Benoit . ha scritto:
The problem is that the "9" vertex becomes black "prematurely", and it depends only from the traversing algorithm (BFS in this case). I'm searching a method to customize the traversing algorithm, but I'm pessimist. The coloration method is cabled in the algorithm.
I am still not sure i understand your problem correctly but i still want to help a bit. You're clearly right, the coloration method is cabled in the algorithm. BUT the color property isn't. You may decide to create a color property that does what you want (possibly returning "white" after being set to "black"?) . This is entirely up to you.
Hope this helps !
Thanks Benoit, yes, it's what I will make. Now the question is: which is the right moment in which to turn a vertex color from black to white, without to risk to distort the sense of the Dijkstra's algorithm? I've tried a pair of strategies, but I must think a little on the question... I would want to be sure to avoid unexplored paths or to create situations of inconsistent vertex states. I hope to post a solution soon... Regards, Cosimo Calabrese