Thread: Overloading Subscripting Operators?????

The problem is that you can overload operator[], but there is no operator[][] for you to overload. The typical workaround is to use a proxy object, i.e., you overload operator[] to return an object of a class that is implementation detail. This class will in turn have operator[] overloaded to return the "real" object that you want to return.

Of course, this might seem more work than desirable just to get this syntactic sugar. If so, another approach is to cop out by saying: let's overload operator() to do the job instead, e.g., use g(10, 5) instead of g[10][5].

Originally Posted by laserlight

The problem is that you can overload operator[], but there is no operator[][] for you to overload. The typical workaround is to use a proxy object, i.e., you overload operator[] to return an object of a class that is implementation detail. This class will in turn have operator[] overloaded to return the "real" object that you want to return.

Of course, this might seem more work than desirable just to get this syntactic sugar. If so, another approach is to cop out by saying: let's overload operator() to do the job instead, e.g., use g(10, 5) instead of g[10][5].

Idle question could you do it like g[10, 5] in C++?

Tim S.

Originally Posted by stahta01

Idle question could you do it like g[10, 5] in C++?

Yeah. It would be equivalent to g[5].

Originally Posted by stahta01

Idle question could you do it like g[10, 5] in C++?

Tim S.

I presume you want x[n, m] to mean the same x[n][m]. If that's the case, then you can't do it directly with integers. You can do it with at least one non-integer type (i.e. a class), but it will result in ugly hacks.
One method I can think of is basically x[index(n), m] where index's constructor basically stores n in a global variable which then x's operator can pick up. Ugly, hacky, but it works.

unless you're already overloading the funciton-call operator with two int parameters, you can do something more like this:

Code:

Cell GameOfLife::operator ()(int row, int column)
{
    // return the relevant cell here
}

This is what I do when I need what amounts to a two-dimensional subscript operator.

On the other hand, why do you need to use operators? what's wrong with just having a GetCell(row, column) member function?

Originally Posted by Elkvis

On the other hand, why do you need to use operators? what's wrong with just having a GetCell(row, column) member function?

Why not? It's used to make the syntax more natural. Why say vector.get(x, y) instead of vector(x, y) or vector[x][y]?