The Research On XML Query Technology

With the rapid development and popularization of Internet, people can receive and send a large number of the latest information from all parts of the world real-timely by linking their computer to Internet. But an outstanding problem exists while information exchanges, it is the varied data formats which have brought the obstacle to the using effectively of information. So how by a most convenient, reliable and effective way to obtain necessary information is a heavy puzzlement. XML as a kind of electronic data interchange standard that is used in WEB applications, with its remarkable advantages, especially in interchanging different structural data over platforms, is becoming the best choice for electronic data interchange through Internet in the future. It has been already widely used in various kinds of fields, such as commerce, education, communication, etc. The research on XML data storing and querying is also in the ascendant.At present, the research on querying XML data mainly concentrates on the query based on path expression because it is one of the core components of most XML query languages. From studying new query methods, index structures, operator recoding and path expression's optimizing, etc., researchers have proposed many valuable solutions to improve the query efficiency. For example: the path expression template-match method, the structural joins based on B+ tree method , the suffix tree index method, recoding path expression into a special index method and utilizing XML schema to optimize path expression method, etc.. All of these have made important contribution to raise XML query efficiency, but they also have their own shortcomings. In this paper, based on the deep studying of all these research results, we proposed a new solution to mostly improve the query efficiency on single XML document by simple path expression, from building a high-efficient index structure and choosing an appropriate realization mechanism. The main works and achievements we have doneas follows:1. Proposed a new index structureBased on the basic tree structure of XML file, we proposed a new index for XML document which utilize Hash index, linked list index and C-forest' structure synthetically and analyzed its efficiency in query.2. Proposed a new query processing methodIn this paper, we proposed a new solution which take advantage of the multithread mechanism, utilize the high-efficient index structure built in this paper, from the bottom of a path expression to the up, to process a XML query simultaneously. To support the multithread mechanisms, a path expression need to be rebuilt before processed. We described the parse tactics in detail also.3. Presented the key algorithms and their programming codesIn this paper, we presented the key algorithms and their programming codes on how to build the desired index structure and how to process a path query using multithread. Mainly include: the algorithm of parsing XML document to build a special index structure, the algorithm of parsing path expression to build the relevant objects and the algorithm of processing a query multithreading simultaneously. The technologies adopted among this paper also have some practical reference values for programmers.4. ExperimentWe compared the traditional method with our method on the query efficiency on some XML documents of different size in an experiment. Through analyzing the experimental result, we pointed out the remarkable improvement of query efficiency by our method and pointed out the problem still existing in it.