Font Size: a A A

Research On XML Database With XQuery Support

Posted on:2003-01-04Degree:MasterType:Thesis
Country:ChinaCandidate:J YaoFull Text:PDF
GTID:2168360065460838Subject:Computer applications
Abstract/Summary:PDF Full Text Request
Because of the XML(Extensible Markup Language)'s capability of representingvedous kind of information, it will be the standard of exchanging information over thenext generation Internet. XML Data Model is hopeful to be a coordinate withRelational and Object-Oriented Data Model. One research focus in the databasecommunity is how can we integrae databases and XML to explore the theory andtechnology of XML-Database.In this thesis, on the basis of analyzing and understanding W3c's specifications ofXML and XQuery (XML Query), we mainly investigate on aspect of XML databases,that is storing and querying XML. First, a processing model of XQuery query engineis proposed frOm the research of XML query language. Second. the XML data modelsare stUdied in order to support XML-query Finally, we investigate and implement thestoring and querying of XML based on relational databases.In the research of XML query language, we consider the requirement of twocategories (for data-centric and for document-centric) of XML aPplications and pointoat that the basic operations in XML-quey are pattem matching, selection andfiltering, and result reconStfUction. The path expression and the FLWR expression inXQuery are tyPical strUCtUrs tha are comPrised of these operations. According toW3C, the execution of XML-query involves four intemal phases, i.e. Parsing,NormaliZaion, Static Type Checking, and Dynamic EvalUation. It also involves threeinPut/outPut phases namely Schema Irnport, XML Loading, and SerialiZaion.In Study of XML data model, we investigate the semi-StructUre charasteristics ofXML: selfdescribe, graPh-based model and irregUlar Then the compedson andanalysis of normal tree-strUcture model and node labeled direct graPh are Presented.We draw the conclusion tha the XQuery data model is a tree-like model with varioustyPes of nodes. The nodes are conStrUcted by consmictOr and accessed by accessor.Both of them are defined in XQuery data model.'It is our vieWPoint that XML database research will receive benefits from thewell-established relational database theory and technology especially in XMLdatabases' early stage. For this reason, in our studX we use relational databasesStoring and querying XML. The key problem of storing XML in relational databasesis how to design a set of table cleverly and rationally for storing XML. To solve thisproblem, the first method is tha we use XML's DTD (or Schema) to generate tabledefinitions for storing XMLSchema-based method. An altemaive way we use isthat we store both the data and the strUcture (especially path) of XML into severalpredefined tableMath-based method. We suPport XQuery by using a method basingon the data model, we call it DM-method. Using thes method, the physical store ofXML is hid from XQuery query engine and the suPPort for XQuery is achieved overrelational databases.In particular, we make two contributons. First we give processing model of XMLquery engine according to W3C XQuery language. Second, we implement tWomethods that can store XML in relational databases, and propose a method that canhide relational storing manner from XQuery engine.
Keywords/Search Tags:Research
PDF Full Text Request
Related items