Research And Implementation Of Structural Join In XML Data Graph

With the development of computer science and web technology,XML is becoming the standard language for data interchange and presention,because of the strong features self-describing,cross-platform,expansibility.The most popular and effective method is use database to manage XML data.XML can not only represent tree-structure data,but also complex graph-structure data. The ID/IDREF property of XML makes a XML document to be of the graph structure rather than the tree structure, the traditional structural join algorithms are not suitable to process this kind of XML data graph. It is necessary to carry out a research on the structural join algorithm with respect to this ID/IDREF property, and also the corresponding XML query syntax should be extended.In this paper, we introduce the XML-related technologies, including XML model,the two main forms of XML query language XPath and XQuery,XML query method,XML index and other aspects of technology. By analyzing the results of existing technologies and query language features, we proposed an XML graph-structural query method design and the design has been implemented.First, the ID/IDREF relationship between nodes is called the referencing relationship in this paper, and a referencing index for this kind of relationship is organized with the B+ tree structure. After that, an XPath extension is presented based on the referencing index,and offer a desing whcich can transform an appropriate XQuery to extended XPath,after that, a novel structural join algorithm based on the index is proposed to support the query processing using the extended XPath,which can avoid some invalid result. Finally, the primary experiments on XML query efficiency of the extended XPath and XQuery are discussed, the experimental results show that our method gains better query efficiency.