In order to manage XML documents, native XML databases use specific encodings that map the hierarchical structure of a document to a flat representation. Several encodings have been proposed that differ in terms of their support for certain query workloads. While some encodings are optimized for query processing, others focus on data manipulation. For example, the Pre/Dist/Size XML encoding has been designed to support queries over all XPath axes efficiently, but processing atomic updates in XML documents can be costly. In this paper, we present a technique, so-called structural bulk updates, that works in concert with the XQuery Update Facility to support efficient updates on the Pre/Dist/Size encoding. We demonstrate the benefits of our technique in a detailed performance evaluation based on the XMark benchmark.

A key difference between traditional humanities research and the emerging field of digital humanities is that the latter aims to complement qualitative methods with quantitative data. In linguistics, this means the use of large corpora of text, which are usually annotated automatically using natural language processing tools. However, these tools do not exist for historical texts, so scholars have to work with unannotated data. We have developed a system for systematic, iterative exploration and annotation of historical text corpora, which relies on an XML database (BaseX) and in particular on the Full Text and Update facilities of XQuery.

Mere storage of personal data in state-of-the-art filesystems is a markedly well done job in current operating systems. Convenient access to and information retrieval from such data, however, is crucial to leverage the stored information. Thereby database style query languages can be of great use. We demonstrate a user level filesystem implementation that is built on recent semi-structured database storage techniques. As such, it serves as a storage layer for the BaseX XQuery processor and, while it appears to the operating system as a conventional filesystem, a large part of its content can be queried using XPath/XQuery.