FSM functor help

[Darkinyuasha1]

I just need someone to help me point out my array = newstate pointer mistake...

#include <iostream>
#include <cstddef>
using namespace std;

typedef void (*state_t)( char * inData, unsigned int , size_t );

class State
{
  state_t Curstate; 
  unsigned int label; 
  
  public:
  
  State(void)
 {  Curstate = NULL; }

  State( state_t & nState , unsigned int nID)
 {
       if( nState != NULL )
      {
          Curstate = nState;
          label = nID;
      }
      
  }

 void operation( state_t & nState , unsigned int nID)
 {
       if( nState != NULL )
      {
          Curstate = nState;
          label = nID;
      }
      
  }

 void operation ( char *& Data , size_t Size )
 {
     Curstate( Data ,label, Size );
 }

 
 unsigned int pass_ID(void) { return label; }

 };

class FSM 
{
  State state;
  
  char * data;

  state_t  * array; 

  unsigned int ID;
  
  public :

  FSM( state_t * newstate  )
 {
   array = newstate;
   ID = 0x0000000A;
   data = NULL;
   state( *array, ID); 
 }

void doAction(void)
{

   state( data , 4);

  if( ID ^ state.pass_ID() )
 {
   ++i;
   state(array[ ID % 10], ID );
 }

};
   
}

 
 void process1( char * data , unsigned int ID , size_t Size)
{
  if( data = NULL )
   data = new char [ Size];
  
  cout<<"this is your ID "<<ID<<" I will now let go to next function"<<endl;

  //passcode
  *data = "hi";

 ++ID;
}

 void process2( char * data, unsigned int ID , size_t Size )
{
  cout<<"the passcode is "<<data<<endl;

  *data = "no";
}

int main(void)
{
  state_t state[2];
 
  state[0] =&process1;
 
  state[1] =&process2;

  return 0;
}

[rags_to_riches]

if( data = NULL )

That's an assignment, not a comparison.

[Elysia]

>>*data = "hi";
Never going to work. *data is a char, "hi" is a const char[3].
Furthermore, you are getting a memory leak because you don't clean up.
Use std::string!

And the value of ID is not saved after the function. Use a reference!

>>*data = "no";
Same problem here as above.

>> State state;
>> state( *array, ID);
You are trying to use a variable as if you are calling a function.

>>if( ID ^ state.pass_ID() )
Did you really mean to use XOR here?

>>NULL
Use nullptr instead.

Lastly, indent properly.

[Darkinyuasha1]

Does visual C++ already have the nullptr idiom? eeh thanks for pointing out the flaws, but i already found them ..

[King Mir]

If it doesn't use 0. NULL is for C.

[Elysia]

Does visual C++ already have the nullptr idiom?...

It does.

If it doesn't use 0. NULL is for C.

NULL is for both C and C++.
Whether or not you want to use NULL for C++ is another matter. There are both advantages and disadvantages to that.

[King Mir]

Whether or not you want to use NULL for C++ is another matter. There are both advantages and disadvantages to that.

NULL is defined in C++, but it is a C idiom. It should not be used in C++.

The short of it is, the disadvantages out-way the advantages. Using NULL in C++ is liable to obscure bugs.

[Elysia]

Using 0 can cause confusion, as well. If you overloaded a function taking an int and one taking a pointer, 0 will call the int one, not the pointer. So how do we call the pointer function? It is likely this will cause just as much confusion as NULL.
Unfortunately, there just aren't any good solutions, other than defining your own nullptr type (which C++11 finally have fixed).
So while evil it may be, it is fine if you know what you're doing. So I call this one a preference.

[King Mir]

NULL causes more confusion, because a programmer using it will not expect it to be treated like an int, and on most compilers it will be (because it's defined as 0). I agree that nullptr is the best option, where supported.

Once such undesired conversion is identified, it can be fixed with an explicit cast to the appropriate pointer type.