Abstract - lic thesis Robert Kaminski
Towards and XML Document Restructuring Framework
An XML document has a set of constraints associated, such as validity
constraints and cross-dependencies. When changing its structure these
constraints must be maintained. In some cases a restructuring involves
many dependent documents; such changes should be automated to ensure
consistency and efficiency.
Most existing XML tools support simple updates, restricted to a single
document. Moreover, these tools often do not support concepts defined by
a specific XML-application (an XML-application defines the set of valid
markup symbols, e.g., tags, and their hierarchical structure). This work
aims at developing a framework for XML document restructuring. The
framework facilitates realisation of document restructuring tools by
providing advanced restructuring functions, i.e., provide an environment
where restructuring operations can easily be realised. To avoid
restricting the framework to a specific set of XML-applications, it is
designed for flexibility.
The conceptual part of this work focuses on the definition of an
operation set for XML document restructuring, called the operation
catalogue. The operations are adapted to a document model defined by
this work. The catalogue is divided into three abstraction layers,
corresponding to the concepts defined by XML, XML-applications, and
XML-application policies. The layer structure facilitates extensibility
by allowing new operations to be defined in terms of existing.
In the practical part, an architecture is presented for a document
restructuring framework which supports realisation of the earlier
presented operations. The architecture is based on a layered approach to
facilitate extensibility with new layers that contain restructuring
operations and functions for an XML-application or an XML-application
policy. A new layer component can be added without recompilation of
existing components. To reduce resource consumption during document load
and restructuring the framework allows its user to specify, upon
initialization, the set of active layer components (each layer component
may perform analysis). This part also includes a prototype
implementation of the presented architecture.
This work results in an event-based framework for document restructuring
that is extensible with restructuring support for new XML-application
and XML-application policies. The framework is also well suited to
manage inter document issues, such as dependencies.
This work has been supported by CUGS (National Graduate School in
Computer Science), EasyComp (Easy Composition in Future Generation
Component Systems), REWERSE (Reasoning on the Web with Rules and
Semantics), and RISE (Integrational Software Engineering).