An empirical evaluation of optimization parameters in XML parsing for performance enhancement

XML is a widely adopted technology in various domains ranging from web services to messaging systems, databases grid and scientific computing. However the processing of XML documents is an expensive operation that can degrade performance and limits the adoption of XML for specific applications. As different applications incorporating XML have come to the fore it becomes imperative to conduct a comprehensive study of the issues involved in design of parser. Hence optimizing parser performance has become a research hotspot. This paper focuses on a detailed study of the optimization parameters and design and implementation of a novel parser NXParse while optimizing for character streams, in-place parsing, string transformation control-flow, data structures and recurrent structures. Our empirical analysis shows that the NXParse gives a better parsing performance.