one spoken language can be written in multiple scripts, so it will be detected as a different alphabet/language
look at the alphabet.rs to understand what languages have already defined alphabets. Some of them need validation
Separates words in text (from iterator CharIndices), and detects language of words by used alphabets (chars).
Warning: can return words with chars from the Unicode private area (for example Lingala, Nuer or Yoruba languages)