Software Evolution Workflow Modeling And Analysis Based On Graph Transformation

The study of the software evolution is significant for the software lifecycle extension and improvement of software adaptation to the new environment and reducing operating costs of software. However, the existing software evolution technology also has a series of problems, such as how to pre-derive the results of change and how to design a flexible planning or mechanism of software evolution. In order to fix these problems, this thesis introduces workflow into the field of software evolution. Thus, the evolution operations can be combinated to deal with more complex evolution. And through the analysis of the evolution workflow, we can pre-derive the results of software evolution or optimize the software evolution workflow.Based on graph transformation, this thesis describes the workflow model with time and resource attributes by attributed type graph (ATG) and designs a set of graph transformation rules for workflow modeling.In order to improve the applicability of the workflow model, this thesis introduces an approach of workflow model correctness analysis using graph reduction for making sure the correctness of workflow model structurally. This thesis defines a set of rules to deal with the workflow time consistancy analysis and workflow deadline analysis and introduces an approach based on the conclusion of the existing research on the field of workflow schedu(?)ability analysis.Finally, the thesis realizes the software evolution workflow modeling and analysis tool, which is used to to finish the workflow modeling on the software evolution of Conditional Accept System (CAS). The thesis uses this tool to finish the correctness analysis, deadline analysis, time consistancy analysis and schedu(?)ability analysis. It is real effective which is proved by the result of case study that the approach of software evolution workflow modeling and analysis.