Design And Implementation Of The Analysis Tool For The Source Code In The J2EE Legacy System

This thesis analyzes a plenty of cases about the current status of Legacy System inthe software area, and finds that a lot of companies and organizations still keep usingthe old systems in their daily work, even though the systems have many problems ontheir quality and technology. Eventually, the thesis concludes the reason why the usersgive up developing a new system to displace the Legacy System is that the normalbusiness will be impacted by cost, timing and other unknown risks when thereplacement happens. Thus, when the concept about Legacy System Reengineeringappears, it not only satisfies the requests from the users, but also upgrades the oldsystem on technology; additionally, reengineering can fix the bugs in the old systemvia re-analyze, re-design, re-program and regression testing, in order that the newsystem can be more stable, safe, compatible, and extendable.This thesis also finds that reverse engineering spends lots of time and has toomuch risk during the process of Legacy System Reengineering; whereas analyzing thesource code holds the key of it. So far there is not a workable and automatic analysistool which can help the engineers to understand the complex source code and thebusiness logic in Legacy System. Quite many artificial subjective factors impact thisprocess and countless software defects will inject into the new system. As the result ofthat the new system and the original system cannot keep consistent anymore;additionally, it also causes the problems of software quality in the new system.This thesis focus on such a problem, then designs and implements an automaticsource code analysis tool in a realistic J2EE Legacy System Reengineering project.This tool bases on the contact between classes and the relationship between objects anddata tables, then sort the source code files, and create the diagrams of classes and thereflection diagrams for the analysts. Thereby this tool can improve the work efficiency,reduce the risks, ensure the correctness of the products, and also cut down the scheduleof the development. However, this automatic tool depends closely on a particular Legacy System, andit cannot cover all the cases during the process of analysis. So about the futuredevelopment, we need to put it into more realistic Legacy System Reengineeringprojects and find out the common points in those projects, then we are going tore-design and re-program the tool in order to improve its independence and coverage.