1473
votes

I've read around about const and static readonly fields. We have some classes which contain only constant values. They are used for various things around in our system. So I am wondering if my observation is correct:

Should these kind of constant values always be static readonly for everything that is public? And only use const for internal/protected/private values?

What do you recommend? Should I maybe even not use static readonly fields, but rather use properties maybe?

19
Here's a very interesting single case I just found in favor of static readonly: try using a const inside an IEnumerator which would trigger an unrecheable yield and you'll get a dreaded "Internal compiler error". I didn't test the code outside Unity3D, but I trust this is either a mono or .NET bug. It is a c# issue nevertheless.cregox
another difference is that you can use a const string in a switch, but not a static readonly stringflagg19
static readonly can't be used in switch-case statement as case variable, const is required for this purpose.Mostafiz Rahman
static readonly can't be used as attribute parameter, tooDread Boy

19 Answers

994
votes

public static readonly fields are a little unusual; public static properties (with only a get) would be more common (perhaps backed by a private static readonly field).

const values are burned directly into the call-site; this is double edged:

  • it is useless if the value is fetched at runtime, perhaps from config
  • if you change the value of a const, you need to rebuild all the clients
  • but it can be faster, as it avoids a method call...
  • ...which might sometimes have been inlined by the JIT anyway

If the value will never change, then const is fine - Zero etc make reasonable consts ;p Other than that, static properties are more common.

254
votes

I would use static readonly if the Consumer is in a different assembly. Having the const and the Consumer in two different assemblies is a nice way to shoot yourself in the foot.

210
votes

A few more relevant things to be noted:

const int a

  • must be initialized.
  • initialization must be at compile time.

readonly int a

  • can use a default value, without initializing.
  • initialization can be done at run time (Edit: within constructor only).
191
votes

This is just a supplement to the other answers. I will not repeat them (now four years later).

There are situations where a const and a non-const have different semantics. For example:

const int y = 42;

static void Main()
{
  short x = 42;
  Console.WriteLine(x.Equals(y));
}

prints out True, whereas:

static readonly int y = 42;

static void Main()
{
  short x = 42;
  Console.WriteLine(x.Equals(y));
}

writes False.

The reason is that the method x.Equals has two overloads, one that takes in a short (System.Int16) and one that takes an object (System.Object). Now the question is whether one or both apply with my y argument.

When y is a compile-time constant (literal), the const case, it becomes important that there does exist an implicit conversion from int to short provided that the int is a constant, and provided that the C# compiler verifies that its value is within the range of a short (which 42 is). See Implicit constant expression conversions in the C# Language Specification. So both overloads have to be considered. The overload Equals(short) is preferred (any short is an object, but not all object are short). So y is converted to short, and that overload is used. Then Equals compares two short of identical value, and that gives true.

When y is not a constant, no implicit conversion from int to short exists. That's because in general an int may be too huge to fit into a short. (An explicit conversion does exist, but I didn't say Equals((short)y), so that's not relevant.) We see that only one overload applies, the Equals(object) one. So y is boxed to object. Then Equals is going to compare a System.Int16 to a System.Int32, and since the run-time types do not even agree, that will yield false.

We conclude that in some (rare) cases, changing a const type member to a static readonly field (or the other way, when that is possible) can change the behavior of the program.

91
votes

One thing to note is const is restricted to primitive/value types (the exception being strings).

31
votes

Static Read Only:

The value can be changed through a static constructor at runtime. But not through a member function.

Constant:

By default static. A value cannot be changed from anywhere (constructor, function, runtime, etc. nowhere).

Read Only:

The value can be changed through a constructor at runtime. But not through a member function.

You can have a look at my repository: C# property types.

27
votes

The readonly keyword is different from the const keyword. A const field can only be initialized at the declaration of the field. A readonly field can be initialized either at the declaration or in a constructor. Therefore, readonly fields can have different values depending on the constructor used. Also, while a const field is a compile-time constant, the readonly field can be used for runtime constants

Short and clear MSDN reference here

17
votes

const and readonly are similar, but they are not exactly the same.

A const field is a compile-time constant, meaning that that value can be computed at compile-time. A readonly field enables additional scenarios in which some code must be run during construction of the type. After construction, a readonly field cannot be changed.

For instance, const members can be used to define members like:

struct Test
{
    public const double Pi = 3.14;
    public const int Zero = 0;
}

Since values like 3.14 and 0 are compile-time constants. However, consider the case where you define a type and want to provide some pre-fab instances of it. E.g., you might want to define a Color class and provide "constants" for common colors like Black, White, etc. It isn't possible to do this with const members, as the right hand sides are not compile-time constants. One could do this with regular static members:

public class Color
{
    public static Color Black = new Color(0, 0, 0);
    public static Color White = new Color(255, 255, 255);
    public static Color Red   = new Color(255, 0, 0);
    public static Color Green = new Color(0, 255, 0);
    public static Color Blue  = new Color(0, 0, 255);
    private byte red, green, blue;

    public Color(byte r, byte g, byte b) => (red, green, blue) = (r, g, b);
}

But then there is nothing to keep a client of Color from mucking with it, perhaps by swapping the Black and White values. Needless to say, this would cause consternation for other clients of the Color class. The "readonly" feature addresses this scenario.

By simply introducing the readonly keyword in the declarations, we preserve the flexible initialization while preventing client code from mucking around.

public class Color
{
    public static readonly Color Black = new Color(0, 0, 0);
    public static readonly Color White = new Color(255, 255, 255);
    public static readonly Color Red   = new Color(255, 0, 0);
    public static readonly Color Green = new Color(0, 255, 0);
    public static readonly Color Blue  = new Color(0, 0, 255);
    private byte red, green, blue;

    public Color(byte r, byte g, byte b) => (red, green, blue) = (r, g, b);
}

It is interesting to note that const members are always static, whereas a readonly member can be either static or not, just like a regular field.

It is possible to use a single keyword for these two purposes, but this leads to either versioning problems or performance problems. Assume for a moment that we used a single keyword for this (const) and a developer wrote:

public class A
{
    public static const C = 0;
}

and a different developer wrote code that relied on A:

public class B
{
    static void Main() => Console.WriteLine(A.C);
}

Now, can the code that is generated rely on the fact that A.C is a compile-time constant? I.e., can the use of A.C simply be replaced by the value 0? If you say "yes" to this, then that means that the developer of A cannot change the way that A.C is initialized -- this ties the hands of the developer of A without permission.

If you say "no" to this question then an important optimization is missed. Perhaps the author of A is positive that A.C will always be zero. The use of both const and readonly allows the developer of A to specify the intent. This makes for better versioning behavior and also better performance.

12
votes

My preference is to use const whenever I can, which, as mentioned in previous answers, is limited to literal expressions or something that does not require evaluation.

If I hit up against that limitation, then I fallback to static readonly, with one caveat. I would generally use a public static property with a getter and a backing private static readonly field as Marc mentions here.

7
votes

A static readonly field is advantageous when exposing to other assemblies a value that might change in a later version.

For instance, suppose assembly X exposes a constant as follows:

public const decimal ProgramVersion = 2.3;

If assembly Y references X and uses this constant, the value 2.3 will be baked into assembly Y when compiled. This means that if X is later recompiled with the constant set to 2.4, Y will still use the old value of 2.3 until Y is recompiled. A static readonly field avoids this problem.

Another way of looking at this is that any value that might change in the future is not constant by definition, and so should not be represented as one.

7
votes

Const: Const is nothing but "constant", a variable of which the value is constant but at compile time. And it's mandatory to assign a value to it. By default a const is static and we cannot change the value of a const variable throughout the entire program.

Static ReadOnly: A Static Readonly type variable's value can be assigned at runtime or assigned at compile time and changed at runtime. But this variable's value can only be changed in the static constructor. And cannot be changed further. It can change only once at runtime

Reference: c-sharpcorner

5
votes

Const: Constant variable values have to be defined along with the declaration and after that it won't change.const are implicitly static, so without creating a class instance we can access them. This has a value at compile time.

ReadOnly: We can define read-only variable values while declaring as well as using the constructor at runtime. Read-only variables can't access without a class instance.

Static readonly: We can define static readonly variable values while declaring as well as only through a static constructor, but not with any other constructor. We can also access these variables without creating a class instance (as static variables).

Static readonly will be better choice if we have to consume the variables in different assemblies. Please check the full details in the below blog post:

Const Strings – a very convenient way to shoot yourself in the foot

4
votes

There is a minor difference between const and static readonly fields in C#.Net

const must be initialized with value at compile time.

const is by default static and needs to be initialized with constant value, which can not be modified later on. It can not be used with all datatypes. For ex- DateTime. It can not be used with DateTime datatype.

public const DateTime dt = DateTime.Today;  //throws compilation error
public const string Name = string.Empty;    //throws compilation error
public static readonly string Name = string.Empty; //No error, legal

readonly can be declared as static, but not necessary. No need to initialize at the time of declaration. Its value can be assigned or changed using constructor once. So there is a possibility to change value of readonly field once (does not matter, if it is static or not), which is not possible with const.

3
votes

const:

  1. value should be given upon declaration
  2. compile time constant

readonly:

  1. value can be given upon declaration or during runtime using constructors.The value may vary depend upon the constructor used.
  2. run time constant
1
votes

A const (being determined at compile-time) can be used in cases where a readonly static can't, like in switch statements, or attribute constructors. This is because readonly fields are only resolved at run-time, and some code constructs require compile time assurance. A readonly static can be calculated in a constructor, which is often an essential and useful thing. The difference is functional, as should be their usage in my opinion.

In terms of memory allocation, at least with strings (being a reference type), there seems to be no difference in that both are interned and will reference the one interned instance.

Personally, my default is readonly static, as it makes more semantic and logical sense to me, especially since most values are not needed at compile time. And, by the way, public readonly statics are not unusual or uncommon at all as the marked answer states: for instance, System.String.Empty is one.

1
votes

Another difference between declaring const and static readonly is in memory allocation.

A static field belongs to the type of an object rather than to an instance of that type. As a result, once the class is referenced for the first time, the static field would "live" in the memory for the rest of time, and the same instance of the static field would be referenced by all instances of the type.

On the other hand, a const field "belongs to an instance of the type.

If memory of deallocation is more important for you, prefer to use const. If speed, then use static readonly.

0
votes

Constants are like the name implies, fields which don't change and are usually defined statically at compile time in the code.

Read-only variables are fields that can change under specific conditions.

They can be either initialized when you first declare them like a constant, but usually they are initialized during object construction inside the constructor.

They cannot be changed after the initialization takes place, in the conditions mentioned above.

Static read-only sounds like a poor choice to me since, if it's static and it never changes, so just use it public const. If it can change then it's not a constant and then, depending on your needs, you can either use read-only or just a regular variable.

Also, another important distinction is that a constant belongs to the class, while the read-only variable belongs to the instance!

0
votes

There is one important question, that is not mentioned anywhere in the above answers, and should drive you to prefer "const" especially for basic types like "int", "string" etc.

Constants can be used as Attribute parameters, static readonly field not!

Azure functions HttpTrigger, not using HttpMethods class in attribute

If only microsoft used constants for Http's GET, POST, DELETE etc.

It would be possible to write

[HttpTrigger(AuthorizationLeve.Anonymous,  HttpMethods.Get)] // COMPILE ERROR: static readonly, 

But instead I have to resort to

[HttpTrigger(AuthorizationLeve.Anonymous,  "GET")] // STRING

Or use my own constant:

public class HttpConstants
{
    public const string Get = "GET";
}

[HttpTrigger(AuthorizationLeve.Anonymous,  HttpConstants.Get)] // Compile FINE!
0
votes

Use const if you can provide a compile-time constant:

private const int Total = 5;

Use static readonly if you need your value evaluated during run-time:

private static readonly int GripKey = Animator.StringToHash("Grip");

This will give a compile error because it is impossible to get the value at compile-time.

private const int GripKey = Animator.StringToHash("Grip");