The Concept of Transformation of XML Documents into Quasi-Relational Model

Abstract

XML has evolved from a document markup language to a data model commonly and widely used for storing, exchanging and sharing hierarchically (semi)structured data. Simultaneously relational model is still a production standard for storing and querying data. Since there are on the market so many advanced and mature, relational- and sql-based tools, systems and solutions just loading xml documents into relational model (for instance as BLOBs) would be the easiest way for storing and then processing XML data. Unfortunately, in such a way, hierarchical and (semi)structured data becomes flat and unstructured. That is why so many effort is made to provide efficient models and tools for storing and processing XML data and such solutions as XPath or XQuery have been proposed and become a standard in this field. It results however in developing heterogenous and not unified data storing and processing standards, models and tools. It is similar to well-known object-relational impedance mismatch. Taking its power, scientific foundations, available tools and flexibility as well as, "assumed and hidden knowledge"(everybody knows it) it would be great to process XML documents effectively with the use of SQL queries. There are of course tools responsible for mapping between XML and relational model on the market. Using them however results often that either hierarchical data structure is lost and/or manual transformation (or at least transformation definition) is required. In this paper the concept of transformation of XML data into quasi-relational model is proposed. The general assumption of this transformation is making it possible to process XML document with the use of SQL-like language without making manual transformation and with preserving hierarchical structure of XML data. © Springer International Publishing Switzerland 2014.