C / C++ / MFC

Re: pointers to functions

16-Jul-22 20:58

Not quite, I started in the transistor age, so things were getting quite sophisticated.

Greg Utas16-Jul-22 7:36

Greg Utas

16-Jul-22 7:36

You're creating a pointer to a free function (one defined outside a class) and then storing it, and invoking it, from within an object. But it's also possible to create a pointer to class member data or a member function. See here[^]. You may need to read several articles about this to gain a good understanding, because I can't point to one that is really good on its own. Search on "C++ pointer to member" and read articles that discuss the type Class::* (a pointer to a class member) and operators .* and ->*. These make the following possible:

C++

int Class::* pm = &Class::m;         // pointer to member data "int m"
int (Class::* pf)(int) = &Class::f;  // pointer to member function "int f(int)"
Class c, *k;
c.*pm = 1;        // c.m = 1;
n = c.*pf(0);     // n = c.f(0);
k->.pm = 2;       // k->m = 2;
n = (k->*pf)(0);  // n = k->f(0);  note that ->* requires parentheses

Robust Services Core | Software Techniques for Lemmings | Articles
The fox knows many things, but the hedgehog knows one big thing.

Convert decimal to binary in C

sahil Ranka14-Jul-22 1:14

14-Jul-22 1:14

I'm attempting to convert a decimal to binary such as 192 to 11000000. I simply need a basic code to do this yet the code I have so far doesn't work:

void dectobin(int value, char* output)
{
    int i;
    output[5] = '\0';
    for (i = 4; i >= 0; --i, value >>= 1)
    {
        output[i] = (value & 1) + '0';
    }
}

Prior to this, done thorough research on decimal to binary in cI and gone through some good Articles, to have a clear understanding of the topic.

Any help would be much appreciated!

modified 23-Sep-22 12:22pm.

Mircea Neacsu14-Jul-22 1:25

14-Jul-22 1:25

You don’t say exactly why it doesn’t work so I’ll just guess: the code works only with numbers up to 2^5=32. If you tried it with 192 it would have failed.

As a general advice, for small algorithms like that, it helps to “play the computer”: take a piece of paper and go through each step as you would be the computer. That gives you a better understanding of how the algorithm works and helps you find eventual bugs.

Mircea

Richard MacCutchan14-Jul-22 1:29

14-Jul-22 1:29

Your code is correct as far as the conversion goes. The problem is that it can not handle any binary value greater than 31. You need to calculate the number of digits in the converted value first, and create an array big enough to store the string.

sahil Ranka14-Jul-22 2:42

14-Jul-22 2:42

Thanks Richard MacCutchan I will Try this.

CPallini14-Jul-22 2:36

CPallini

14-Jul-22 2:36

As noted in other answers, your code handles at most four bits.
Moreover, your function is assuming a (at least) 5-bytes buffer is provided by the caller. That's a flawn: the caller should provide the size of the buffer and the called should check if the size is big enough to produce the requested output.
Try

#include <stdio.h>
#include <stddef.h>
#include <stdbool.h>

bool uint2binstr(unsigned value, char output[], size_t output_size)
{
  unsigned v = value;
  unsigned bits = 0;
  while ( v )
  {
    v >>= 1;
    ++bits;
  }
  if ( ! bits ) ++bits;

  if ( output_size > bits )
  {
    output[bits] = '\0';
    while (bits)
    {
      --bits;
      output[bits] = (value & 1) + '0';
      value >>= 1;
    }
    return true;
  }
  return false;
}

enum { N = 33 }; // assuming 'unsigned' is 32 bits

int main()
{
  char out[N];

  unsigned a[] = { 0, 1, 128, 192, 65535, 65536, -1 };

  for (size_t n=0; n<sizeof(a)/sizeof(a[0]); ++n)
  {
    if ( uint2binstr(a[n], out, N) )
    {
      printf("success: '%u' decimal is '%s' binary\n", a[n], out);
    }
    else
    {
      printf("failed to represent '%u' in binary\n", a[n]);
    }
  }
  return 0;
}

"In testa che avete, Signor di Ceprano?"
-- Rigoletto

sahil Ranka14-Jul-22 2:42

14-Jul-22 2:42

Well explained. Thank You!

CPallini14-Jul-22 3:14

CPallini

14-Jul-22 3:14

You are welcome.

"In testa che avete, Signor di Ceprano?"
-- Rigoletto

Richard MacCutchan14-Jul-22 21:15

Prey/predator c++ project

14-Jul-22 21:15

You have updated your question, but it is not clear what has changed since yesterday when we gave you the answer.

sahil Ranka12-Jul-22 0:26

Re: Prey/predator c++ project

12-Jul-22 0:26

Hello everyone, new to programming and algorithms, I have an obligation to respond to a project.
That of creating a grid in which there is a prey/predator system. The beginning is to initialize the structures of the 3 elements. We are instructed to create a (dynamic) list to which only the prey and predators entered by the user are added.
I totally don't understand how.
I had the idea of creating a static list and leaving blanks when nothing is there. But our tutor told us that it's heavier and it's better to use a dynamic list rather than a static one.

Thanks for your help.

Richard MacCutchan12-Jul-22 0:29

Re: Prey/predator c++ project

12-Jul-22 0:29

The question is not clear. What do you mean by "prey/predator system", what are "the 3 elements"? I know what both of those terms mean, but have no idea what sort of system you are supposed to design. I suggest you go back to your tutor and get a clearer description of what you are supposed to create.

Greg Utas12-Jul-22 1:20

Greg Utas

12-Jul-22 1:20

If the size of the grid is static (m x n), you can just hard-code it:

C++

enum GridOccupant {EMPTY, PREY, PREDATOR};
GridOccupant Grid[m][n] = {EMPTY};

If the grid is dynamic, the question is whether you're allowed to use things from the Standard Template Library (STL), like std::vector[^], or whether you're supposed to allocate the memory yourself.

When the user specifies the location of Prey or Predator, you create one and add it to PreyAnimals or PredatorAnimals. Again, the question is whether you're allowed to use the STL. Each Animal (the base class for Prey and Predator, if this helps your design) keeps track of its location in the Grid to make it easy to find.

You might discover that you don't need Grid at all, only its dimensions (m x n).

Robust Services Core | Software Techniques for Lemmings | Articles
The fox knows many things, but the hedgehog knows one big thing.

How to share a pointer between an EXE and a DLL

Mircea Neacsu11-Jul-22 13:50

Re: How to share a pointer between an EXE and a DLL

11-Jul-22 13:50

Apologies if the description is long; I just want to make the question clear(er).

I have a WIN32 project in which there are some applications and some DLLs that may or may not use a framework. The framework itself is statically linked to apps and DLLs and it has an important data structure that should be unique. Currently, when a DLL is loaded it has no way of knowing if the process in which it was loaded uses the framework and if the data structure has been created or not.

If DLL uses the framework and the EXE doesn't, the DLL can initialize the data structure and all is fine.
If the EXE uses the framework and the DLL doesn't, the EXE will initialize the structure and all is fine again.

The problem appears when both the DLL and the EXE use the framework and only the EXE should initialize the structure and the DLL should just somehow find the address and use the data structure.

I know I could use a shared memory area but this is not inter-process communication so I'd rather use something simpler. I even thought of setting an environment variable with the address of the structure but somehow the solution seems cheesy.

Do you have any suggestion? Is there any standard/better way of doing this?

Mircea

CPallini11-Jul-22 20:12

CPallini

11-Jul-22 20:12

Couldn't the executable simply call a function of the DLL, passing the address of the data?

"In testa che avete, Signor di Ceprano?"
-- Rigoletto

Re: How to share a pointer between an EXE and a DLL

Mircea Neacsu12-Jul-22 4:01

Re: How to share a pointer between an EXE and a DLL

12-Jul-22 4:01

Thanks for the suggestion but the DLL-EXE API is set in stone. With over 300 DLL modules changing the API is not easy/feasible. I have to find a different solution. I'll probably go with the environment variable; I'll just have to hold my nose while coding Big Grin | :-D

Mircea

Victor Nijegorodov12-Jul-22 6:22

Victor Nijegorodov

12-Jul-22 6:22

Rather than use the environment variable you could use the Registry.

Re: How to share a pointer between an EXE and a DLL

Mircea Neacsu12-Jul-22 7:52

Re: How to share a pointer between an EXE and a DLL

12-Jul-22 7:52

Interesting idea! However I would have to take care of cleaning the registry when the app finishes. The environment variable just "evaporates" on exit. Do you see any technical advantage to using the registry?

Mircea

Victor Nijegorodov12-Jul-22 8:04

Victor Nijegorodov

12-Jul-22 8:04

Delete a registry key or value is very simple.
And (almost) no one else (except your app) would like to change it.

Re: How to share a pointer between an EXE and a DLL

JudyL_MD12-Jul-22 11:31

JudyL_MD

12-Jul-22 11:31

Is the "framework" API set in stone? It sounds like it should be the one controlling the existence of this structure. You said it was statically linked to both the DLL and the EXE, hence two copies of the library, but you can force it to use a single shared memory location to hold this structure's pointer. Both the DLL and the EXE would ask the framework for a pointer and the framework would use some sort of mutual exclusion / atomic operation to synchronize its access to the shared memory location to either allocate or use the existing structure.

Be wary of strong drink. It can make you shoot at tax collectors - and miss.
Lazarus Long, "Time Enough For Love" by Robert A. Heinlein

Re: How to share a pointer between an EXE and a DLL

Mircea Neacsu12-Jul-22 12:17