Research On Native XML Database And Implementation Of The Engine

With the rapid development of the Internet technique, extensible Markup Language (XML) is fast becoming the standard to conformity isomerous data. It support semi-structure data and can self-description, independent of platform. At the same time, there come forth so many XML data continually that how to store them efficiently has become a research hotspot. Many databases support XML data have appeared in a short time. There are old manufacturers of relation database and object-oriented database among them, and also many new-established database companies. Native XML databases, which store XML data in native format, perform much more efficient than other databases in web applications because of XML data's characteristic. And then it receives wider attention. But in such an early time of native XML database, there hasn't formed any fixed standard. More or less, there are many problems in storage structure, query optimize, index mechanism and concurrency control etc.This paper has an in-depth study, analysis and summarization on current native XML databases. And describe the implementation of the new main memory resident database engine called "myNXD", which is a native XML database too. The primary tasks of the paper are as follows:1. A new storage model to store XML data in native format is proposed, which can efficiently parse and store XML documents. The engine acquire related structure information after analyze the DTD and Schema definition of the XML documents. Use these metadata to parse XML documents and write the data into main memory in native format for farther query operation.2. Improve the current index techniques of native XML database, use simple and efficient indices, which called path index and value index, to lookup and orientation the XML node data and improve the query efficiency of the database.3. Support various query strategies, such as bottom-up, top-down, through different APIs to traverse the tree-like XML data. Combined with different index, myNXD engine can be used as a testbed to evaluate the performance of various query strategies when search XML data. MyNXD engine can take suitable query strategies to optimize the query base on different query conditions.4. Analyze the current problems of using lock mechanism when deal with concurrency transactions of XML data. Propose some simple rules for locking...