Design And Implementation Of SLIQA System

In the process of building software systems,software designers and developers usually directly use existing software architecture patterns to guide the development of software systems,and the hierarchical architecture pattern is one of the commonly adopted patterns.The hierarchical mode organizes the software system into a hierarchy,and the complexity of the system implementation process can be effectively reduced through hierarchical division,so the mode has good ease of understanding and ease of use in practical applications.By using the hierarchical mode to develop software systems,software development efficiency and software quality can be effectively improved,and it is also conducive to the maintenance and expansion of software systems.However,as software systems continue to evolve,the architectural state of the hierarchical architecture pattern may shift,resulting in inconsistencies with the design goals of earlier architectures,which in turn leads to problems that the software architecture is difficult to understand and maintain.Architectural drift may be caused by the frequent development actions that violate the principle of hierarchical architecture in the process of software system construction,and the lack of standard architecture design documents makes the understanding of different developers inconsistent.In order to avoid the structural deviation or mitigate the serious impact of the architectural deviation of the software system using the hierarchical architecture pattern,it is necessary to detect and restore the software architecture when the software evolves to a certain stage.The detection of the schema is to reflect the actual application of the hierarchical architecture pattern by obtaining the current schema state,so as to determine whether the current schema meets the expectations,and then find the key nodes that cause the schema to be offset according to the relevant design principles and constraints,so as to carry out refactoring and other maintenance activities.The main goal of this thesis is to improve the deficiencies in the existing identification methods according to the research and analysis of the existing hierarchical architecture pattern recognition methods,and form a complete set of hierarchical architecture pattern recognition methods.In addition,in order to make the pattern recognition process promote the subsequent recovery work,on the basis of the architecture information extraction and pattern recognition process,according to the characteristics in the hierarchical architecture pattern,the factors affecting the quality of the architecture in the pattern are identified and extracted,a simple and easy-to-understand model is established to represent and quantify the quality of the architecture,and finally the architectural quality change trend between different versions in the evolution of the software system is used to assist in the discovery of architecture deviation problems and problem analysis.The work carried out to achieve the above objectives is as follows:(1)A layered architecture pattern recognition method based on program dependency information extraction and processing is proposed.On the basis of using the source code-based architectural information extraction method to obtain the architectural pattern data,the preprocessing mechanism is added to improve the feasibility of identifying large-scale target systems,and then the entity structure similarity is proposed as the clustering basis to improve the accuracy of hierarchical schema pattern recognition in the improved schema recognition algorithm.(2)Firstly,according to the background of the formation of the hierarchical architecture pattern and the relevant research of the application case,the important characteristics of the hierarchical architecture pattern are extracted and used as the constraints guiding the architecture detection and recovery process.Then,the products of the pattern recognition process and other necessary information are used to establish the architecture quality factors and evaluate the model,and then analyze the measurement results based on different versions of the evolution process to reflect the evolution trend of the architecture and assist in locating the offset point.(3)The SLIQA(Software Layer Identification and Quality Assessment)system is designed and implemented.It has two main functions: layered architecture pattern recognition and layered architecture quality analysis.It can automatically carry out dependency extraction,architecture information analysis,layered pattern recognition,pattern violation analysis and quality evolution trend analysis for Java software.Several layered open-source software and their different versions are selected as test objects to verify the functions of the methods and tools in this thesis.The results show that the pattern recognition method has high accuracy and can effectively restore the reality of the target architectural state;the generated architectural abstract representation can effectively help understand the structure and state of the system,reducing the difficulty of maintenance work required in the evolution of the system;the architectural quality analysis method has a high degree of recognition for the trend of system evolution,it can detect the structure deviation phenomenon in the evolution process of the software system in time and designers and developers can quickly locate the deviation point to carry out subsequent reconstruction work through the analysis of relevant factors.