A Study On The Impact Of Python Coding Style On Software Maintenance

Coding style is a set of rules or guidelines used when writing source code for a computer program,bringing together the best practices that have been recognized in the industry.Code written in violation of coding style is not conducive to the understanding of developers and maintainers about its functionality,resulting in enhancing the possibility of introducing errors,affecting the readability and maintainability of program code,thus reducing the quality of software products.The relationship between coding style and software maintenance is worthy of further study,but there is no related work to study the impact of Python coding style on software maintenance.In order to explore the relationship between Python coding style and software maintenance reasonably and scientifically,this paper uses empirical research methods to select seven popular Python projects with 4-8 years of maintenance time on open source websites,then analyzes in detail the relationship between coding style and module change-proneness(fault-proneness)in 57 project versions.In the process of problem analysis,this paper uses Fisher exact test,Wilcoxon rank sum test and Logistic regression model to explore:whether modules violating coding style are more change-prone(fault-prone),whether modules with a larger violation density of coding style are more change-prone(fault-prone),and whether particular categories of coding style are more relevant to change-proneness(fault-proneness)than others.The results show that:(1)modules violating coding style are more likely to change than modules that do not violate coding style;(2)modules with a bigger violation density of coding style are easier to change than modules with a smaller violation density of coding style;(3)in some of the selected projects,modules violating certain types of coding style are easier to change than modules violating other types of coding style;(4)there are similar conclusions about the relationship between the violation of coding style and module fault-proneness.In a word,this innovative research work is of great significance,both the researches who focus on coding style and the researchers who study software maintenance can get inspiration and reference from experimental results.Based on the experimental data,we suggest that software developers should increase the level of attention to coding style,because the violation of coding style will reduce the readability and comprehensibility of the code,thereby increasing the cost of software maintenance;we also suggest that software quality maintainers should pay more attention to the modules violating coding style,thus they could find and solve more problems in finite time and energy.