what does the following c++ code do

Question

I know its very naive question, but i am not able to understand what the following code does.

#include <malloc.h>
#define MAXROW 3
#define MAXCOL 4
int main(){
int (*p)[MAXCOL];
p = (int (*)[MAXCOL])malloc(MAXROW*sizeof(*p));
}

Please provide a complete explanation including the type and size of p.

It is just for learning purpose. I am not using this code in any real application.

Smells like homework? If yes, please tag accordingly. Also, we are no service site where you can just ask us to do your job. — Xeo, Jul 09 '11 at 18:13
Code smells like C, title says C++ (which makes this terrible code), and tags say both. Which is it? — GManNickG, Jul 09 '11 at 18:19
@Tomalak: right. I should have said ... @Nitin: trying to write multi-language source files is hard. I suggest you stick to one of C or C++ — pmg, Jul 09 '11 at 18:19
The code was wrriten by Dr. P.J. Plauger (google him), and he tagged it as C++ in his book. I was just trying to get hang of it. — Nitin Garg, Jul 09 '11 at 18:33
@Nitin: Nobody "owns" Stack Overflow (unless you're talking legally). — Lightness Races in Orbit, Jul 09 '11 at 18:41
@Nitin: No-one writes code like that in a real-world C++ application. C++ developers use `new`, `std::vector` or `std::array`... and if they _really_ need legacy C features, then they include non-deprecated `cX`, not deprecated `X.h`. I recommend [these resources](http://jcatki.no-ip.org/fncpp/Resources) if you're serious about learning C++. — Lightness Races in Orbit, Jul 09 '11 at 21:29

score 5 · Accepted Answer · answered Jul 09 '11 at 18:24

5

As far as I can tell, it's gibberish. You probably meant (int(*)[MAXCOL]).

In C it means that the programmer who wrote it doesn't know how void pointer typecasts work.

In C++ it means that you are allocating an array of arrays. p is an array pointer, so *p is an array of size MAXCOL, and you allocate MAXROW such arrays. The result is a "mangled" 2D array. The avantage of using this rather obscure syntax is that you get a 2D array which has every cell in adjacent memory, something you wouldn't achieve with the more commonly seen pointer-to-pointer dynamic 2D array.

answered Jul 09 '11 at 18:24

Lundin

195,001
40
254
396

How are pointer-to-pointer dynamic 2D arrays "more commonly seen"? They are an order of magnitude more complex, and you need loops to populate them. – Lightness Races in Orbit Jul 09 '11 at 18:45
It appears to be the most common way to teach dynamic 2D arrays, for some reason, I don't know why, as pointer-to-pointer arrays aren't compatible with static 2D arrays, making it a pain to write generic functions handling either kind. – Lundin Jul 09 '11 at 19:51
Well how very odd! They are _not_ dynamic 2D arrays. They are dynamic arrays of pointers (as you said). – Lightness Races in Orbit Jul 09 '11 at 21:28

pmg · Answer 2 · 2011-07-09T18:47:34.230

4

Supposing you meant the uncommented line (the other is the original, which is not valid C)

// p = (*)[MAXCOL]malloc(MAXROW*sizeof(*p));
p = (int(*)[MAXCOL])malloc(MAXROW*sizeof(*p));

my answer is:

In C do not cast the return value of malloc. It is at best redundant and may hide an error when present. Simply do

p = malloc(MAXROW * sizeof *p);

edited Jul 09 '11 at 18:47

answered Jul 09 '11 at 18:15

pmg

106,608
13
126
198

score 2 · Answer 3 · answered Jul 09 '11 at 18:21

2

It's not valid code in C or C++.

So, it doesn't "do" anything at all.

answered Jul 09 '11 at 18:21

Lightness Races in Orbit

378,754
76
643
1,055

2

@Nitin: Have you decided which language you're trying to write, yet? Will you tell us? – Lightness Races in Orbit Jul 09 '11 at 18:40

score 1 · Answer 4 · edited May 23 '17 at 12:05

1

As you can learn from this question, int (*p)[MAXCOL] is a pointer to an array of MAXCOL integers.

The line p = (int (*)[MAXCOL])malloc(MAXROW*sizeof(*p)); allocates the memory for an array of MAXROW arrays of MAXCOL integers (i.e. two dimensional array), and sets p to point to it.

edited May 23 '17 at 12:05

Community

1
1

answered Jul 09 '11 at 19:04

Igor

26,650
27
89
114

score 1 · Answer 5 · answered Jul 10 '11 at 09:03

I have not compiled the following code, but i think it's valid c++-code:

    typedef int[MAXROW][MAXCOL] table;

    table *p = new table;

This code is only posible if the dimensions of the array are known at compile-time.

This is the way most c++-prgrammers would define p:

    using namespace std;
    vector<vector<int> > p;

This allows for a more flexible way of programming.

G, folks, it's been a long time since i have programmed K&R C!

what does the following c++ code do

5 Answers5