Question

I’m trying to write a function to traverse a tree with depth first search.

My current algorithm goes something like:

If children
 go to first child

If no children
 go to next sibling

If no siblings
 go to parent

The problem I’m running into is that I can’t mark nodes on the tree as having been visited, so when I go to the parent the cycle just resets and it goes to the child again, getting stuck in a loop. Does anyone have any idea as to how I could solve this?

(It’s in java using the ANTLR plugin)

EDIT:

Following one of the suggestions I wrote this:

public void traverseTree(Tree tree){

    if (tree.getChildCount() > 0){
        tree = tree.getChild(0);
        traverseTree(tree);
        System.out.println(tree.toString());
    }
    if (tree.getParent().getChild(tree.getChildIndex() + 1) != null){
        tree = tree.getParent().getChild(tree.getChildIndex() + 1);
        traverseTree(tree);
        System.out.println(tree.toString());
    }
    if (!tree.getParent().toString().contains("ROOT_NODE")){
        tree = tree.getParent();
        traverseTree(tree);
        System.out.println(tree.toString());
    }
}

Root node is the name of the root node, but I’m getting a stack overflow error. Anyone have any idea why?

Thanks.

Answer 1

I would use recursion in this case.

class Node {
   public List<Node> getChildren() { .... }

   public void traverse(Visitor<Node> visitor) {
      // If children
      // go to first child - by traversing the children first.
       for(Node kid: getChildren())
           kid.traverse(visitor);
           // If no children
           //  go to next sibling, - by continuing the loop.

       visitor.visit(this);
       // If no siblings
       // go to parent - by returning and letting the parent be processed
   }
}


interface Vistor<N> {
   public void visit(N n);
}