"Test String" would result in NN(Norwegian Nynorsk?) detection

[AutisticShark]

I am doing a translator project that involves automatic language detection, I set up the lingua lib as instructed, however, it exhibited strange behavior during testing.

Here is part of my code:

    if source_language == "auto":
        detector = LanguageDetectorBuilder.from_all_languages().with_preloaded_language_models().build()
        source_language = detector.detect_language_of(text).iso_code_639_1.name.lower()
        print(source_language)
        if source_language not in ['en', 'zh','ja','ko','fr','de','es','it','pt','ru','tr']:
            return "Unsupported language"

And here is the console log:

test string auto zh
nn
Unsupported language
Successful Response auto zh
en
成功回应
Test String auto zh
nn
Unsupported language
This is a Test String auto zh
en
这是一个测试弦。
A Test String auto zh
nn
Unsupported language

Note the lib can successfully detect Successful Response and This is a Test String as English, but it failed to detect test string, Test String, and A Test String.

Is this some kind of edge case, or I am doing something wrong here? Please advise, thanks.