Where we left off

// Start the list
Node *head = new Node;
head->data = 1;

// Add another element
head->next = new Node;
head->next->data = 2;
head->next->next = NULL;

Today's topics

• Algorithms for working with linked lists
  • Inserting a node
  • Searching for a value
  • Deleting a node
• Passing linked lists to functions
• Linked list variations

Textbook Chapter 13

Traversing a list

In C++ syntax:

Node *current = head;
while (current) { // or while (current != NULL)
    // do something with current->data
    current = current->next;
}

• This is a sentinel loop that stops when current is NULL
  • As always, the LCV update must be the last line of the loop body
  • Very important that the last node point to NULL!
• current does not allocate new memory (other than for the pointer itself)

Passing linked lists to functions

• Many algorithms, like printing a list, would be most useful as a function

• What should be passed to the function? By value or by reference?

  • The whole list is attached to the head pointer
  • Does the head need to point somewhere else?

  // by value
  void print(Node *head);
  

  // by reference
  void insert(Node *&head, int value);
  

• If Node *& is confusing, you can use a typedef:

  typedef Node * NodePtr;
  void insert(NodePtr &head, int value);
  

Linked list + function example

Write a function to calculate and return the length of a linked list.

Inputs: Head pointer (by value or by reference?)

Outputs: Length of list (what datatype?)

Caution: passing by reference vs value

Say we defined an insertion function as:

void insert(Node *head, int value); // pass head by value

When testing with the value 5, this will work!

• What about 9?
• What about 12?
• What about 0?

Finding the proper position - v1

• We need to find the node that comes before the insertion point
• This means that we need to examine the next pointer of each node

  Node *current = head;
  while (current->next && current->next->data < value) {
      current = current->next;
  }
  

• Very important to check current->next before accessing current->next->data!
• What if the list is empty?
• What if the new node needs to be inserted at the head?

Finding the proper position - v2

• Alternatively, we could use two pointers traversing in parallel:

  Node *prev = NULL;
  Node *current = head;
  while (current && current->data < value) {
      prev = current;
      current = current->next;
  }
  

• Makes handling of special cases easier, but need to keep track of two pointers
• Which is better? Whichever makes more sense to you!

Special cases

For every linked list operation, think about the special cases!

Linked List check-in 1/2

What am I forgetting in the following code? Assume that a list of Nodes already exists with a pointer to head defined.

Node *temp = new Node;
temp->data = 0;
temp->next = head;

Linked List check-in 2/2

What is the following code doing? Again, assume head is defined.

Node *temp = new Node;
temp->data = 5;
temp->next = head->next;
head->next = temp;

Deleting a node

• We need to:
  • Find the node to delete
  • Copy the address of the next node
  • Disconnect the node by changing the next pointer of the previous node
  • Free the memory using delete
• More traversing!
• More special cases!

Deleting a node - code example

Node *prev = NULL;
Node *current = head;
// Find the node to delete
while (current && current->data != value) {
    prev = current;
    current = current->next;
}

prev->next = current->next; // Disconnect the node
delete current; // Free the memory

What special cases do we need to consider?

Special cases for deletion

Linked list variations

• By creating your own data structure, you get to define the rules!
• Maybe you're doing a lot of adding/deleting at the end of the list, so you might want to keep track of a pointer to the last node (the tail)
• You might want to keep track of the length of the list or the sort order
• Once you start getting fancy, you probably want to create a class to encapsulate all of this information - we'll do this next week
• First though, let's look at doubly linked lists

Doubly linked lists

• So far we've only been able to travel one direction - the next node has no knowledge of its predecessor
• Solution: add a prev pointer to each node

    struct Node {
        int data;
        Node *next;
        Node *prev;
    };
  

• Advantages: easier to delete nodes, easier to traverse backwards
• Disadvantages: more memory, more complexity (two pointers to maintain)

Circularly linked lists

• Instead of the last element pointing to NULL, it points to head
• There is no real head anymore, but you still need to keep a pointer to somewhere in the list
• Only really useful when there's no start or end
• Advantages: easier to traverse, no need to check for NULL
• Disadvantages: more complexity, usually can't be empty

Lists of lists

• You can have a linked list of linked lists!
• Each node in the "outer" list is the head of another list
• Example: list of courses, each course has a list of students

Cross-linked lists

Expanding on the previous example, consider the situation where:

• Each student has a list of courses
• Each class has a list of students
• Each instructor has a list of courses

struct Student {
    string name;
    int id;
    Course *courses;
};

struct Course {
    string name;
    int number;
    Student *students;
    Instructor *instructor;
};

struct Instructor {
    string name;
    Course *courses;
};

Object oriented programming preview

• So far we've been implementing solutions in a procedural style
• The object oriented approach is based on the idea that different objects can be interacted with in a different way
  • You can sit on a chair
  • You can draw with a pen
  • You can (probably) pick up a chair and a pen
  • Can you draw with a chair?
• In the OO approach, we can encapsulate data and functions in a class - an abstract data type that defines how an object can be interacted with

Classes

• A class is a blueprint for creating objects, much like how a struct is a blueprint for creating data structures

  struct Student {
      string name;
      int id;
  };
  

  class Student {
      string name;
      int number;
  
      void print();
  };
  

• A class is a type of object, just like int or string or Node

• Member functions are accessed using . or -> just like member variables

Coming up next

• Tutorial: Linked lists
• Assignment 3 Linked lists
• Next week: Classes and objects

Textbook Chapter 10.2-10.3