Toward Generating Optimization Actions Of Twig Pattern

XML is a markup language proposed by the World Wide Web Consortium (W3C),is widely used in the data interaction between the cross-platform. Twig pattern queryas a natural way of XML query, due to its central position in the XML query processhas been widespread concerned.In order to improve query efficiency, a lot of research was made on Twig patternoptimization. Optimization methods can be broadly divided into two categories, thefirst category is to improve the efficiency of Twig pattern matching algorithm usingauxiliary data structures and coding techniques, accelerate Twig pattern matching;second category is by modifying the query itself, that is, the so-called query rewritetechnology to optimize the Twig pattern, the common practice is for a given Twigmode, transforming Twig pattern into a price, but having a higher efficiency form, forexample tree mode is minimized.Twig pattern minimizing usually considers the constraints in the DTD, XMLSchema. In the current work, different constraints have different cutting methods,most of these cutting strategies are for specific XML query processing platform, it isdifficult for the same query expression performing a variety of optimizations, andwhen the discovery of new optimization method, the algorithm must be redesigned ormodified. The correctness of algorithm itself can only rely on designers to manuallycheck and correct. To solve this problem, this paper proposes a generic Twig patternoptimization program, the program does not generate new optimization methods andrules, but to present a unified optimization rules description language ORS and itsplatform. According to optimization rules described by ORS language as well aspending Twig pattern, the system will automatically output the cutting list of Twigpattern. The importance of the study is that: on the one hand, simplifing the process ofTwig pattern optimization and cutting; on the other hand, the formal methods andtheory are used in Twig pattern optimization, using model checking and predicatelogic theory and tools, as well as the strict definition of the syntax and semantics ofthe rule description language-ORS, ensursing the correctness of the resultingoptimization actions.