Software Source Code Similarity Analysis Technology And Application System

Computer software industry rapid development, the industry is more likely to have intellectual property disputes and plagiarism phenomenon, which is different from the traditional industry, as the source code is the core value of a computer software, to determine whether there is plagiarism, this is very important, which can be treat as the evidence for judicature. To detect source code whether there has plagiarism can help avoid some commercial disputes, as it can be used in detecting whether there is open source code in commercial software.Code Similarity Detection been widely used, which can be used to judge whether there is plagiarism in the source code and if it violates the rules of the use of open source code. The existing tools often use one comparison algorithm, and only aim at one language, the ability of comparison is very limited. This subject is designed and realized a kind of similarity analysis system, it uses the three core algorithms based on the character, tags sequence and grammar, in addition, it also use the whole file comparison and part of the semantic comparison as the assistant, and according to proportion of each algorithm we designed a very accurate comprehensive comparison algorithm.Similarity detection is a kind static analysis method, mainly used in source code similarity comparison. This system can detect the source code by line, and adopted many advanced and mature combined algorithms. Considering that the detection process is calling for an amount of code samples or database, which is composed by a large number of open source code. So we can build a large Sample with unsafe code, that suitable for detecting the security of other software. This paper constructed a large number of test sample, involving a variety of defect types, by comparing large number of source projects,we proved the superiority and reliability of the idea.