Develop Custom Configuration Type

Updated 9 months ago

Creating Configuration Types

Configuration types are the validators for the values that a configuration can accept. They check and accept or deny values and serialize them to transmit or store data in files.

If you just want to access the default configuration types, use the class CTypes.

Extending the class Using the type

Extending The Base Class

To create you own configuration type you must extend the CType class.

Parameters

public object Default {get; }

This property request the default value to be used for this configuration type. It must always be acceptable or errors will occur.

This value is only used as a fallback value, realistically it will never be used.

public string TypeName {get; }

This is the name that will be shown in the menu tooltip for each configuration that uses this type.

It should represent a short text saying what values are acceptable.

protected ConfigDisplay {get; }

This property must return a new instance of a Configuration display everytime it's called.

I recommend that if you create a custom CType you also create a custom ConfigDisplay.
If you want to use the default ones, they are located under the namespace Amrv.ConfigurableCompany.Core.Display.ConfigTypes.

Methods

public bool IsValidValue(object value) {}

This method should return true whenever a value is directly compatible to the ones that this configuration type accepts.

public bool TryConvert(object value, out object result, IFormatProvider formatProvider = null) {}

This method should return true or false depending if the configuration was succesful.

With this method you should try (best effort) to convert the provided object to an acceptable value by this configuration type. If it needs to be formatted you must use the provided formatter or an invariable one (otherwise the value will depend on the region of the player).

public bool TryGetAs<T>(object value, out T result, Type type, TypeCode code, IFormatProvider formatProvider = null) {}

This method should try to convert object value (which is not granted to be acceptable by the configuration in edge cases) to a resulting object of the specified type T.
You can convert it however you want or even deny the conversion (returning false) but if you make a lossy conversion, try to make it reasonably.

TIP
Configurable Company provides the basic utility methods for conversion in the class NumberUtils.

protected bool Serialize(in object value, out string serialized) {}

This method should convert object value into a serialized string that could be converted back into the same object value without losing any information.

You must not throw any exception but return false when the serialization cannot be performed.

protected bool Deserialize(in string data, out object value) {}

This method is the reversal of the Serialize method, the provided data should be converted back into the original value.

Using The Custom Type

To use your custom type you should decide if it needs to be a singleton or be instantiated once per configuration. For optimization purposes you should try to keep it as a singleton if possible.

You can manually assign the type of any configuration to an instance of this class by setting the CConfigBuilder::Type.

// You store the instance on a variable
CType myCustomType = new CustomConfigType();

CConfig config = new CConfigBuilder() {
    // And then you assign it to any configuration
    Type = myCustomType,
}

You can also register this custom type so any configuration that contains a value of an acceptable type will be automatically assigned to this type.

This can be done by mapping it on your plugin class (make sure this method gets executed before registering any configuration).

CType.SetMapping<T>(new CustomConfigType());