[GE] XOR Linked List – A Memory Efficient Doubly Linked List | Set 2


XOR Linked List – A Memory Efficient Doubly Linked List 

In the previous post, we discussed how a Doubly Linked can be created using only one space for address field with every node. In this post, we will discuss the implementation of memory efficient doubly linked list. We will mainly discuss the following two simple functions.
1) A function to insert a new node at the beginning.
2) A function to traverse the list in forward direction.

In the following code, insert() function inserts a new node at the beginning. We need to change the head pointer of Linked List, that is why a double pointer is used (See this). Let us first discuss few things again that have been discussed in the previous post. We store XOR of next and previous nodes with every node and we call it npx, which is the only address member we have with every node. When we insert a new node at the beginning, npx of new node will always be XOR of NULL and current head. And npx of the current head must be changed to XOR of new node and node next to the current head.


printList() traverses the list in forward direction. It prints data values from every node. To traverse the list, we need to get pointer to the next node at every point. We can get the address of next node by keeping track of current node and previous node. If we do XOR of curr->npx and prev, we get the address of next node.


/* C++ Implementation of Memory 
efficient Doubly Linked List */
#include <bits/stdc++.h>
#include <inttypes.h> 
using namespace std;
  
// Node structure of a memory 
// efficient doubly linked list 
class Node 
    public:
    int data; 
    Node* npx; /* XOR of next and previous node */
}; 
  
/* returns XORed value of the node addresses */
Node* XOR (Node *a, Node *b) 
    return (Node*) ((uintptr_t) (a) ^ (uintptr_t) (b)); 
  
/* Insert a node at the beginning of the 
XORed linked list and makes the newly 
inserted node as head */
void insert(Node **head_ref, int data) 
    // Allocate memory for new node 
    Node *new_node = new Node(); 
    new_node->data = data; 
  
    /* Since new node is being inserted at the 
    beginning, npx of new node will always be 
    XOR of current head and NULL */
    new_node->npx = XOR(*head_ref, NULL); 
  
    /* If linked list is not empty, then npx of 
    current head node will be XOR of new node 
    and node next to current head */
    if (*head_ref != NULL) 
    
        // *(head_ref)->npx is XOR of NULL and next. 
        // So if we do XOR of it with NULL, we get next 
        Node* next = XOR((*head_ref)->npx, NULL); 
        (*head_ref)->npx = XOR(new_node, next); 
    
  
    // Change head 
    *head_ref = new_node; 
  
// prints contents of doubly linked 
// list in forward direction 
void printList (Node *head) 
    Node *curr = head; 
    Node *prev = NULL; 
    Node *next; 
  
    cout << "Following are the nodes of Linked List: \n"
  
    while (curr != NULL) 
    
        // print current node 
        cout<<curr->data<<" "
  
        // get address of next node: curr->npx is 
        // next^prev, so curr->npx^prev will be 
        // next^prev^prev which is next 
        next = XOR (prev, curr->npx); 
  
        // update prev and curr for next iteration 
        prev = curr; 
        curr = next; 
    
  
// Driver code 
int main () 
    /* Create following Doubly Linked List 
    head-->40<-->30<-->20<-->10 */
    Node *head = NULL; 
    insert(&head, 10); 
    insert(&head, 20); 
    insert(&head, 30); 
    insert(&head, 40); 
  
    // print the created list 
    printList (head); 
  
    return (0); 
  
// This code is contributed by rathbhupendra

Output:
Following are the nodes of Linked List:
40 30 20 10














































































































Comments

Popular posts from this blog

Important Program Collection

[16 Feb 2020] Given an array where every element occurs three times, except one element which occurs only once.

Which data structure is used in redo-undo feature?