Xml-based Heterogeneous Database Integration Middleware Technology Research

Today, more and more enterprise applications need to integrate the different information from diverse database sources, and a system to integrate heterogeneous database sources dispersed over a computer network must be developed and several challenges must be overcome to establish such a system: heterogeneity, integrality, performance, semantic conflict, authorization bottleneck, constraints subjoining and content restriction. Aimed to this goal, an appropriate middleware (Heterogeneous Database Integration Middleware Based on XML, HDIMBX for short) framework based on XML is proposed in this paper. Successful data transformation from relational schema to XML schema and data integration makes the middleware work as a bottom level service to support data publishing, sharing and access. The following five issues are studied in this paper:1) Design of the middleware architecture. A Mediator/wrapper middleware architecture based on XML has been designed for heterogeneous database integration. It provides a bottom level service for application via schema transformation and business logic encapsulation.2) Implementation of the middleware architecture. The middleware architecture has been implemented based on the J2EE technology, which is one of the most important technologies in enterprise computing. EJB,JNDI,JTA,RMI,JDBC,XML and more other technologies are synthetically applied to get a good extensibility and a good flexibility.3) Presentation and implementation of Key-based Relational Schema Reverse Reconstruction Algorithm (KRR). Based on discrete mathematics theory, a KRR algorithm is proposed which can realize the transformation from relational schema to directed graph structure and get a good data preparation for the sequential schema mapping.4) Presentation and implementation of Constraints-preserving Mapping Algorithm from Relational Schema to XML Schema (CPM). A schema mapping templete is defined by the segmental modeling for XML schema and the relational schema will be transformed to XML schema according to this schema mapping templete.5) Presentation and implementation of a Constraints-Subjoining Integration algorithm (CSI). The results of transformation are integrated by separating constraints and containing constraints defined by CSI algorithm.