Recognize the primary alphabet for different alphabets per text

[JohannesRie]

Hi,

I didn't know exactly where to post it. It's not a problem as such, just an addition.

For an offline translator based on Meta AI's NLLB, I use this detector in conjunction with langdetect, among other things, to recognize text modules if no source language is specified. This occasionally causes problems if Latin characters are present in a predominantly Arabic, Cyrillic or other alphabet, as all alphabets are returned by the detection. The following code provides a remedy if the text is predominantly written in one alphabet:

def alphabet_detect(text : str) -> str:
    ad = AlphabetDetector()
    result = ad.detect_alphabet(u'{}'.format(text))
    if len(result) > 1:
        result = []
        for word in text:
            alpha = ad.detect_alphabet(u'{}'.format(word))
            if len(alpha) == 1:
                result.append(next(iter(alpha)))
            alphabet_with_highest_count = max(set(result), key=result.count) 
        return(alphabet_with_highest_count)
    else:
        return next(iter(result))