Unions are their own type. They if I remember right can hold only one type at a time? e.g. int or float but not both. Hope that helps? And pretty sure you can use them in classes just like any other data type. 
Please forgive me! =)
Please forgive me! =)
A union is a group of variables that you want to share the same memory. When you assign a value to one, all the other ones are invalidated, and you can only access one at a time. The amount of space a union actual takes is determined by the size of the largest member.
well
all elements of a union share the same memory space.
declaration of a union is same as a struct
{ Stating the obvious never helped any situation !! }
union { int x; char y; } void myfunc() { x = 65; // will print A printf("%c",y); } all elements of a union share the same memory space.
declaration of a union is same as a struct
{ Stating the obvious never helped any situation !! }
Wrong on that union example
union surface
{
int a;
float b;
char a;
};
surface front_surface;
front_surface.a = 1;
cout << front_surface.a << endl;
//this will display 1
//if you define char a like this
front_surface.c = 'A';
cout << front_surface.c << endl;
//you will get a A displayed
//if you define variable b and a also you will get garbage
//you do this you will get garbage
front_surface.a = 1;
front_surface.b = 1.00F;
front_surface.c = 'A';
cout << front_surface.a << endl;
cout << front_surface.b << endl;
cout << front_surface.c << endl;
//just define only one variable at a time not two or three or more
hope that helps
from what I have read unions are used to help save memory? But I could be wrong. And if thats the case look at my signature!
Please forgive me! =)
Edited by - mars_999 on October 17, 2001 3:54:50 PM
union surface
{
int a;
float b;
char a;
};
surface front_surface;
front_surface.a = 1;
cout << front_surface.a << endl;
//this will display 1
//if you define char a like this
front_surface.c = 'A';
cout << front_surface.c << endl;
//you will get a A displayed
//if you define variable b and a also you will get garbage
//you do this you will get garbage
front_surface.a = 1;
front_surface.b = 1.00F;
front_surface.c = 'A';
cout << front_surface.a << endl;
cout << front_surface.b << endl;
cout << front_surface.c << endl;
//just define only one variable at a time not two or three or more
hope that helps
from what I have read unions are used to help save memory? But I could be wrong. And if thats the case look at my signature!
Please forgive me! =)
Edited by - mars_999 on October 17, 2001 3:54:50 PM
Yip, unions are used to save memory - but you don't want to think of it as "invalidating data" when you change it. For example
I can access the top or bottom 16 bits of the union with highword/loword, or the whole 32bits with wholeInt. In this case I am using a union and struct as tricks for data access.
You also commonly see
You have a union within a struct and use the variable isFloat to determine the state of the union.
Hope it helps
Brad
Edited by - brad_beveridge on October 17, 2001 7:08:59 PM
Edited by - brad_beveridge on October 17, 2001 7:10:15 PM
Edited by - brad_beveridge on October 17, 2001 7:11:00 PM
Edited by - brad_beveridge on October 17, 2001 7:12:05 PM
union { struct { unsigned short highword; unsigned short loword; } unsigned int wholeInt;} I can access the top or bottom 16 bits of the union with highword/loword, or the whole 32bits with wholeInt. In this case I am using a union and struct as tricks for data access.
You also commonly see
struct { union{ float f; int i; } bool isFloat;} You have a union within a struct and use the variable isFloat to determine the state of the union.
Hope it helps
Brad
Edited by - brad_beveridge on October 17, 2001 7:08:59 PM
Edited by - brad_beveridge on October 17, 2001 7:10:15 PM
Edited by - brad_beveridge on October 17, 2001 7:11:00 PM
Edited by - brad_beveridge on October 17, 2001 7:12:05 PM
can somebody run off some common uses for unions? I see how they work, but I don''t see any important uses...
The way I (and XLib, coincidentally) use them is like this: To get information about any even that happens you have to pass a pointer. The event function will copy the information into this pointer, and you''ll use a certain part of it depending upon what the event is. Now, this is where unions come in. I don''t know what event will occur, so I don''t know what size the struct I have to pass is. So, I pass a union that looks like this (sort of):
Now, if I find out that the event happened to be a Mouse event, I know that EventThingy.Mouse will have valid data. This prevents me from declaring 10 different structs, one for each event, and saves memory.
Another common use is like this:
That only uses 12 bytes (like normal), but now you can use the vertex as either an array or use as an individual element. This is also common with Matrices.
[Resist Windows XP''s Invasive Production Activation Technology!]
union EventThingy { struct { int X, Y, Action; } Mouse; struct { int Key; } Keyboard;}; Now, if I find out that the event happened to be a Mouse event, I know that EventThingy.Mouse will have valid data. This prevents me from declaring 10 different structs, one for each event, and saves memory.
Another common use is like this:
union Vertex { float V[3]; struct { float X, Y, Z; };}; That only uses 12 bytes (like normal), but now you can use the vertex as either an array or use as an individual element. This is also common with Matrices.
[Resist Windows XP''s Invasive Production Activation Technology!]
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
