Clarity between Pointer-to-Pointer, Value-of-Pointed address and 2D-Array access

Question

Can someone set the path straight on the disparities between these ubiquitous, as It can get a bit mind blowing.

For example if I have a 2D array such as rec[3][2] , the following access means the same;

rec[0][0] = **rec
rec[i][0] = **(rec + i)
*(*(rec + i) + j) = rec[i][j]

If this is the case then what are the meaning of these:

#include <stdio.h>
double *recptr[3];
int i=1;

main()
{
 double n1=12.0;
 doublw n2=3.4;

 recptr[0]= &n1;
 recptr[1]= &n2;

 printf("Amt: %.2f\n", **(recptr + i));
}

What is **(recptr + i) , Is this access to 2D pointer or ponter-to-pointer reference?

foo(ptr2ptr)double **ptr2ptr;
{
 int i=1, j=0;
 if(**(ptr2ptr +i) > **(ptr2ptr + j))
 {
  double *tmp= *(recptr +i);
 }
}

Again what is the difference between *(recptr +i) and **(ptr2ptr +i) ?! is the later also 2D access or access to pointer-2-ponter reference and the earlier the object pointed to?

Answer 1

double *recptr[3];

This creates an array of 3 pointers to double (double*) .
Now, in the main() , we have double n1 = 12.0f and double n2 = 3.4f .

Now we assign the address of n1 (&n1) to recptr[0] (First element of the array which just happens to point to doubles, so now it points to an address of a double (our n1) which is ok.) -> Please do note this is NOT a reference (in theory). We are taking the address of n1.

We do the same thing with n2, so now recptr[1] is a pointer that points to the address of n2.

Now we print **(recptr + i) -> This means the following:

We take the address of the first element of the array (recptr evaluates to the address of the aray), we offset this address by i (which just happens to be 1). Now please do note that we don't move 1 byte further, we move sizeof(double*) bytes further.

Now, we dereference this position *(recptr + i) which is (address of first element + sizeof(double*) bytes) away and what do we see ? Well, it's a pointer to a double (Or in other words, it's just an address of n2). So we need to dereference yet again to actually see the value that this particular pointer points to, hence **(recptr + i) .

So, the first dereference just gives us the correct "cell" of our recptr array but in that cell, there's a pointer to a double, so in order to get to it's value, we need to dereference yet again.

(We then indeed print 3.40 since it's address happens to live in the second cell of recptr array.)