Research And Application Of Structured Testing Method For Disjunctive Logic Program

Answer Set Programming(ASP),a logic programming based on answer set semantics,is one of the most widely used descriptive problem solving methods to achieve non-monotonic knowledge representation and reasoning,and has been widely used in many fields such as planning and diagnosis of artificial intelligence.How to improve the reliability(or ensure the correctness)of ASP programs is currently difficult for ASP programs in software engineering.Structured program(automatic)testing technique is an important method commonly used in the field of software engineering to improve the reliability of programs,and has been initially applied to the testing of normal logic programs.Since disjunctive logic programs are strictly more expressive than normal logic programs,it is important to promote the application of(automatic)testing methods such as structured testing in software engineering to the(automatic)testing of disjunctive logic programs in order to further explore and improve the reliability of disjunctive logic programs and to promote the engineering of answer set programming.The main work of this paper is as follows.(1)The basic theory of structured testing of disjunctive logic programs is proposed.Based on the theory of structured testing of normal logic programs,concepts of test coverage for disjunctive logic programs: subprogram coverage are defined,such as rule coverage,definition coverage,loop coverage,etc.,and discuss the basic properties of these coverages,which provide theoretical guidance for the selection of test cases for(automatic)structured testing of disjunctive logic programs.(2)A prototype system for automatic testing of disjunctive logic programs was implemented.Based on the theoretical work on the testing of disjunctive logic programs,a method is proposed to automatically generate test cases for disjunctive logic programs: a logic program is generated based on the logic program under test,and its answer set is a test case for the logic program under test;the experimental results of systematic automatic testing of five typical(disjunctive)logic programs show that the above-mentioned test coverage concept is effective for the selection of test cases.Which is to say that the test purpose can be achieved with less test cases without reducing the coverage rate,thus improving the efficiency of automated testing.