Research On Method Of Detecting Unary Polynomial Inequality Likely Invariant

Computer software although experienced several decades developments, it has already got an extensive application in each social aspect. For example, medical and health, aviation aerospace, nuclear power plant control etc... At the same time, quality and safety of software got people extensive concern. People start studying how to raise the quality of software, and have obtained certain result. Design by contract is a very important widely applied method of building quality software. Contracts describe properties, expectations and guarantees of a program. Contracts can be classified into four levels, grammar contract, data contract, behavior contract and service contract. Program invariant is a common kind of contracts which includes class invariants, pre-condition and post-condition. Invariant is the basic fabric of data contract, which is valuable for program revolution, verification, rebuilding, testing, debugging and updating. Pass analysis the invariant at the key point of program can examine the run-time abnormality of program. The technique of dynamical detecting program invariants is a kind of effective method that solves contract formalization establishment.This paper focuses on dynamic detecting techniques in discovering unary polynomial inequality likely invariant. At first we introduce the concept of design based on the contract. After then, we introduce the technology of dynamic detecting program invariant and a traditional approach (Daikon). At the same time, we analysis the defects of Daikon and propose the method of detecting unary polynomial likely invariants. By using a test, we prove the validity of this method. We use the result of test comprising with Daikon, and finally we find ours method that is superior to Daikon. At the last of this paper, we design and enable a tool of detecting unary polynomial likely program invariant,which is very simple , flexible. The customer can define the variable which is observed by them.