An Abstractor provides the user with a view on a topicmap from the perspective of a topicmap elememt.

The result of that abstraction is returned as an AbstractModel.

How information is organised by an abstractor, is up to the concrete implementation. An Abstractor may perform such different operations on data as

.

In order to customize the abstraction process, an Abstractor accepts an instance of AbstractionContext as a parameter to the generateModel(TopicMapObject, AbstractionContext)-method.

An abstractor that shall be used with panckoucke, must implement the Interface org.tm4j.panckoucke.abstraction.InstantiableAbstractor.

The main method to implement is

  AModel generateModel(TopicMapObject, AbstractionContext)

This method returns an abstract model for the given TopicMapObject and the given AbstractionContext

Currently Panckoucke comes with five different abstractors

Compact- and Extended Abstractor both create an AbstractModel that represents the environment of the topic map object used as input. Both models are suitable to be rendered as a graph. The concrete result depends on the kind of the input topic map object as well as on the given abstraction context. See the description of both abstractors for further details.

The Section-Abstractor was introduced with panckoucke 0.3. Just as the Compact- and the Extended Abstractor it represents the enviroment of a given topic. In difference to the latter two, the resulting model is split up into several sections that consequently uses the MatrixPattern. This makes this Abstractor especially suitable to be used with more strictly organized, static output technologies like html for example (in contrary to an interactive hypergraph)

The Tolog-Abstractor takes a tolog query as input and returns result in a simple model, that follows the MatrixPattern.

An Index-Abstractor is constructed for a particular TM4J-Index. Once created, it takes a TopicMapObject as input and returns a simple model, that is suitable to be displayed as a list.