Unlike classes, a struct
is a value type, and is created on the local stack and not on the managed heap, by default. This means that once the specific stack goes out of scope, the struct
is de-allocated. Contained reference types of de-allocated struct
s are also swept, once the GC determines they are not longer referenced to by the struct
.
struct
s cannot inherit and cannot be bases for inheritance, they are implicitly sealed, and also cannot include protected
members. However, a struct
can implement an interface, as classes do.
public struct Vector
{
public int X;
public int Y;
public int Z;
}
public struct Point
{
public decimal x, y;
public Point(decimal pointX, decimal pointY)
{
x = pointX;
y = pointY;
}
}
struct
instance fields can be set via a parametrized constructor or individually after struct
construction.
Private members can only be initialized by the constructor.
struct
defines a sealed type that implicitly inherits from System.ValueType.
Structs cannot inherit from any other type, but they can implement interfaces.
Structs are copied on assignment, meaning all data is copied to the new instance and changes to one of them are not reflected by the other.
A struct cannot be null
, although it can used as a nullable type:
Vector v1 = null; //illegal
Vector? v2 = null; //OK
Nullable<Vector> v3 = null // OK
Structs can be instantiated with or without using the new
operator.
//Both of these are acceptable
Vector v1 = new Vector();
v1.X = 1;
v1.Y = 2;
v1.Z = 3;
Vector v2;
v2.X = 1;
v2.Y = 2;
v2.Z = 3;
However, the new
operator must be used in order to use an initializer:
Vector v1 = new MyStruct { X=1, Y=2, Z=3 }; // OK
Vector v2 { X=1, Y=2, Z=3 }; // illegal
A struct can declare everything a class can declare, with a few exceptions:
struct
instance fields can be set via a parameterized constructor or individually after struct
construction. Private members can only be initialized by the constructor.With constructor:
Vector v1 = new Vector();
v1.X = 1;
v1.Y = 2;
v1.Z = 3;
Console.WriteLine("X = {0}, Y = {1}, Z = {2}",v1.X,v1.Y,v1.Z);
// Output X=1,Y=2,Z=3
Vector v1 = new Vector();
//v1.X is not assigned
v1.Y = 2;
v1.Z = 3;
Console.WriteLine("X = {0}, Y = {1}, Z = {2}",v1.X,v1.Y,v1.Z);
// Output X=0,Y=2,Z=3
Point point1 = new Point();
point1.x = 0.5;
point1.y = 0.6;
Point point2 = new Point(0.5, 0.6);
Without constructor:
Vector v1;
v1.Y = 2;
v1.Z = 3;
Console.WriteLine("X = {0}, Y = {1}, Z = {2}",v1.X,v1.Y,v1.Z);
//Output ERROR "Use of possibly unassigned field 'X'
Vector v1;
v1.X = 1;
v1.Y = 2;
v1.Z = 3;
Console.WriteLine("X = {0}, Y = {1}, Z = {2}",v1.X,v1.Y,v1.Z);
// Output X=1,Y=2,Z=3
Point point3;
point3.x = 0.5;
point3.y = 0.6;
If we use a struct with its constructor, we aren't going to have problems with unassigned field (each unassigned field has null value).
Unlike classes, a struct doesn't have to be constructed, i.e. there is no need to use the new keyword, unless you need to call one of the constructors. A struct does not require the new keyword because is a value-type and thus cannot be null.
public interface IShape
{
decimal Area();
}
public struct Rectangle : IShape
{
public decimal Length { get; set; }
public decimal Width { get; set; }
public decimal Area()
{
return Length * Width;
}
}
Sinse structs are value types all the data is copied on assignment, and any modification to the new copy does not change the data for the original copy. The code snippet below shows that p1
is copied to p2
and changes made on p1
does not affect p2
instance.
var p1 = new Point {
x = 1,
y = 2
};
Console.WriteLine($"{p1.x} {p1.y}"); // 1 2
var p2 = p1;
Console.WriteLine($"{p2.x} {p2.y}"); // Same output: 1 2
p1.x = 3;
Console.WriteLine($"{p1.x} {p1.y}"); // 3 2
Console.WriteLine($"{p2.x} {p2.y}"); // p2 remain the same: 1 2