Hashing table

It consists in implementing a list of triples in $Q \times \Sigma \times Q$using a hash table mapping couples of $Q \times \Sigma$ on P(Q). This is a good trade-off between speed and memory : transition access is performed in almost constant time (depending on the quality of the hashing function) and memory use is proportional to the number of transitions. It is recommended anyway to have an idea of the number of transitions beforehand to avoid reallocation during computation (see table 5).