casting void*

Hello,

How could I cast a void* into an integer pointer that points to integers of 1 byte size?

I was thinking about something like this,

Code:

---

void* = vp;
short int* ip = reinterpret_cast<short int*>(vp);

---

but that will give me a pointer to a 2 byte integer.

Seron

Cast it to a char*.

Thanks, that works.

Is there a way to make custom casts, like for a word of 15 bytes or for one of half a byte?

Another problem I have are the following casts. The first cast works. The 2nd gives the error "Cannot cast from 'int' to 'unsigned char'".

Code:

---

int value

unsigned char uchvalue = (unsigned char)value;

unsigned char uchvalue = reinterpret_cast<unsigned char>(value);

---

Quote:

---

Is there a way to make custom casts, like for a word of 15 bytes or for one of half a byte?

---

Assuming that char is 8 bits:

Code:

---

#include <iostream>
#include <iomanip>

using namespace std;

typedef unsigned char byte;
inline byte get_first(byte b)  { return b >> 4; }
inline byte get_second(byte b) { return b & 0xf; }
inline void set_first(byte& b, byte d)  { b |= d << 4; }
inline void set_second(byte& b, byte d) { b |= d & 0x0f; }

int main() {
    // Declare two half-bytes
    byte b = 0;
    set_first(b, 0xa);
    set_second(b, 0x5);

    cout << "1st half-byte = 0x" << hex << static_cast<int>(get_first(b)) << '\n';
    cout << "2nd half-byte = 0x" << hex << static_cast<int>(get_second(b)) << '\n';
}

---

Or something of the kind. Of course you can wrap this up in a class and together with operator overloading provide a cleaner interface. The 15 byte type will be more difficult and quite expensive in computing power. Are you sure you need such big numbers?

Quote:

---

Another problem I have are the following casts.The first cast works. The 2nd gives the error "Cannot cast from 'int' to 'unsigned char'".

Code:

---

int value

unsigned char uchvalue = (unsigned char)value;

unsigned char uchvalue = reinterpret_cast<unsigned char>(value);

---

---

Use static_cast for such casts. You can also let it convert it implicitly, though most compilers will give a warning about possible loss of data.

Thanks Imov, just what I needed.

There is no need for a 15 byte type. I only wish to understand casting better.

Seron

I made a little mistake. The set_* functions should be:

Code:

---

inline byte set_first(byte& b, byte d)  { b &= 0x0f; b |= d << 4; }
inline byte set_second(byte& b, byte d) { b &= 0xf0; b |= d & 0x0f; }

---