Subscribe by Email


Friday, January 6, 2012

What are different mutation testing methods?

Mutation testing is another methodology of software testing. It can be defined as the software testing methodology which involves modification of source code or programs or the byte code of the program in a way that is particularly small but significant.

Mutation is known by two other names:
- program mutation
- mutation analysis.

SOME IMPORTANT FACTS
- The mutated code is tested using various cases and the test suite that doesn’t detects and rejects the code that has been mutated is usually considered to be defective in nature.

- The required mutations are carried out on the basis of some well defined operators known as mutation operators.

- These mutation operators are well known to mimic typical programming bugs or errors and some times they also force the creation of tests which are significantly valuable tests.

- Best example of such test is given by the test which derives each expression to the value zero.

- The purpose of the mutation testing is basically to allow the tester or software developer to develop effective test cases so that the weaknesses or the vulnerabilities of the software system or application can be located or detected in the test data that is being used as the input for the program or in the sections of the program code.

- These sections are chosen as such that they are rarely used and accessed during the execution of the software system or the application.

- The mutation testing is carried out to ensure or verify that the implemented software program or application is correct or not.

Still the creation of the mutation testing poses some doubts.
- The doubt is whether the mutation tests are correct or not?
- Whether they sufficiently cover all the specifications and the requirements that have lead to the origination of the implementation of the test cases?

Mutation testing represents a technological problem of “who will guard the guards?”
- In the 19s, the mutation testing was pioneered to detect or locate the weaknesses and the potential vulnerabilities of the test suites.

- A theory was put forward regarding the behavior of the mutation testing.

- It stated that if a mutation was done without affecting the behavior of the program or output, then it meant that either the code that had been mutated did not execute or the employed testing suite wasn’t able to detect the injected mutation.

- In order to make one mutation work, one has to introduce many mutations all over the program.

- This in turn leads to the compilation of a large number of copies of the program code.

- The increase in the use of unit testing frame works and object oriented programming languages has led to the increase in the creation new mutation testing tools for other programming languages.

METHODS FOR MUTATION TESTING
- Mutation testing is carried out by selecting few mutation operators and then applying them to the source code using one at a time for each correct piece of the program code.

- The resultant of application of a mutation operator is known as mutant.

- If the mutant is detected during the testing, then the mutant is said to be killed.
- Strong mutation testing or strong mutation coverage ensures that the test suite is effective in detecting the mutants.
- Weak mutation coverage or weak mutation testing can be related to code coverage methods.
- It takes less number of calculations to ensure the effectiveness of the weak mutation testing.
- In some cases equivalent mutants are also used.
- In those cases it becomes impossible for a test suite to kill such a mutant.

- Equivalent mutation proves to be the greatest obstacle in the path of practical usage of the mutants.
- Equivalent mutants require more efforts for testing.
- A variety of mutation operators have been discovered by the researchers.


No comments:

Facebook activity