Object Owning Layers and Drawables¶

Object layers are functionally split into two classes - DKObjectOwnerLayer and DKObjectDrawingLayer, which subclasses it.

The names may be slightly misleading, as a DKObjectOwnerLayer is perfectly able to draw the objects it owns. The name is meant to reflect the main functional purpose of the class - that is, to own objects. In DrawKit, the term ‘object’ is rather ambiguous and open to misinterpretation, so a graphical object that has its own distinct identity is called a “drawable”. Every single shape, path or other distinct selectable item in DrawKit is a subclass of the semi-abstract drawable base class, DKDrawableObject.

DKObjectOwnerLayer provides the basic ownership of drawables, and is responsible for maintaining them as a related set. It also deals with the front- to-back (Z) ordering of the drawables it owns, and provides methods for changing these. Like DKLayerGroup, it provides user-action methods that can be simply hooked to menus to provide commands such as Move To Front, Move Backwards, etc.

DKObjectDrawingLayer is a subclass of DKObjectOwnerLayer that brings the concept of a selection into the picture. The reason for the functional split is twofold: first, one of convenience in that even with the split, both are quite large classes in terms of number of methods, so this keeps it manageable. Second, it permits a DrawKit developer to subclass at either level if they wish, giving a more fine-grained opportunity to do this. DKObjectOwnerLayer undoably supports the adding and removing of objects to the layer, changing their Z-order and other basic operations that do not require the concept of a “selection”. It also handles the essential requirements of handling drags of external data into the layer.