A framework for effective exception handling in software requirements phase

Abstract

The exception handling structure allows software developers to reduce software maintenance cost through preventing faults, errors, and failures that may occur after exception arising. Forecasting possible exceptions and presenting powerful exception handling structures are noticeable in decreasing software modification workload and maintenance costs. But most of the developers neglect proper exception handling (EH) in early software development phases which make difficulty in software maintenance, indeed they underestimate EH. Since Focusing on EH only in the last phases of software life cycle is not a good policy, so we propose a framework, including principles, components, and metrics to present EH at the software requirement phase just while system scenarios are written. The proposed framework is a means for early exception discovery and leads to improve in software metrics: reliability, robustness, and maintainability. Applying the components of the framework: exception classification, scenario dependency graph and etc. and measuring proposed metrics in relation with exceptions allows to select proper EH strategies. At the end of the research, we present guidelines for the software tester to test all parts of the software according to the framework as a facility for verifying and correcting EH structures and discovering new possible exceptions.