N元树的二进制表示

Question

I am currently trying to make a tree structure with Nodes having an unknown amount of children but I also have to keep track of parent Nodes. I looked at this question N-ary trees in C and made a structure similar to that advised in the link:

template <class T>
struct treeNode {

public: 

T  * data;
treeNode *parent;
treeNode *kids; //left 
treeNode *siblings; //right


treeNode();
~treeNode();
treeNode(T data);
void treeInsert(T newItem);



};

It says that by making the tree in this way, it makes certain algorithms easier to code. However, I am having a difficult time figuring out how I would create insert() and search() methods for this structure, seeing that I need to keep track of parent nodes. Are there any suggestions as to how I may go about this?

EDIT:

Here is my insert() method:

template<class T>
bool NewTree<T>::insert( T *data, treeNode<T> * parent)
{
if(this->root == NULL)
{
    this->root = new treeNode<T>();
    this->root->data = data;
    this->root->parent = NULL;
    this->root->children = NULL;
}
else
{
    treeNode temp = new treenode<T>();
    temp.data = data;
    temp.parent = parent;
    parent->children->siblings = temp; // add node to siblings of parent's   children
    parent->children = temp; // add node to children of parent

}

}

Does this look correct?

Answer 1

With any tree structure, searching is going to use relatively the same algorithm (depth-first recursion if unsorted, or simple tree-searching if sorted (like binary search tree)). When you insert, all you need to do is assign the new node's .parent to the parent. Then assign the new node's .parent.child[1] to the new node (thus linking the parent to the child). Then, check the other children of the parent node to assign your new node's siblings (if any).

Okay, so here's some pseudocode (mostly like java -- sorry, it's what i've been writing today) that will implement node creation and the series of assignments to maintain it in a tree structure, using the second example in the link you provided:

(node source):

class Node {
  // generic data store
  public int data;
  public Node parent;
  public Node siblings;
  public Node children;
}

(tree source):

class NewTree {
  // current node
  public Node current;
  // pointer to root node
  public Node root;

  // constructor here
  // create new node
  public boolean insert(int data) {
    // search for the node which is immediately BEFORE where the new node should go
    // remember that with duplicate data values, we can just put the new node in
    // front of the chain of siblings
    Node neighbor = this.search(data);
    // if we've found the node our new node should follow, create it with the 
    // found node as parent, otherwise we are creating the root node
    // so if we're creating the root node, we simply create it and then set its
    // children, siblings, and parent to null
    // i think this is the part you're having trouble with, so look below for a 
    // diagram for visual aid
  }

  public Node search(int target) {
    // basically we just iterate through the chain of siblings and/or children 
    // until this.current.children is null or this.current.siblings is null
    // we need to make certain that we also search the children of 
    // this.index.sibling that may exist (depth-first recursive search)
  }
}

When we find the spot (using search()) where our new node should go, we need to reassign the parent, children, and siblings "links" inside the new node to its new parent, children, and siblings. For example, we take this:

A-|
|
B-C-|
| |
| F-G-|  
| |
| -
|
D-E-|
| |
- H-|
  |
  -

And we will insert a new node (X) where F is. This is just to illustrate how we reassign each of the new node's links. The finer details may be slightly different, but what's important here is the implementation example of the link reassignment:

A-|
|
B-C-|
| |
| X-F-G-|  
| | |
| - -
|
D-E-|
| |
- H-|
  |
  -

What we do is: 1) Create X. 2) Assign x.parent to c. 3) Reassign c.children to x. 4) Assign x.siblings to f. This inserts a new node (be mindful that an insertion is different than a sort and your tree may need resorting if you explicitly require a specific ordering).