protected internal override PuzzleState popFrontier()
{
//remove a state from the top of the fringe so that it can be searched.
PuzzleState lState = Frontier.RemoveFirst();
//add it to the list of searched states so that duplicates are recognised.
Searched.AddLast(lState);
return(lState);
}
protected internal override PuzzleState popFrontier()
{
//remove an item from the fringe to be searched
PuzzleState thisState = Frontier.First.Value;
Frontier.RemoveFirst();
//Add it to the list of searched states, so that it isn't searched again
Searched.AddLast(thisState);
return(thisState);
}
public override bool addToFrontier(PuzzleState aState)
{
//We only want to add the new state to the fringe if it doesn't exist
// in the fringe or the searched list.
if (Searched.Contains(aState) || Frontier.Contains(aState))
{
return(false);
}
else
{
Frontier.AddLast(aState);
return(true);
}
}
public override bool addToFrontier(PuzzleState aState)
{
//if this state has been found before,
if (Searched.Contains(aState) || Frontier.Contains(aState))
{
//discard it
return(false);
}
else
{
//else put this item on the end of the queue;
Frontier.AddLast(aState);
return(true);
}
}
public override direction[] Solve(nPuzzle aPuzzle)
{
//keep searching the fringe until it's empty.
//Items are "popped" from the fringe in order of lowest heuristic value.
//Add the start state to the fringe
addToFrontier(aPuzzle.StartState);
while (Frontier.Count > 0)
{
//get the next State
PuzzleState thisState = popFrontier();
//is this the goal state?
if (thisState.Equals(aPuzzle.GoalState))
{
return(thisState.GetPathToState());
}
//not the goal state, explore this node
List <PuzzleState> newStates = thisState.explore();
for (int i = 0; i < newStates.Count; i++)
{
PuzzleState newChild = newStates[i];
//if you can add these new states to the fringe
if (addToFrontier(newChild))
{
//then, work out it's heuristic value
newChild.HeuristicValue = HeuristicValue(newStates[i], aPuzzle.GoalState);
newChild.EvaluationFunction = newChild.HeuristicValue;
}
}
//Sort the fringe by it's Heuristic Value. The PuzzleComparator uses each nodes EvaluationFunction
// to determine a node's value, based on another. The sort method uses this to sort the Fringe.
//
// TODO: is this the correct way to sort the frontier as specified in the Assignment:
// When all else is equal, nodes should be expanded according to the following order:
// the agent should try to move the empty cell UP before attempting LEFT, before
// attempting DOWN, before attempting RIGHT, in that order.
Frontier.Sort(new PuzzleComparator());
}
//no more nodes and no path found?
return(null);
}
