[Coding] Sigscanning offsets // dynamic class padding topic

Hi

I have been wondering for a while how people exactly implement sigscanning for offsets in their projects. Ofc I know how to sigscan for them, but I would still like to use my classes, instead of scanning and using every single offset individualy. This might sound a bit vague, so here is an example of what I mean:

Code:

---

DWORD somepadding;

void sigscan()
{
    somepadding = Sigscan("\x01\x02\x03\x04");
}

class cSomeclass
{
public:
    char padding[somepadding];
    int otherstuff;
}

---

I know this code doesn't work because the char array needs a constant size. Anyways, is there any way to make this possible?
thanks!