类型为'double *'和'double'的无效操作数到二进制'operator +'

Question

Here is my template class I am trying to implement:

#ifndef __ENCRYPTION_HEADER_INCLUDED__
#define __ENCRYPTION_HEADER_INCLUDED__

#include <iostream>  // in/out 
#include <fstream>   // file
#include <string>    // string
#include <sstream>   // string stream

using namespace std; // standard namespacing eliminates std:: from code


template <class mytemp>
class encryption 
{   
    public:

        mytemp * p;   

        encryption(); 

        mytemp encrypt_function (mytemp);
};


template <class mytemp>
encryption<mytemp>::encryption ()
{
    string line_in, file_name;
    mytemp input_value;

    cout << "input file name which contains encryption keys: " ;
    getline(cin, file_name);
    cout << '\n';   
    ifstream val_file (file_name.c_str());

    p = new mytemp[10];

    if (val_file.is_open())
    {
        while (getline(val_file, line_in))
        {
            istringstream line_out(line_in);
            line_out >> hex >> input_value;

            *p = input_value;
            p++;
        }

        p -= 10;
        val_file.close();
    }
    else
    {
        cout << "file not found" << '\n';
    }
}

template <class mytemp>
mytemp encryption<mytemp>::encrypt_function (mytemp input_en)
{
    mytemp output_en = input_en;

    for (mytemp k = 0; k < 10; k++)
    {
        output_en ^= *(p + k); // does not work with double or float but works with int?
    }
    return output_en;
}

#endif // __ENCRYPTION_HEADER_INCLUDED__ 

the line shown below is the one I am having a problem with:

output_en ^= *(p + k); // does not work with double or float but works with int? 

My understanding is that this is correct for an integer but it does not work for double or float and I am wondering if there is an easy way to get this to work other than rewriting the code to increment the pointer instead of incrementing through the pointed to values?

Answer 1

Pointer arithmetic expects a signed integer (ie, std::ptrdiff_t ) type, which double or float are neither. In this loop here:

for (mytemp k = 0; k < 10; k++)
{
    output_en ^= *(p + k); // does not work with double or float but works with int?
}

It makes no sense to use a float or double. Just use a signed int (which int is).

Answer 2

Bitwise xor isn't defined for those types. If you're interpreting the inputs as just bits, then you should do what memcpy() or memset() do and just interpret them as an untyped array of bytes. That is, your template wrapper could determine the address and size of the data and pass those to a function that xors a block of memory.