Single-data components in ECS with inheritance : downsides? - c++

Question

Let's say I have a component Position. This component has only one data, its... position. So it could be defined as this:
struct Position
{
Vector2 value;
};
But it's really redondant in access like entity.getComponent<Position>.value = {x, y};so I end up doing like so:
struct Position : Vector2
{
};
Is there any downside of doing this (syntacticly speaking) or some deviance related to the hypotetic ECS holy bible?

Answer 1

This solution is not canonical, but you can overload the "->" operator in order to access the fields of the child structure, but then you will have to take this into account if you pass Position by pointer to any functions. It is better to pass Position always by reference, then there will be no ambiguity. Operators "." and "::" cannot be overloaded, so it will not be possible to overload the behavior of p.x or p::x.
Here is an example code:
#include <iostream>
struct Vector2
{
float x = 0, y = 0;
};
struct Position
{
Vector2* const operator ->()
{
return &value;
}
Vector2 value;
};
int main()
{
Position p;
p->x = 1;
p->y = 2;
std::cout << p.value.x << " " << p.value.y << std::endl; // prints "1 2"
Position& pref = p;
pref->x = 3;
pref->y = 4;
std::cout << p.value.x << " " << p.value.y << std::endl; // prints "3 4"
Position* pp = &p;
(*pp)->x = 5;
(*pp)->y = 6;
std::cout << p.value.x << " " << p.value.y << std::endl; // prints "5 6"
return 0;
}