Automated verification and test case generation for input validation

Input validation is essential for any software that deals with input from its external environment. It forms a major part of such software that has intensive interaction with its environment. Through the integration of invariant and empirical properties for implementing input validation, this paper proposes a novel approach for the automation of the following tasks from processing the source code of a program: (1) verification of existence of input validation; (2) generation of test cases to test and demonstrate all the input validations; (3) classification of each validation into the various types defined along with its test case generated. All the empirical properties in the theory have been validated statistically based on open source systems. Our evaluation shows that the proposed approach can help in both testing of input validation features and verifying the adequacy of input control.