Sethi, Mridu
15.9 MB of textual records (PDF)
Audience: Undergraduate. -- Dissertation: Thesis (B. A.). -- Algoma University, 2007. -- Submitted in partial fulfillment of course requirements for COSC 4235. -- Includes figures. -- Contents: Thesis.
The importance of testing a source code is universally acknowledged. However there are more excuses prevalent to skip the process of testing with few common reasons being "My source code doesn't need testing" and a few like "I do not wish to waste time in testing; code development is what I need to work on". However, testing plays an imperative role in the construction phase of a software application or project. The concept of unit testing evokes bipolar reactions among people, either they are totally for it or absolutely deny its usefulness and strengths. The best way to begin would be to write some code, entrap the encountered bugs by testing them, mend them, and work on these iterations of writing, testing, fixing repeatedly until the desired output is achieved. It is important to note that testing reveals the presence of faults and thus its goal is to find faults, the more the better. Unit testing is the first and foremost level of testing. It is used to test the components in isolation, so as to uncover irregularities between the behavior and interface specification. It is widely recommended that developers devote 25-50% of their time to testing. In this thesis, I have worked towards developing an automated test generator which solves the purpose of generating appropriate tests which are specific and general at the same time, and helpful in catching bugs in a .class file of Java. Testing yields good returns during development by the way of adding new functionality and efficiency to the project at hand, by discovering bugs which prevent a software program to give the desired results.
