如何在此处打印结构的名称变量？

Question

I have an ArrayList struct and Department struct that go as follows:

typedef struct ArrayList {
   void** elements;
   int size;
   int length;
} ArrayList;

typedef struct Department {
   char* name;
   ArrayList* courses;
} Department;

To print my list, I'm using these two methods:

void* get(ArrayList* arraylist, int i) {
    if (i < 0 || i >= arraylist -> size) {
        return (void*) NULL;
    }
    return arraylist -> elements[i];
}

void printAL(ArrayList* arraylist) {
    for (int i = 0; i < arraylist -> size; i++) {
        printf("%s\n", (char*) get(arraylist, i));
    }
}

The issue I'm facing, however, is that when I add a Department to my ArrayList, the line 'return arraylist -> elements[i];'returns the address of that struct. I'm trying to get it to print the name of the struct using 'return arraylist -> elements[i] -> name' but I keep getting a warning that I'm dereferencing a void* pointer, followed by an error that says 'request for member 'name' in something not a structure'. This obviously means that 'arraylist -> elements[i]' isn't a struct but rather an address. How can I reference the name of the struct at that address then? I'm quite confused because of the double pointer in the ArrayList struct.

TIA!

Answer 1

You need different printing functions for each different type of data element that could be in the ArrayList . You need one function to print departments; you need a different function to print courses. You pass the function pointer to the printing function — printAL() — along with a pointer to other data (which in this case is probably just a FILE * , but could be a more general structure).

This is analogous to the qsort() function in standard C. It can sort any data type; you just need to pass it a different comparator function for different data types.

Like this:

#include <stdio.h>
#include <stdlib.h>

typedef struct ArrayList
{
    void **elements;
    int size;           /* Allocated size */
    int length;         /* Space in use */
} ArrayList;

typedef struct Department
{
    char *name;
    ArrayList *courses;
} Department;

static void *get(ArrayList *arraylist, int i)
{
    if (i < 0 || i >= arraylist->size)
        return NULL;
    return arraylist->elements[i];
}

static void printAL(ArrayList *arraylist, void (*function)(const void *data, void *thunk), void *thunk)
{
    for (int i = 0; i < arraylist->length; i++)
    {
        (*function)(get(arraylist, i), thunk);
    }
}

static void put(ArrayList *al, void *data)
{
    if (al->length >= al->size)
    {
        size_t new_size = (al->size + 2) * 2;
        void  *new_data = realloc(al->elements, new_size * sizeof(void *));
        if (new_data == 0)
        {
            fprintf(stderr, "Failed to allocate %zu bytes memory\n", new_size * sizeof(void *));
            exit(1);
        }
        al->elements = new_data;
        al->size = new_size;
    }
    al->elements[al->length++] = data;
}

/*
typedef struct Course
{
    const char *name;
    const char *code;
    // ...
} Course;

static void print_courseinfo(const void *data, void *thunk)
{
    FILE *fp = thunk;
    const Course *cp = data;
    fprintf(fp, " - %s (%s)\n", cp->name, cp->code);
}
*/

static void print_deptname(const void *data, void *thunk)
{
    FILE *fp = thunk;
    const Department *dp = data;
    fprintf(fp, "Name: %s\n", dp->name);
    /*
    if (dp->courses != 0)
        printAL(dp->courses, print_courseinfo, thunk);
    */
}

int main(void)
{
    ArrayList al = { 0, 0, 0 };
    Department dl[] =
    {
        { "Engineering", 0 },
        { "Physics", 0 },
        { "Mathematics", 0 },
        { "Chemistry", 0 },
        { "Biology", 0 },
        { "English", 0 },
        { "Computational Astronomy and Universe-Scale Data Modelling", 0 },
        { "Economics", 0 },
    };
    enum { DL_SIZE = sizeof(dl) / sizeof(dl[0]) };
    for (size_t i = 0; i < DL_SIZE; i++)
        put(&al, &dl[i]);
    printAL(&al, print_deptname, stdout);
    return 0;
}

Sample output:

Name: Engineering
Name: Physics
Name: Mathematics
Name: Chemistry
Name: Biology
Name: English
Name: Computational Astronomy and Universe-Scale Data Modelling
Name: Economics

You didn't document what the length and size members of the ArrayList represent. I've annotated what I've assumed, but I had to change the printAL() function to iterate over length elements instead of size elements, so I may have inverted the meaning you intended. It's easy to reverse them. I tend to use names like max_elements and num_elements for the job; it is more obvious what they're for, perhaps, since length and size are ambiguous or even equivalent in many contexts.

There's skeletal code in there to show how to handle the ArrayList of courses offered by each department. I couldn't be bothered to write code to initialize a separate ArrayList for each department, though it wouldn't be particularly hard to do.

I still prefer the pre-standard notation (*funcptr)(arg1, arg2) notation for invoking a function designated by a function pointer — it was necessary when I learned C, and I still find it clearer than the alternative. You're excused if you prefer funcptr(arg1, arg2) instead, though that can leave me wondering where funcptr is defined.

You can also find some code closely related to what you're doing in my SOQ (Stack Overflow Questions) repository on GitHub as files aomcopy.c , aomcopy.h , aommngd.c , aommngd.h , aomptr.c , aomptr.h , aoscopy.c , aoscopy.h , aosptr.c and aosptr.h in the src/libsoq sub-directory.

• aomcopy.c , aomcopy.h : array of memory blocks, copied.
• aommngd.c , aommngd.h : array of memory blocks, managed.
• aomptr.c , aomptr.h : array of memory blocks, 'raw'.
• aoscopy.c , aoscopy.h : array of strings, copied.
• aosptr.c . aosptr.h : array of strings, 'raw'.

The 'raw' versions simply take the pointer passed and store it. The onus is on the user to ensure the data pointed at remains valid while the array lasts. The 'copied' versions allocate a simple copy of the data passed to it; it doesn't matter if the data passed is reused to store the next value. The 'managed' version calls user-defined functions to create copies of the data structures. This would be necessary if you have a complex structure (like a department) where you need a 'deep copy' of the data.