XML-based documents play a major role in modern information architectures and their corresponding work-flows. In this context, the ability to identify and represent differences between two versions of a document is essential. A second important aspect is the merging of document versions, which becomes crucial in parallel editing processes. Many different approaches exist that meet these challenges. Most rely on operational transformation or document annotation. In both approaches, the operations leading to changes are tracked, which requires corresponding editing applications. In the context of software development, however, a state-based approach is common. Here, document versions are compared and merged using external tools, called diff and patch. This allows users for freely editing documents without being tightened to special tools. Approaches exist that are able to compare XML documents. A corresponding merge capability is still not available. In this thesis, I present a comprehensive framework that allows for comparing and merging of XML documents using a state-based approach. Its design is based on an analysis of XML documents and their modification patterns. The heart of the framework is a context-oriented delta model. I present a diff algorithm that appears to be highly efficient in terms of speed and delta quality. The patch algorithm is able to merge document versions efficiently and reliably. The efficiency and the reliability of my approach are verified using a competitive test scenario.    «

XML-based documents play a major role in modern information architectures and their corresponding work-flows. In this context, the ability to identify and represent differences between two versions of a document is essential. A second important aspect is the merging of document versions, which becomes crucial in parallel editing processes. Many different approaches exist that meet these challenges. Most rely on operational transformation or document annotation. In both approaches, the operations...    »