I have a static class with a static constructor that takes some time (10-15 seconds) to execute and fully initialize the class. In order to improve performance, I've decided to enable this static class to be explicitly initialized instead of when it is first needed so that it is ready to go when it actually needs to be used.

My first thought was to create an Initialize() method for the class, but since I already have a static constructor, this method doesn't seem to need to do anything other than be the method to call to explicitly initialize the class without accessing any of its other public methods or properties. Having a method that does nothing directly doesn't seem right to me.

Then I thought I can just move the code from the static constructor into this Initialize() method, but I'd also like the class to be initialized when it is first needed and the Initialize() method wasn't explicitly called.

To summarize, I want the following criteria to be observed:

  • I want to allow the static class to be explicitly initialized (likely by using a public Initialize() method).
  • I don't want to have to access any other public methods or properties on the class when I don't need them, even though this would initialize the static class.
  • If the class has not been explicitly initialized, I still want to initialize it when it is first needed (i.e. when accessing other public methods or properties with the intent of using the functionality or data they provide).
  • This is a helper class and using the Singleton design pattern is not necessary for my purposes.

What would be the proper way to observe the above criteria for a static class written in C#? This can also apply to other programming languages (e.g. Java), but I'm personally interested in a solution written in C#.

Simply calling any member on the class will force the constructor to run, so you could just get an existing property for example like this:

// trigger static initialization
var _ = StaticClass.MyProperty;

To make the code a little more self documenting, you could add an empty, parameterless, void method called something like Initialize so the code reads with a little more intentionality.

public static class StaticClass
{
    static StaticClass()
    {
       // any logic you want in the constructor
    }

    /// <summary> Blank Method which will force constructor of static class </summary>
    public static void Initialize() { }
}

Then you can call during startup like this:

StaticClass.Initialize();

Demo in .NET Fiddle

See Also: Is the order of static class initialization in C# deterministic?

Update: I had previously suggested putting startup logic in the Initialize method, and also calling from the ctor, but, as Smart Humanism pointed out, this will result in a race condition where Initialize is called twice