide me some detail on how you used the tool? It would be good if I can detect potential encoding issues beforehand and try to address them, to make the tool as simple as possible for the end user.

[vamseekm]

I forgot to supply utf8 encoding option to mysql connection while trying to get some unicode text from db. So I was essentially passing garbled mess to langid.py and asking it to identify the language. Again langid.py is a great tool it was my stupid mistake. :)