Parallel processing model for XML parsing
Adriana Georgieva, Bozhidar Georgiev · International Conference on Signal Processing · 2012
In this paper, are presented some development problems and solutions concerning the parallel implementation of an algebraic method for XML data processing. It is in tight connection with modern concepts of the parallel programming. The proposed parallel algorithm first partitions the XML document into chunks and then apply the parallel model to process each chunk of XML tree. In the article are shown some theoretical aspects of XML functional parsers and parallel navigating mechanisms on XML source. The authors suggest a different point of view about XML parsers with the creation of advanced algebraic processor (including all necessary software tools, search techniques and programming modules). The possibilities of this linear algebraic model, combined with principles of parallel programming allow efficient solutions for parsing, search and manipulation over semi-structured data with hierarchical structures. Thus presented paper combines the building of an algebraic formalism for navigation over XML hierarchy with concepts of modern XML parser and their mutual work in parallel. So proposed parallel parsing mechanism is easy accessible to the Web consumer, who is able to control XML file processing, to search different elements in it, to delete and to add a new XML content. The presented various tests show higher rapidity and low consumption of resources in comparison with some existing commercial XML parsers.