As I said to @tiago
the purpose for this exercise is just fun.
I think that it would be hard implementation-wise but I was thinking a DFS that backtracked whenever the second different character was found. It would be hard keeping track of the number of different characters found for a given node but not impossible.
We're a place where coders share, stay up-to-date and grow their careers.
We strive for transparency and don't collect excess data.