Wouldn't it make more sense for the switch to pattern 2 prefix matching to ignore the repeated characters? So if the user typed LLA the prefix matched would be LA (ignore the first L), not LLA.