Software Migration Based On Refactoring Partial-Order Planning

Software migration is one of the most important means of expanding the usage scope of software and extending software life-cycle. But most existing migration methods rely on the experience of developers. It is difficult to keep system consistency between the new system and the original one.This thesis introduces the software refactoring method into software migration by using the behavior keeping feature of refactoring to maintain the migration quality in this field, and simultaneously to adapt complicated system's migration. This thesis applies the intelligent planning theory into software refactoring to instruct the refactoring operation choice. It causes the work of software migration to be partial automated, while it could avoid the blindness of operation choosing in refactoring, and then it raises the efficiency in software migration.This thesis first introduces the partial order planning method in detail, then addresses the related principle and the technology of the Scheme Interpreter under the Eclipse IDE as a plug-in, thus it realizes the partial order plan under the software development platform. In order to ensure software architecture described visualization, this thesis presents a software architecture description method based on the Attributed Type Graph theory, and transforms software structure graph description into predicate logic description by GXL language for achieving planning. Further, this thesis applies the partial order plan method into software refactoring, proposes a software refactoring planning method based on the partial ordering planning, which could be useful for the solution of software refactoring operation choice question.Finally, this thesis takes a concrete software migration project as an example, analyses the partial ordering plan software refactoring method in detail, and proves that this method is effective in software migration.