Unit Testing of Software - Essay Example

Unit Testing of Software Unit Testing Unit testing is a phase of software testing development that tests the individual components of a system or software. The purpose of unit testing is to authenticate that the performance of each software unit is as designed (Craig & Jaskiel, 2002). A unit is the smallest testable part of software that normally has one or a few inputs and a single output. A unit may be an individual function, program, or procedure in procedural programming. In object-oriented programming, a unit is a method that may belong to a derived class, base class or abstract class (Perry, 2006). Some experts treat modules of applications as units. However, this is inappropriate, as there would be many individual units within the module. Drivers, mock objects, stubs, and unit testing frameworks help in unit testing (Craig & Jaskiel, 2002). Benefits of Unit Testing Unit testing increases confidence in maintaining or changing code. Good unit tests also increase the detectability of defects that may arise when changing codes. The development of unit testing is faster. The test runs take very less time as there is no need of firing the GUI and providing all the inputs. Additionally, fixing defects detected during unit testing costs lesser than in higher levels of testing (Craig & Jaskiel, 2002). Debugging at this level is easy since only the latest changes need to be debugged contrary to higher levels where changes made over the span of several days or weeks may need to be debugged (Software Testing). Limitations of Unit Testing There are limits to the number of scenarios and test data that a developer can use to verify the source code. This means that after exhausting all options the developer has no choice but to merge the code segment with other units (Software Testing). Integration Testing Integration testing involves testing combined parts of an application to determine if they function correctly together. Integration testing is performed after unit testing and before system testing. Integration testing is necessary to identify faults in the interaction between integrated units. Test stubs and test drivers re useful in performing integration testing (Tutorials Point). Analogy When manufacturing a ballpoint pen, the body, the cap, the clip and tail, the ballpoint, and the ink cartridge are unit tested separately. When the units are ready, they are assembled and integration testing is performed. For instance, one would test whether the cap fits into the body (Tutorials Point). Method White box testing, black box testing, and gray box testing can be used to perform integration testing. Generally, the method depends on one’s definition of unit (Craig & Jaskiel, 2002). Approaches to Integration Testing Big Bang Big bang approach involves combining all or most of the units together and testing them together. This approach is applicable when the testing team receives the entire software in a bundle. Unlike system testing which tests the entire system, big bang integration tests only the interactions between units (Computer Testing). Top Down  Top down integration testing approach involves testing top-level units first before systematically testing of lower level units (Perry, 2006). This approach is applicable in cases where developers follow top down development approach. Test Stubs are essential to simulate lower level units (Computer Testing). Bottom Up Bottom up approach to integration testing involves testing bottom level units first before systematically testing of upper level units. This approach is suitable when following bottom up development approach (Perry, 2006). Test drivers are needed to simulate higher-level units (Computer Testing). Sandwich/Hybrid Approach The sandwich approach to integration testing involves combining bottom up and top down approaches (Computer Testing). In a comprehensive software development environment, bottom-up testing is usually done before top-down testing. The process ends with several tests of the complete application (Tutorials Point). System Testing Once all the components of a system are integrated, a test on the application is performed to verify if it meets quality standards. Special teams of experts perform system testing. System testing is performed after integration testing and before acceptance testing (Craig & Jaskiel, 2002). Analogy During the process of manufacturing a ballpoint pen, the body, the cap, the ballpoint, and the ink cartridge are produced and unit tested separately. Integration testing is performed when two or more units are readily assembled. System testing is performed after the complete pen is integrated. Normally, black box testing method is used to perform system testing (Tutorials Point). Importance of System Testing System testing is the first step in the software development life cycle that involves testing the application as a whole. The application is tested systematically to prove that it meets the technical and functional and specifications (Software Testing). Normally, the application is tested in an environment that is close to the production environment where the application will be set up. System testing helps experts in testing and validating the business requirements and the applications architecture (Perry, 2006). Acceptance Testing Acceptance testing is arguably the most important type of testing as it involves gauging whether the application meets the intended conditions and meets customers’ expectations. Acceptance testing is performed after system testing and before making the system available to end-users. The quality assurance team uses a set of test cases and pre-written scenarios to test the application. More tests are performed and ideas about the application shared to gauge its precision and justify the initiation of the project (Craig & Jaskiel, 2002). Acceptance tests are not only identifies interface gaps, cosmetic errors, and simple spelling mistakes but it also points out any bugs in the application that may result in errors in the applications or lead to system crashers. Acceptance testing helps in deducing how an application will perform in production. Moreover, there are contractual and legal requirements for acceptance of the system (Tutorials Point). Analogy During the process of manufacturing a ballpoint pen, the body, the cap, the ballpoint, and the ink cartridge are produced and unit tested separately. Integration testing is performed when two or more units are readily assembled. System testing is performed after the complete pen is integrated. Normally, black box testing method is used to perform system testing. Acceptance testing is performed on the completion of system testing to verify that the ballpoint pen is ready to be made available to the end-users (Tutorials Point). Normally, black box testing method is used in Acceptance testing, which may not follow a strict procedure and is ad-hoc rather than scripted (Perry, 2006). References Craig, R., & Jaskiel, S. (2002). Systematic Software Testing. Norwood: Artech House. Top of Form Bottom of Form Perry, W. (2006). Effective methods for software testing (3rd Ed.). Indianapolis, IN: Wiley. Top of Form Bottom of Form Software Testing Fundamentals. (n.d.). Retrieved December 29, 2014, from http://softwaretestingfundamentals.com Top of Form Bottom of Form Tutorials Point Simply Easy Learning. (n.d.). Retrieved December 29, 2014, from http://www.tutorialspoint.com/software_testing/levels_of_testing.htm Read More