Role of Value Transformers
Value transformer classes are used to transform the value of an object in some manner. This is particularly useful when using Cocoa bindings and creating a binding between a model property of a controller and a user interface element, or another controller object. By using the built-in value transformers, or creating custom value transformers, you can further reduce the amount of glue code required by your application.
For example, it’s often necessary to disable a user interface element if a model property is a
nil value. Instead of writing a method that returns
YES if the property is
nil, you specify that the binding use a “is not nil” transformer. The transformer acts as the “middleman”, providing a
YES value to the user interface element if the property is
Value transformation is done immediately before a value is passed to a user interface element’s
setObjectValue: method. Likewise, the reverse transformation is applied before the value in the user interface is set in the model. See “Bindings Message Flow” in Cocoa Bindings Programming Topics for a detailed description of when value transformers are applied in the context of Cocoa bindings.
All value transformers are subclasses of
NSValueTransformer. In addition to providing the abstract methods for subclasses, the
NSValueTransformer class maintains a mapping of value transformer names and the corresponding value transformer objects. This name is used in Interface Builder to specify the value transformer that is used for a binding. You register instances of your custom value transformers in order to expose them, allowing them to be used by Cocoa bindings in Interface Builder.
A value transformer can be reversible, able to convert a value to a new value, and back again. A reversible transformer can be thought of as “read-write”, it transforms the original property value, but will also return any changes made to the transformed value. A non-reversible transformer is “read-only”, only able to reflect changes in the original property.