Telling wether two files are the same in C, but cannot tell the fallacy of this algorithm

My goal is to make a function telling wether two different files are the same (they contain the same content).

This is my code:

int are_files_equal(char* filename1, char* filename2){  
if(strcmp(filename1, filename2) == 0)
        return 1;

    struct stat stat1, stat2;

    if ( stat(filename1, &stat1) != 0 || stat(filename2, &stat2) != 0)
        return -1; // error opening files

    if(stat1.st_size != stat2.st_size)
        return 0; // files are not the same as they have a different dimension
    
    FILE *file1 = fopen(filename1, "rb");
    FILE *file2 = fopen(filename2, "rb");
    if (file1 == NULL || file2 == NULL)
    {
        return -1; // error opening files
    }
    #define BYTES_TO_READ_AT_ONCE 512000
    unsigned char databuffer1[BYTES_TO_READ_AT_ONCE];
    unsigned char databuffer2[BYTES_TO_READ_AT_ONCE];
    size_t bytes;
    while ((fread(databuffer1, 1, BYTES_TO_READ_AT_ONCE, file1)) != 0)
    {
        fread(databuffer2, 1, BYTES_TO_READ_AT_ONCE, file2);
        if(memcmp(databuffer1, databuffer2, BYTES_TO_READ_AT_ONCE) != 0)
            return 0;
    }
    fclose(file1);
    fclose(file2);
    return 1;
}

But for some reason I am not able to understand, it fails. In particular the if(memcmp(databuffer1, databuffer2, BYTES_TO_READ_AT_ONCE) != 0) holds true. Why does it happen? Thank you.

>Solution :

You are not checking the correct buffer length. Try

size_t bytes;
bool matching = true;
while ((bytes = fread(databuffer1, 1, BYTES_TO_READ_AT_ONCE, file1)) != 0)
{
    if(bytes != fread(databuffer2, 1, BYTES_TO_READ_AT_ONCE, file2) ||
            memcmp(databuffer1, databuffer2, bytes) != 0) {
        matching = false;
        break;
    }
}

if(fread(databuffer2, 1, BYTES_TO_READ_AT_ONCE, file2) != 0) {
    matching = false;    // check if the whole file was read
}
fclose(file1);
fclose(file2);
return matching;