如何在没有双指针的情况下更改 function 内的结构？

Question

I'm trying to implement a hash table that contains contacts, so every line of the hash table is a chained list. When the hash table has too many contacts, I have to resize it. However, I don't have a say in the functions' prototypes, meaning the function resize always takes a single pointer to my hash table as an argument.

These are my structures:

struct dir {
uint32_t len;
uint32_t contactsNumber;
struct contact* contactList;
};

struct contact {
char *name;
char *num;
struct contact *next;
};

This is how I initialize my directory:

struct dir *dir_create(uint32_t len)
{
struct dir * directory = malloc(sizeof(struct dir)*len);
directory->contactsNumber = 0;

for (uint32_t i = 0; i<len;i++) {
  directory[i].contactList = contact_sentinel();
}

directory->len = len;
return(directory);
}

contact_sentinel() returns a chained list with an empty dummy node.

This is my resize prototype (imposed):

extern void dir_resize(struct dir *dir,uint32_t new_size);

Obviously, when I try to manipulate my struct within the resize function, everything works well and the table is resized, however as soon as I leave the function, my directory goes back to normal. The reason is that I'm not using a double pointer, meaning my directory still points to the same old directory and nothing was changed.

So my question is: How can I resize effectively without using a double pointer? Should I change something in my structure? Are there ways to do it?

This is my resize function:

void dir_resize(struct dir *dir,uint32_t new_size)
{
uint32_t old_size = dir->len;
struct dir *new_dir = dir_create(new_size);

for (uint32_t i = 0; i<old_size;i++) {
      struct contact *current_contact = dir[i].contactList;
      while (current_contact->next != NULL) {
          dir_insert(new_dir, current_contact->next->name,current_contact->next->num);
          // contact_removeTop(current_contact->next);
          current_contact->next = current_contact->next->next;
      }
    }
  }

I don't know what to do after that, now that I've created the new hash table (by calculating the new hash and inserting it into the right spot), how do I transfer the new_dir to the old dir with this kind of prototype?

Answer 1

Thanks for your answer Reinstate Monica!

Unfortunately doing dir->contactList = new_dir didn't change the table as it reverted back to what it was outside of the function.

However, this is my updated resize function:

void dir_resize(struct dir *dir,uint32_t new_size)
{
uint32_t old_size = dir->len;
struct dir *new_dir = dir_create(new_size);

for (uint32_t i = 0; i<old_size;i++) {
      struct contact *current_contact = dir[i].contactList;
      while (current_contact->next != NULL) {
          dir_insert(new_dir, current_contact->next->name, current_contact->next->num);
          // contact_removeTop(current_contact->next);
          current_contact->next = current_contact->next->next;
      }
    }

  
dir_free(dir);
dir->contactsNumber = new_dir->contactsNumber;
dir->len = new_size;
for (uint32_t i = 0; i<new_size;i++) {
    dir[i].contactList = new_dir[i].contactList;
}
}

The lines I added at the end seem to work. And my table is changed outside of the function. Now I just gotta figure out how to free the memory space previously used by dir. (idk if free(dir) is enough since it has *contacts within it).