开发者

In C++, how to I stop my pointers from being 'overridden' when a function is called again?

开发者 https://www.devze.com 2023-01-19 10:07 出处:网络
all! I don\'t know if \'overridden\' is the right word here. In my programming class, I have to create a circular list, such that each node

all! I don't know if 'overridden' is the right word here. In my programming class, I have to create a circular list, such that each node node contains a pointer pointing to the next node and the final node points to the first node. In addition, there is a tail node that points to the last node added (its null before any nodes are added).

I cannot populate my list (called a ring) because every time I call the Ring::Insert(const int& d) function, which inserts a single node, and it gets to the line "RingNode newNode(d);", the new RingNode object overwrites the previous RingNode object that was created when I last called the Ring::Insert(const int& d) function. Obviously, I don't want this because it messes up my list. How do I make it so every time the function creates a brand new RingNode object it doesn't interfere with the previous RingNode objects?

Source code from my header file, just in case:

class RingNode {
public:
    RingNode(const int& i=0 ): data(i), next(NULL){}
private:
    int data;  /* ID of player */
    RingNode* next;
friend class Ring;

And here is the function in question

RingNode* Ring::Insert(const int& d){
    RingNode newNode(d); //This line overwrites previous RingNode objects
    RingNode* refNode = &newNode; //Probably bad form, but that's not my main concern right now
    if (tail==null){
            tail = refNode;
            newNode.next = refNode;
            return refNode;
    }
    newNode.next = (*GetTail()).next;
    (*GetTail()).next = refNode;
    tail = refNode;
    开发者_JAVA百科return refNode;
}

So, for example, if I execute the following snippet in my main...

Ring theRing;
theRing.Insert(5);
theRing.Insert(2);
theRing.Insert(7);

If I debug my project I can see that theRing contains only one RingNode, first it's the 5 RingNode, then the 2 RingNode overwrites it, then the 7 RingNode overwrites that. Thanks for reading and double thanks for your replies!

EDIT: I replaced

RingNode newNode(d); 
RingNode* refNode = &newNode;

with

RingNode *newNode = new RingNode(d);

tweaked the rest of the code, and it's working properly. Thanks so much for the help, guys! Very informative and best of all I now understand why it was messing up.


You're re-using the same local variable on the stack every time.

RingNode newNode(d); //This line overwrites previous RingNode objects

is a local variable - it lives on the stack. So it's only valid during the lifetime of your insert method. However, since you're calling insert several times in a row from the same calling function, your different "newNode"'s wind up at the same place on the stack.

What you probably want to do is

RingNode *refNode = new RingNode(d);

This will dynamically allocate your RingNode on the heap.

However, now you have to worry about using delete to clean up all the nodes when your Ring is destroyed.


RingNode newNode(d);
RingNode* refNode = &newNode;

Replace with:

RingNode* refNode = new RingNode(d);

BTW, replace

RingNode(const int& i=0 )

with:

RingNode(int i=0 )

There is no need to use reference for small type like integer.


You need to create objects that last beyond the scope of your function... so you need to use the new operator.

RingNode* Ring::Insert(const int& d){
    RingNode* refNode = new RingNode(d); // this line creates a ring node not bound to the scope of the function.
    if (tail==null){
            tail = refNode;
            newNode.next = refNode;
            return refNode;
    }
    newNode.next = (*GetTail()).next;
    (*GetTail()).next = refNode;
    tail = refNode;
    return refNode;
}


You can't add an object allocated as an "auto" variable, i.e. on the stack, like that. You need to do RingNode *newNode = new RingNode(d); and add that.

0

精彩评论

暂无评论...
验证码 换一张
取 消