Being able to get not only the "most possible" language but an ordered list of the languages and their respective score

[allan-simon]

Internally the detector rank the languages using two technic

absolute ranking (which use a score based on the absolute number of time n-grams appear in the database for a given language, i.e English have more chances to get high score than Shanghainese for exemple)

relative ranking which use a score based on the proportion of the n-gram in that language , regardless of how many sentences this language has

so it would be good to be able to expose this to the user like

{
"absolute" : {
"eng" : 67
"fra" : 42,

 },
 "relative" :  {
       "fra" : 350
       "eng" :  34 ,
      "epo" :  6
 }

}

This way they'll be able to do more computation on their side (i.e testing a better detection algorithm etc.)