Twig Query Optimization Technology Based On Partial-evaluation

With the rapid development of Internet technology, more and more web applications use XML language to exchange and store data, XML has become one of the current mainstream data forms. In order to standardize XML query and processing, W3 C presents XQuery language according to the data characteristic of XML language.The XQuery language uses XPath as a sublanguage to orientate and select elements, XPath provides navigation operation for XML document.But the semistructured characteristics of XML data makes its query efficiency lower relatively compared with relational data, and XQuery has a more complex semantic and implementation process as a programming language and a combination of query language. To this end, many XQuery queries can use Tree pattern to model, known as the Tree Pattern Query(TPQ), which tree structure with branching node is called Twig Pattern Query. Twig query is the core of the XML data query processing operation, its performance directly affects the XML data query processing efficiency.In recent years, Twig query matching algorithm has been widely attention and indepth research, TreeMatch algorithm is widely considered to be one of the best Twig query algorithms. According to the characteristics of TreeMatch algorithm re-peatedly analysis on Twig pattern, partial evaluation in the field of compilation tech-nology will be applied to the Twig pattern matching, through some partial evaluations were finished ahead of schedule on Twig model analysis, to generate the query that special instruction sequence replace the original query, and provide special execution engine of query machine, can effectively eliminate the repeated calculation, so as to realize the optimization of the XML tree pattern queries.Comparative experiments show that the optimization technology improve the execution efficiency of XML query to different extent for different Twig models.