Possible feature that could be added is having a uk and us keyboard layout in the adjacency graph. At the moment doing !"£$%^&*() gets a score greater than 1 which is probably wrong because its a fairly simple sequence.
I forked it and tried to implement it myself but the graph generation script doesn't play nice with the £ sign because its a unicode character and all the other characters are ASCII (i.e., when doing len(3£) it returns 3 instead of 2 because the £ is considered by bytes.
Another workaround for this could just be to add !"£$%^&*() to one of the dictionaries of penalised words.
Possible feature that could be added is having a uk and us keyboard layout in the adjacency graph. At the moment doing !"£$%^&*() gets a score greater than 1 which is probably wrong because its a fairly simple sequence.
I forked it and tried to implement it myself but the graph generation script doesn't play nice with the £ sign because its a unicode character and all the other characters are ASCII (i.e., when doing
len(3£)
it returns 3 instead of 2 because the £ is considered by bytes.Another workaround for this could just be to add !"£$%^&*() to one of the dictionaries of penalised words.