Unions

[dalawh]

Since unions occupy the same address space, what happens when you have something like...

union unionTest{
bool tF; //1 byte
char alpha; //1 byte
int num; //4 byte
};

What happens in this scenario? In all the scenarios, there would of been an array of 4 for tF and alpha to match up with num, but in this scenario, what happens?

[firedraco]

An array of 4? I don't understand.

You'd just have this:

[b][ ][ ][ ]
[c][ ][ ][ ]
[i][i][i][i]

There is no array.

[TheIdeasMan]

You haven't declared a union. You need the union keyword for that (not struct).

With a union, the compiler set aside enough space for the largest member.

[Nexius]

In unions, the members share the same space. Unions are useful for defining some kind of abstract data type.

union shared_data
{
    char character;
    bool boolean;
    int integer;
    float decimal;
};

For differently-sized members, the union takes the size of the largest member.
Union's members may not have some sort of constructor. They must be C-style POD.
Structures may also be a member of union.

[TheIdeasMan]

Unions' members may not have some sort of constructor. They must be C-style POD.

This article is about C++11 unions

http://www.informit.com/guides/content.aspx?g=cplusplus&seqNum=556

One observation is that there are a lot of new things in C++11.

[dalawh]

@firedraco
In the example in the tutorials, they had an array of 4, which is created "int arrayTest[4];".

That was what I was looking for. I wanted to know how it would look like if first two data types were smaller than the last. Thanks for explaining this.