This algorithm collects the words of a dictionary identical to a given words, one letter excepted.
To achieve a faster speed, the dictionnary is transformed into a tree of nodes containing the words letter by letter, as such:
{
"a": {
"b": {
"c": {
"d": {