Nrthugu

Hi I'm trying to create a function that takes a void* pointer as parameter and print all element of the vector using printf

the declaration of the function is:

void print_vec(void *vec,int dime_se,int dime,char *format);

and implementation is:

void print_vec(void *vec,int dime_se,int dime,char *format){

    for(int i=0;i<dime;i++)

        printf(format,*(vec+dime_se*i));

}

the problem is that when i compile the compiler returns:

error: invalid use of void expression printf(format,*(vec+dime_se*i));

So the question there is a way to do this task without make this?

void print_vec(void *vec,int dime_se,int dime,char *format){

for(int i=0;i<dime;i++)

    switch(format[1]){

        case 'c':

            printf(format,*((char*)(vec+dime_se*i)));

            break;



        case 'd':

            printf(format,*((unsigned int*)(vec+dime_se*i)));

            break;





        case 's':

            printf(format,*((char**)(vec+dime_se*i)));

            break;



        case 'i':

            printf(format,*((int*)(vec+dime_se*i)));

            break;







        case 'f':

            printf(format,*((float*)(vec+dime_se*i)));

            break;



        default:

            break;

    }

}

You don't need the dime_se param. The size is more or less implicit in the format character.
Also, if you cast to the right pointer type you can let C do the scaling instead of doing it
manually:

#include <stddef.h>

#include <stdio.h>

void print_vec(void const*vec,size_t dime,char const*format)

{

    for(size_t i=0;i<dime;i++)

        switch(format[1]){

            #define X(Char,Tp) break;case Char: printf(format,((Tp*)vec)[i])

            X('c',char);

            X('d',unsigned);

            X('s',char*);

            X('i',int);

            X('f',float);

            break;default: return;

            #undef X

        }

}

I said more or less because type promotion kind of blurs the line. %f could well mean you've got a double instead of a float.

If you want to use a dime_se parameter to disambiguate that case, you can:

         break;case 'f': 

                if(sizeof(double)==dime_se)

                     printf(format,((double*)vec)[i]);

                else { 

                     assert(sizeof(float)==dime_se); 

                     printf(format,((float*)vec)[i]); 

                }

but I'm not sure if a print_vec function like that is a good idea.