Trouble implementing a templated singly linked list - c++

I'm trying to implement a templated singly linked list and I'm fairly new to C++
#include <iostream>
#include <string>
#define NEWL "\n"
#define PRINT(s) std::cout << s
#define PRINTL(s) std::cout << s << NEWL
#define PRINTERR(e) std::cerr << e << NEWL
////// Class for a Node
template<class Data> class Node {
Node<Data>* next_ptr;
Data data;
public:
Node(Node<Data>* nxt_ptr) :next_ptr(nxt_ptr) {};
Node(Data d, Node<Data>* nxt_ptr) :data(d), next_ptr(nxt_ptr) {};
Node<Data>* get_next() { return next_ptr; }
Data& get_data() { return data; }
friend std::ostream& operator<<(std::ostream& out, const Node<Data>& node) {
out << node.data;
return out;
};
};
////// Class for a SinglyLinkedList
template<class Data> class SLinkedList {
Node<Data>* head_ptr;
int max_size;
public:
SLinkedList() : head_ptr(nullptr) {};
bool is_empty() {
return head_ptr == nullptr;
};
bool is_full() {
return get_size() == max_size;
};
int get_size() {
if (is_empty()) {
return 0;
}
int count = 0;
for (Node<Data>* it_ptr = head_ptr; it_ptr != nullptr; it_ptr = it_ptr->get_next()) {
count++;
}
return count;
};
void add(Data d) {
if (is_full()) {
throw std::exception("List is full!");
}
Node<Data> new_node(d, head_ptr);
head_ptr = &new_node;
};
void print_content() {
int count = 1;
PRINTL("This list contains:");
for (Node<Data>* it_ptr = head_ptr; it_ptr != nullptr; it_ptr = it_ptr->get_next()) {
PRINTL("\t["<< count << "]" << " at " << it_ptr << " : " << *it_ptr);
count++;
}
}
};
////// Main function
int main()
{
SLinkedList<int> sll;
sll.add(42);
sll.print_content();
}
I can't get this to work. Somehow iterating the list with for-loops does not work. It always results in an Reading Access Violation Exception about a pointer to 0xCCCCCCD0 and I have no idea how to fix this.

Your add function is incorrect
Node<Data> new_node(d, head_ptr);
creates a new function local Node in add. You then set head to the address of that local variable. When the function ends all local variables are destroyed so now head points to an object that no longer exists.
To fix that you need to use the new keyword to create a dynamic object that will live on after the function ends.
Node<Data>* new_node = new Node(d, head_ptr);
head_ptr = new_node;
The down side with this is you need to remember to call delete on all of the nodes you created in the list destructor.
You also have some other bugs in your code. You never set max_size in your constructor so using it at all except to give it a value is undefined behavior as we have no idea what the value of it is going to be. You also never increase the size of the list when you add nodes into the list.

Related

Can't pass private member directly from main to function in C++

This my C++ code:
#include <iostream>
class Node
{
public:
int data;
Node* prev;
Node* next;
};
class Doublyll
{
private:
Node* head;
Node* tail;
public:
Doublyll();
Doublyll(int A[], int num);
~Doublyll();
friend std::ostream& operator<<(std::ostream& os, const Doublyll& src);
int Length(Node* p);
};
// Default Constructor will SET head and tail to NULL
Doublyll::Doublyll()
: head(NULL), tail(NULL)
{
}
// Explicit Construcor
Doublyll::Doublyll(int A[], int num)
: head(NULL), tail(NULL)
{
// std::cout << "Explicit Constructor called!\n";
Node** p = &head;
for (int i = 0; i < num; i++)
{
Node* t = new Node;
t->data = A[i];
if (head == NULL)
t->prev = NULL;
else
t->prev = tail;
t->next = NULL;
*p = t;
p = &(t->next);
tail = t;
}
}
// Destructor
Doublyll::~Doublyll()
{
// std::cout << "Desctructor called!\n";
Node* p = head;
Node* tmp;
while (p != NULL)
{
tmp = p;
p = p->next;
delete tmp;
}
}
// Display using Overloading << Operator
std::ostream& operator<<(std::ostream& os, const Doublyll& src)
{
Node* tmp;
for (tmp = src.head; tmp != NULL; tmp = tmp->next)
std::cout << tmp->data << " ";
std::cout << std::endl;
return os;
}
// Find Length how much Node in linked list
int Doublyll::Length(Node* p)
{
static int count = 0;
if (p != NULL)
{
count++;
Length(p = p->next);
}
return count;
}
int main()
{
int A[] = {2, 4, 6, 8, 10, 12, 14};
int size = sizeof(A) / sizeof(A[0]);
// Create object and linked list
Doublyll l1(A, size);
// Display linked list
std::cout << l1;
// Get length of linked list
int c = l1.Length(l1.head);
std::cout << c << std::endl;
return 0;
}
As you can see, I try to practice Doubly Linked List. Then, I want to count total Node in my linked list.
You can see in int Doublyll::Length(Node* p) I try to Count it using Recursion. Just because I want to practice it with Recursion. But, in my main() somewhow this code: int c = l1.Length(l1.head); said "Head is inaccessible"
I know that is because Head is Private in my Doublyll class. And I can simply change it to Public. OR I can write a function getHead() which will return Head pointer and then pass it as arguments.
So, Is there a way to dircetly pass it from my main() without change the member to public or write a getHead() function? or maybe there's another way to write a Recursion based on my problem, which in the future can also implement it to another recursion like display()? Because it seems like difficult to access if everything is inside class.
Maybe you can also review how I create a Doubly Linked List. Thank you!
Make int Doublyll::Length(Node *p) a private member function and add a public int Doublyll::Length() that takes no arguments and does:
int Doublyll::Length()
{
return Length(head);
}
(also you should probably make both of them const - int Doublyll::Length() const since they shouldn't modify anything)
Then just call l1.Length() in main.
Users of Doublyll shouldn't know about the internals of the class, and it doesn't make sense to ask a Doublyll object for the length from some node that it might not even own. Making Length(Node *p) private prevents nonsense things like l1.Length(l2.head).
As for your implementation of int Doublyll::Length(node *p) it's just wrong. As a comment mentions, you're using a static int count to track the length which will give you the wrong answer if you call the function multiple times. Plus your recursion is wrong since you aren't using the result of the recursive call. Do something like this instead:
int Doublyll::Length(Node *p) const
{
// Base case - no nodes
if (p == nullptr)
return 0;
// Recursive case
return 1 + Length(p->next);
}
Or a solution that allows for tail call optimization:
int Doublyll::Length(Node *p, int count) const
{
// Base case - no more nodes - return count
if (p == nullptr)
return count;
// Recursive case - increment count and go to the next node
return Length(p->next, count+1);
}
int Doublyll::Length() const
{
return Length(head, 0);
}
A common technique when implementing recursive functions is to have one public non-recursive function to start things off (say, Doublyll::Length()), and a second private helper function that actually performs the recursion (something like Doublyll::LengthRecursive()).
This could look something like:
int Doublyll::LengthRecursive(Node* p)
{
static int count = 0;
if (p != NULL)
{
count++;
Length(p = p->next);
}
return count;
}
int Doublyll::Length()
{
return LengthRecursive(head);
}
One way of handling a situation like this is to use two member functions: one that is the public interface and one that is private but has the signature you want for the recursive call.
For example:
class Doublyll
{
private:
Node* head;
Node* tail;
int LengthAux(Node* p); //private recursive implementation
public:
Doublyll();
Doublyll(int A[], int num);
~Doublyll();
friend std::ostream& operator<<(std::ostream& os, const Doublyll& src);
int Length(); // public interface
};
int Doublyll::Length() {
return LengthAux(head);
}
// Find Length how much Node in linked list
int Doublyll::LengthAux(Node* p)
{
static int count = 0;
if (p != NULL)
{
count++;
LengthAux(p->next);
}
return count;
}
...
This is a pretty common pattern used by implementations involving recursion. It is the nature of recursive calls that the signature of the recursive guts of the function is often different than the natural signature of calling the function externally.

c++ generic singly linked list string print fails

I am implementing a generic singly linked list. It works with char and int, but not for string. What mistake have I made? I get an error:
C2679 binary '<<': no operator found which takes a right-hand operand of type 'T' (or there is no acceptable conversion)
#include <iostream>
#include <vector>
#include <cstdlib>
#include <ctime>
using namespace std;
template <typename T>
struct Node
{
T data;
Node*next;
//friend class List<T>;
};
template<typename T>
class List
{
public:
List() :headNode{ nullptr } {} // empty list constructor
~List() // destructor
{
Node<T>*current = headNode;
while (current)
{
Node<T>*temp = current;
current = current->next;
delete temp; cout << "\nDeleting Nodes!!!";
}
}
bool empty() const // is list empty?
{
return headNode == nullptr;
}
const T& front() const // get front element
{
return headNode->data;
}
void addFront(const T& e) // add to front of list
{
Node<T>* tempNode= new Node<T>;
tempNode->data = e;
tempNode->next = headNode;
headNode = tempNode;
}
void removeFront()
{
if (empty()) cout << "List is empty" << endl;
else
{
Node<T> *tempNode = headNode;
headNode = tempNode->next;
delete tempNode; cout << "\nDeleting Node!!!";
}
}
void printList()const
{
Node<T>*current = headNode;
while (current)
{
cout << current->data<<" |-> ";
current = current->next;
}
}
private:
Node<T> *headNode;
};
int main()
{
srand(time(NULL));
List<int> mylist;
for (int i = 0; i < 15; ++i)
{
mylist.addFront(rand()%200 + 20);
}
mylist.printList();
mylist.removeFront(); cout << endl;
mylist.printList();
List<string>family;
family.addFront("aaa");
family.addFront("bbb");
family.addFront("ccc");
family.addFront("ddd");
family.printList();
List<char> charlist;
charlist.addFront('a');
charlist.addFront('b');
charlist.addFront('c');
charlist.addFront('d');
charlist.printList();
return 0;
}
You need to include the header <string>.
#include <string>
where the operator << for streams and the class std::string is declared.

Creating objects with new in function and "this is nullptr" exception

I've been trying to make a template class (called List) which stores different type of objects. I created Base class to be like base in my program and Human class. Base can create new Human and to have access to them all, has a (private) pointer to List * first_h (in every List is stored Human* me, List * next and List * first_h (first_h in list)).
The problem is, when I add like more than 1 Human to my Base, I can't display them properly. I think it's because of creating new Human's in Base method (void Base::create_human(string name)) but everything I did don't work it out.
There are my classes:
class Human
{
private:
string name;
public:
Human(string name) { this->name = name; }
void display() { cout << "My name: " << name << endl; }
};
template <class T>
class List
{
private:
T* me;
List <T>* next;
List <T>* first;
public:
void set_me(T* me) { this->me = me; }
T* get_me() { return this->me; }
void set_next(List* next) { this->next = next; }
List <T>* get_next() { return this->next; }
void set_first(List* first) { this->first = first; }
List <T>* get_first() { return this->first; }
void add(T*& created);
void display();
};
class Base
{
private:
List <Human>* first_h;
public:
void set_first_h(List <Human>*& first) { this->first_h = first; }
List <Human>* get_first_h() { return this->first_h; }
void create_human(string name)
{
Human* created = new Human(name);
this->first_h->add(created);
}
};
and methods:
template <class T>
void List<T>::add(T*& created)
{
List <T>* temp = this->get_first();
List <T>* new_list;
if ((this->get_me()) == nullptr)
{
this->set_next(nullptr);
this->set_me(created);
this->set_first(this);
}
else
{
new_list = new List <T>;
temp = this->get_first();
while (temp != nullptr)
{
temp = temp->get_next();
}
new_list->set_next(nullptr);
new_list->set_first(this->get_first());
temp->set_next(new_list);
}
}
template <class T>
void List<T>::display()
{
List <T>* temp_list = this;
T* temp;
if (temp_list == nullptr)
{
std::cout << "There is nothing!" << endl;
}
while (temp_list != nullptr)
{
temp = temp_list->get_me();
temp->display();
temp_list = temp_list->get_next();
}
std::cout << "End!" << endl;
}
and my main function:
int main()
{
Base Main;
List <Human>* first_h = new List <Human>();
Main.set_first_h(first_h);
Main.create_human("Jane");
Main.create_human("John");
Main.create_human("Mary");
Main.get_first_h()->display();
system("pause");
return 0;
}
Sorry for my English and thank you in advance!
Edit:
I found out what was wrong:
in add function:
new_list->set_next(nullptr);
new_list->set_me(created);
new_list->set_first(this->get_first());
temp->set_next(new_list);
I forgot about:
new_list->set_me(created);
the mistake in add function as one of you wrote.
Your loop
while (temp != nullptr)
{
temp = temp->get_next();
}
runs till temp is nullptr and then you do
temp->set_next(new_list);
So, as you see, inside set_next() the this pointer is nullptr.
Please learn how to use a debugger and look at the call stack.

Postincrementation operator in linked list

I had to write a program that handle this main code:(not allowed to change it)
list<int> iv;
iv["john"] = 23;
int ia = iv["john"]++;
int ib = iv["john"];
cout << ia << " " << ib << endl; // prints 23 24
try{
cout << iv["jack"] << endl; // should throw an exception
}catch(list<int>::Uninitialized&)
{
cout << "Uninitialized map element!" << endl;
};
Here is my code:
#ifndef EXAM_H
#define EXAM_H
#include <iostream>
#include <string>
using namespace std;
template <class TYPE>
class list
{
private:
struct node
{
TYPE value;
string index;
bool isInit;
node *next;
};
node *head;
node *current;
public:
class Cref
{
friend class list;
list& s;
string position;
Cref (list& ss, string pos): s(ss), position(pos) {};
public:
operator TYPE() const
{
return s.read(position);
}
Cref& operator = (TYPE val)
{
s.write(position,val);
return *this;
};
Cref& operator = (const Cref& ref)
{
return operator= ((TYPE)ref);
};
};
class Uninitialized{};
list ()
{
cout << "constructor\n";
head = NULL;
current = NULL;
}
~list ()
{
while (head)
{
node *t = head->next;
delete head;
head = t;
};
}
TYPE read (string ind) const
{
cout << "read\n";
node *t = head;
while(t)
{
if(t->index == ind && t->isInit == true) return t->value;
else t = t->next;
}
throw Uninitialized();
}
void write (string ind, TYPE value_)
{
cout << "write\n";
node *t = new node;
t->next = head;
head = t;
head->value = value_;
head->index = ind;
head->isInit = true;
}
TYPE operator[] (string ind) const
{
cout << "read\n";
node *t = head;
while(t)
{
if(t->index == ind && t->isInit == true) return t->value;
else t = t->next;
}
throw Uninitialized();
}
Cref operator[] (string ind)
{
return Cref(*this, ind);
}
};
#endif
Everything works great, but only when I comment out postincrementation operation in main program
int ia = iv["john"]++;
As you can see I have a struct node where I put all variables and I want to increment value by one in node where the key is "john". Is there any way to implement operator++ for this code ?
I am not allowed to use std::map.
The usual approach to your problem is defining the array subscript operators as
const TYPE& operator[](string ind) const;
TYPE& operator[](string ind);
In this way, you do not have to bother a single bit about the operator++: Since iv["John"] returns a reference to int, iv["John"]++ will call the int post-increment operator which is built-in.
Yes, I have already tried this solution, but compiler do not distinguish between reading and writing and still using non-const version. So I had to build proxy class Cref that helps to distinguish.
I have also already find a solution to operator++ problem.
This operation had to be from Cref level. I created
Cref& operator++ (int val)
{
s.increment(position,val);
return *this;
};
And increment function in main class body as follows:
void increment (string ind, int value_)
{
cout << "increment\n";
node *t = head;
while(t)
{
if(t->index == ind && t->isInit == true) t->value = t->value + 1;
t = t->next;
}
}
That fully solved my problem.

Linked list problem with constructor and destructor

Hi
I have some issue regarding constructor and destructor. I have list class, which has two inner classes, one private class for the list nodes, and one public iterator class.
Now for the issue, I have written a non-member print function which uses the inner iterator class. When i use this non-member function it will end calling the destructor for the iterator. It doesn't end here though because for some reason it will also call for the list class's destructor. Which causes some problem when I want to print the list content again.
I don't understand why it call the list class destructor as well and wonder if someone kindly can tell me that, and how I should fix it.
I have attached all the code related to the problem
Main
#include <iostream>
#include "sorted_list.h"
#include "iterator.h"
using namespace std;
void list_print(ostream& os, sorted_list list)
{
sorted_list::iteratorn it(&list);
while( ! it.iterator_end())
{
os << "key = " << setw(3) << it.iterator_get_key() << ", "
<< "value = " << setw(5) << it.iterator_get_value() << endl;
it.iterator_next();
}
os << endl;
}
int main()
{
sorted_list a;
a.insert(4,4);
a.insert(5,5);
list_print(cout,a);
list_print(cout,a);
}
sorted_list.cc
#include "sorted_list.h"
sorted_list::sorted_list()
{
cout << "construct sorted_list" << endl;
this->first = 0;
}
sorted_list::~sorted_list()
{
cout << "destruct sorted_list" << endl;
destroy(this->first);
}
void sorted_list::destroy(list_link* item)
{
cout << "destroy list_link" << endl;
if(item)
{
destroy(item->next);
delete item;
}
}
void sorted_list::insert(int key, double value)
{
list_link *curr;
list_link *prev = 0;
curr = first;
while(curr)
{
if(value < curr->value)
break;
prev = curr;
curr = curr->next;
}
if(this->first == 0 || prev == 0) //if empty or add first
{
//first = create(key, value, this->first);
first = new list_link(key, value, this->first);
}
else if(curr == 0)
{
//prev->next = create(key, value, 0);
prev->next = new list_link(key, value, 0);
}
else
{
//prev->next = create(key, value, curr);
prev->next = new list_link(key, value, curr);
}
}
void sorted_list::remove(my_key_type key)
{
list_link *curr = first;;
list_link *prev = 0;
while(curr)
{
if(curr->key == key)
{
list_link *remove;
if(prev == 0)
{
first = curr->next;
delete curr;
curr = first;
}
else
{
remove = curr;
curr = curr->next;
prev->next = curr;
delete remove;
}
continue;
}
prev = curr;
curr = curr->next;
}
}
sorted_list::list_link* sorted_list::clone(list_link* item)
{
list_link* copyItem= new list_link(item->key,item->value,0);
if(item->next!= 0)
copyItem->next=clone(item->next);
return copyItem;
// ADD YOUR CODE HERE ( 4 well formatted lines in reference solution )
}
void sorted_list::copy(sorted_list* my_this_destination)
{
if (my_this_destination->first == 0) // copy if empty
{
cout << "Copy" << endl;
//list_destroy(my_this_destination);
my_this_destination->first = clone(first);
}
}
double sorted_list::find(int key)
{
list_link *travel = this->first;
while(travel)
{
cout << travel->key << "==" << key << endl;
if(travel->key == key)
return travel->key;
travel = travel->next;
}
return -1;
}
int sorted_list::size()
{
list_link *travel = this->first;
int i = 0;
while( travel )
{
travel = travel->next;
i++;
}
return i;
}
sorted_list.h
#ifndef _SORTED_LIST_H_
#define _SORTED_LIST_H_
#include <iostream>
#include <iomanip>
using namespace std;
typedef int my_key_type;
typedef double my_value_type;
class sorted_list
{
public:
sorted_list();
~sorted_list();
void insert(int key, double value);
void remove(my_key_type key);
void copy(sorted_list* my_this_destination);
void destroy();
void init(struct my_list* my_this);
void print();
void print2();
double find(int key);
int size();
private:
class list_link // An inner class inside sorted_list
{
public:
list_link (my_key_type key, my_value_type value, list_link* next = 0);
~list_link();
my_key_type key;
my_value_type value;
list_link *next;
};
list_link* first;
list_link* clone(list_link* item);
void destroy(list_link* item);
// More declarations
public:
class iteratorn
{
public:
iteratorn();
~iteratorn();
iteratorn(sorted_list *item);
list_link* list_begin();
bool iterator_end();
void iterator_next();
int iterator_get_key();
double iterator_get_value();
private:
sorted_list::list_link* current;
};
};
#endif
iteratorn.cc
#include "iterator.h"
#include "sorted_list.h"
sorted_list::iteratorn::iteratorn()
{
}
sorted_list::iteratorn::iteratorn(sorted_list *list)
{
cout << "construct iteratorn" << endl;
this->current = list->first;
}
sorted_list::iteratorn::~iteratorn()
{
cout << "destruct iteratorn" << endl;
}
sorted_list::list_link* sorted_list::iteratorn::list_begin()
{
return current;
}
void sorted_list::iteratorn::iterator_next()
{
current = current->next;
}
int sorted_list::iteratorn::iterator_get_key()
{
return current->key;
}
double sorted_list::iteratorn::iterator_get_value()
{
return current->value;
}
list_link.cc
#include "sorted_list.h"
sorted_list::list_link::list_link(my_key_type key, my_value_type value, list_link* next)
{
this->key = key;
this->value = value;
this->next = next;
}
sorted_list::list_link::~list_link()
{
cout << "list_link destructor" << endl;
}
Your function void list_print(ostream& os, sorted_list list) takes a sorted_list parameter by copy. A quick and dirty fix (that you should do anyways for performance reasons) is the following:
void list_print(ostream& os, const sorted_list& list)
Now, your iteratornclass takes a mutable list, so this won't work as you expect. You will have quite a few methods to change to make this work.
In any case, your real problem is the lack of a proper copy-constructor. Right now, when you "copy" a list, both end up sharing the same elements, but your destructor is written as if each list owns it's own nodes. Define a proper copy operation and it will solve your problem.
More elaborate help on how to solve the problem: (untested)
Change signature:
void list_print(ostream& os, const sorted_list& list);
Declare + define copy constructor:
sorted_list::sorted_list (const sorted_list& other);
Change iteratorn interface to support a const sorted_list:
class sorted_list::iteratorn
{
public:
iteratorn();
~iteratorn();
iteratorn(const sorted_list& list);
const list_link* list_begin() const;
bool iterator_end() const;
void iterator_next();
int iterator_get_key() const;
double iterator_get_value() const;
private:
// You *should* make this `const` but it is not required.
sorted_list::list_link* current;
};
As you can see, the changes are rather minimal, but need to be applied in various places.
const + non-const iterators:
I applied changes here based on the fact that your iteratorn was currently only defining read-only operations on your sorted_list. If you want to support write access to allow changing the value stored in list nodes (never allow changing the key or you won't have a sorted list anymore), you should define two iterator classes. See the STL iterator interface for more details.
You're copying the list by value, so the local copy in list_print() destructs at end of scope. Pass it by const-reference instead.
This in turn means you will have to change your sorted_list to support working with const lists. In particular you need to have a function that returns a const iterator pointing to the beginning of the list:
sorted_list::const_iteratorn begin() const
{
// returns a const_iteratorn pointing at the beginning of this list
}
Notice you need a new kind of iterator: a const_iteratorn, which promises it won't change the list.
Then, inside print_list() initialize a const_iteratorn with the start iterator that sorted_list returns, by copy:
sorted_list::const_iteratorn s(list.begin());
Finally create a second iterator instance that initializes with an end iterator coming from a member function of sorted_list, similar to the begin() function. This will maintain the const-correctness in print_list().
sorted_list::const_iteratorn e(list.end());
while( s != e ) { // two iterators should be able to compare
// ...
s.iterator_next(); // consider ++s
}
Also, as André mentioned, the fact you don't have a proper copy-constructor and assignment operator is a severe issue. Make sure that copying a sorted_list means copying all its elements, so that the new object owns its own list of elements. Do recall the Rule of Three.