Dsign And Implementation Of Detection System Based On Source Structure Similarity

The program language course is a very important in computer science,and it is important to strengthen the students’ programming ability in teaching.Code similarity automatically detecting can effectively help teachers to view plagiarism of programming.This paper analyzes the similarity detection which is based on the code structure information.The main research contents are as follows:(1)Preproccess the source code,and remove irrelevant information which has no effect on the content of the code structure.This process makes the source code greatly reduced,and it is relatively simple for the subsequent analysis of the grammar to prepare.(2)The main content of the source code similarity detection system using structural metrics is to express the source code as a feature string containing structural information.This paper analyzes the grammatical semantics from the perspective of grammar,and divides the main grammatical unit by c language grammar,and the main grammatical unit embodies the structural characteristics of the program.And achieve the source code structure information with the help of YACC parsing technology,generate file which is contains the structural information of the source code.(3)According to the obtained file,and use similarity matching algorithm to get the similarity measure of the program code which is to be compared.According to the calculated similarity measure and the criteria to determine whether the plagiarism of the phenomenon,and come to the final result.The system will give several different levels of similarity assessment criteria to improve correctness.