The literature on finite state automata generally assumes a relatively small alphabet, often 128 or fewer characters. Small alphabets allow one to implement an FSA efficiently and in a straight-forward manner.

Large alphabets (of which Unicode is a prime example) can make the efficient and compact implementation of automata difficult. This talk presents the problems encountered when handling large alphabets in an FSA implementation and describes some methods to handle them.

The presentation presumes some knowledge of automata, though time will be spent in the beginning to present a brief overview of the necessary theory.