[Nauty] Nauty implementation files

[Edward Turner]

I would like to implement in prolog the algorithm for computing a canonical
form of a graph - to aid in detecting isomorphic graphs (as part of our
symmetry detection module of our ProB Model Checker). We understand that
this implementation language will not be as efficient as the C
implementation of nauty (due to lack of certain data structures), but
believe that the overhead of interfacing prolog with C thousands of times
per second (or more) would outweigh this.

Currently I am reading "Practical Graph Isomorphism", which Brendan McKay
directed me to (thanks!) but am finding it fairly hard going (having a
Formal Methods background). Is there a more practical description of how
this algorithm works that anyone could point me towards? If anyone has any
more help that would be much appreciated.

Best,

Edd.