Research On Performance Optimization Of Code Testing System

Unit testing is the first stage of software testing. Sufficient and effective unit testing can find a lot of software failures and reduce the cost during software developing period. The efficiency of manual testing is low and many mistakes happen during testing through this method. Code Testing System (CTS), which implements automated unit testing, can make up for these defects of manual testing. It can also provide reliable results of sufficient unit testing to the user. Good time performance can enhance the competitiveness of CTS among similar software testing products. So it is significant to optimize the performance of CTS.The main performance issues existing in the original CTS are as follows:1) The consumption of time for completing the feature that extracts functions from all files in the project is especially high;2) The consumption of time for generating project tree is also especially high;3) The editor of CTS doesn’t have the feature of syntax coloring;4) The interface of CTS is not friendly enough, and the time required for the user to understand the information on the interface is a bit long.This paper has done deep research on performance optimization of CTS, and proposed some optimization methods to solve those four performance problems existing in the original CTS.1) Dynamic file cache technology was proposed to optimize the performance of the process that extracts functions from all files in the project;2) Lazy loading technology and a mechanism to reduce the number of I/O operations were proposed to optimize the process of generating project tree;3) Syntax coloring technology was proposed to optimize the editor of CTS;4) Optimization to the interface of CTS has been done according to the principles of interface design. Dynamic file cache consists of file cache and cache estimation model. Cache estimation model can not only make the file cache of CTS have the ability to change its size, but also make the size of file cache be as large as possible to minimize the number of I/O operations.Experimental results show that all performance optimization strategies proposed in this paper can effectively improve the time performance of CTS. Dynamic file cache technology can effectively reduce the consumption of time for extracting functions from all files in the project. Project tree optimization technology can minimize the consumption of time for generating project tree. Syntax coloring technology can reduce the time consumption of compiling the code for it can let user find the keyword written wrong error as soon as possible. The optimized interface of CTS can reduce the time required for the user to understand the information on the interface for it is much friendlier.