Generation of Character Test Input Data Using GA for Functional Testing

Typically, test input data is manually created for the functional testing of software applications hence is a time consuming and error prone activity. In this paper, we present an approach to generate test input data from structured requirements specifications models such as UML use case activity diagrams (UCADs). We propose Constraint Representation Syntax (CRS) for framing software attribute properties as a part of structuring the Software Requirements Specifications (SRS). Then, structured models are parsed into a set of functional paths along with their predicates containing attribute constraints. Genetic algorithm is used to generate test input data that satisfy these predicates. Based on our approach a prototype tool has been developed and a case study results are evaluated.