A Web Application Runtime Application Self-protection Scheme against Script Injection Attacks

Script injection vulnerabilities are popular vulnerabilities in dynamic web applications. Necessary conditions were analyzed for the generation and exploitation of script injection vulnerabilities to provide protection against different injection types. Combined with the analysis of the host language and the object language, the statements were located with their types in the HTML statements. Based on the control flow graph, the data dependency relation subgraph containing source points and sink points was built. A filter insertion algorithm is designed for this sub-graph to define different input data type filtering strategies. Then a solution was implemented based on data flow analysis and automatic insertion of filters before relevant sink statements.