:: Re: [DNG] Experiencing with GtkBuil…
Top Pagina
Delete this message
Reply to this message
Auteur: Edward Bartolo
Datum:  
Aan: Rainer Weikusat
CC: dng
Onderwerp: Re: [DNG] Experiencing with GtkBuilder
Yes, of course, C structures can be declared that way, but the fact
remains that the contents of the ancestor's structure are not merged
into the heir. This means, to access a member N ancestors deep, one
has to specify all of them in the correct order. The purpose of
inheritance is to avoid this requirement and to make objects behave in
a polymorphic way without actually specifying or assigning any virtual
methods used.



On 22/11/2015, Rainer Weikusat <rainerweikusat@???> wrote:
> Edward Bartolo <edbarx@???> writes:
>> Is it possible to use classes and objects while using gtk2/3? I
>> noticed that only functions are used and that class use is ovoided by
>> prepending functions with a group string. As far as I know, C can
>> still use structs that are attached to a set of data and a set of
>> functions, but without inheritance and polymorphyism.
>
> C structs can be embedded into each other which means they can 'inherit'
> other structs, eg, for a totally contrived example
>
> struct something {
>     int number0, number1;
> };

>
> struct another {
>     struct something parent;
>         int drei;
> };

>
> Assuming
>
> struct another *pa
>
> the elements of the embedded struct can now be accessed as
>
> pa->parent.number0
>
> and a this pointer can be casted to a struct something * pointing at the
> first member which can thus be passed to functions expecting such a
> pointer as an argument.
>
> Further, the relatively puny "polymorphism support" of Pascal or C++ is
> easily implemented without help from the compiler. One can create
> structures of function pointers like this
>
> struct something_vtable {
>     int (*calc)(struct something *);
> };

>
> embed a pointer to that into the structure. Derived 'classes' can have
> their own vtable structures and hence, supply alternate functions to
> execute.
>
> This mechanism as certain tendency make C++ developers go bezerk with a
> particular type of strong rage, but it's entirely usuable in practice,
> although more work than having it all laid out for oneself.
>
> Working example:
>
> ------------
> #include <stdio.h>
>
> /* metainformation */
> struct something;
>
> struct something_vtable {
>     int (*calc)(struct something *);
> };

>
> /*  something */
> struct something {
>     struct something_vtable *vt;
>     int number0, number1;
> };

>
> static int calc_something(struct something *sth)
> {
>     return sth->number0 + sth->number1;
> }

>
> static struct something_vtable sth_vt = {
>     .calc = calc_something
> };

>
> static void init_something(struct something *sth, int n0, int n1)
> {
>     sth->vt = &sth_vt;
>     sth->number0 = n0;
>     sth->number1 = n1;
> }

>
> static int calc(struct something *sth)
> {
>     return sth->vt->calc(sth);
> }

>
> /*  another */
> struct another {
>     struct something parent;
>     int drei;
> };

>
> static int calc_another(struct something *sth)
> {
>     struct another *ath = (void *)sth;
>     return ath->drei + calc_something(sth);
> }

>
> static struct something_vtable ath_vt = {
>     .calc = calc_another
> };

>
> static void init_another(struct another *ath, int n0, int n1, int d)
> {
>     init_something((void *)ath, n0, n1);
>     ath->parent.vt = &ath_vt;
>     ath->drei = d;
> }

>
> /*  example program */
> void print_result(struct something *sth)
> {
>     printf("Result: %d\n", calc(sth));
> }

>
> int main(void)
> {
>     struct something sth;
>     struct another ath;

>
>     init_something(&sth, 1, 2);
>     init_another(&ath, 1, 2, 3);

>
>     print_result(&sth);
>     print_result((void *)&ath);

>
>     return 0;
> }
> _______________________________________________
> Dng mailing list
> Dng@???
> https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/dng

>