The Optimization Of XQuery Implementation Based On Indexing

With widespread application of XML for exchanging and expressing data on the Web, efficient access to XML data has become a key problem need to be solved, which touch on the research on the optimization of XQuery,a XML query language.There are many ways to optimize XQuery:by undertaking strategy,physical optimizing,algebra optimizing and so on.Undertaking strategy:by means of selecting traveling node method, exp, bottom-up or top-down, on the basis of path condition. Physical optimization:raising integrate rate by employing fast-cache to cashion the difference between CPU operating and memory visiting.Algebra optimization: optimizing path expression on the basis of XML data model and traditional logic operator, to achieve the purpose of optimizing XQuery.Currently there are many ways to optimize XQuery processing on the basis of index,exp, Dataguide in Lore system,XISS,DBXI.In Dataguide, special support on operator"//", which turn up in XQuery,was never offered.And path,which depart from root node,was record only. So it deprivate many probabilities of optimization.In XISS system,more spaces must be searched in XML data tree ,involved path expression. The frequency of dealing with fabric connection operating in XML-structured data at lease has N-1,so it aggrandized the querying complication and effect querying efficiency.DBXI is a XML index method on the basis of DTD. Searching spaces may be very large while circles are existed in DTD.This paper first analyze these optimization methods,and point out their advantages and faults. We expound a index mechanism based on R tree. The problem of fast-speed judging ancestor, descendant relationship between nodes in XML-structured data is resolved. We store XML-structured data by R tree and index the nodes in the tree,and converse the judgement of relationship between nodes to containing relationship while querying.At the same time,querying efficiency was...