Research On Synchronization Evolution Method For Online And Off - Line Software Components

Software components such as well-encapsulated functional components and software frameworks has made the coders capable of invoking the exposed interfaces directly, which enables them to organize the business logic without having to be familiar with the related details. Software components can be divided into offline and online ones. The traditional offline components need to be downloaded to the same place as the applications that depend on them so as to be compiled or deployed. On the other hand, the online ones realize online binding and visiting through network without having to download the components to the local file system.However, due to the issues of defect repair and quality enhancements, the reused software components are always in change. The application program (namely client-side program) relying on a specified version needs to evolve in sync so as to be well adapted to the new version. This kind of synchronization has connections with a large quantity of software components, thus calling for huge amounts of modification workloads. Consequently, this paper will lay emphasis on the offline and online components, coming up with automatic and semi-automatic application level synchronized evolution methods, which help mitigate the influence on the applications brought by the constantly evolving software components.As to the offline components’evolution issues, this paper chooses the component evolution that exists in the form of Java language library and researches the relevant client side synchronized evolutionary issues. We are going to implement a code generator which automatically generates adaptation layer code according to the specified client side code. On that basis, the developers need only make a few modifications to the client side code to make it compatible with the new versioned components.When it comes to the online components, this paper focuses on the Web Service technology that supports cross-language invocation. Web service, being a kind of networking application, publishes and manages its computing service on the network through XML. For that reason, we are going to include an intermediate layer monitor responsible for the modification of SOAP messages and redirections of the service while the client side can finish the adaptation process by simply replacing the Web Service binding address with the monitor’s listening address. This paper demonstrates the feasibility of the two methods by two case studies. Finally, we are going to point out the offline and online components’commonalities and difference in the evolution patterns and the corresponding adaptation methods in the summary part and give out the reason that causes it.