Effective spectrum-based technique for software fault finding

  • PDF / 1,166,746 Bytes
  • 6 Pages / 595.276 x 790.866 pts Page_size
  • 66 Downloads / 221 Views

DOWNLOAD

REPORT


ORIGINAL RESEARCH

Effective spectrum-based technique for software fault finding Debolina Ghosh1



Jagannath Singh1

Received: 7 November 2018 / Accepted: 31 July 2019 Ó Bharati Vidyapeeth’s Institute of Computer Applications and Management 2019

Abstract The process of software testing can be divided into two phases, first is the testing of software where we search for bugs in the software and second is the software debugging which helps to find the root cause of the bug. Finding and correcting the bug takes much more time and effort than coding. In this paper, we present a study of different existing debugging techniques and their classifications. We found that most of the debugging techniques are manual, but the spectrum-based debugging technique is automatic. The spectrum-based software debugging is based on statistical formulas. Hence, we have presented the implementation and comparison of the six widely used statistical formulas and developed a debugging tool for JAVA programs. Keywords Debugging  Software testing  Java  Spectrum-based debugging  Omniscient debugging

1 Introduction There is a demand of high quality software and the quality of the software ensured by proper testing of the software. Testing of complex software is very effortful and time consuming because software testing consists of two phases. In the first phase, test cases are generated and passed to the software under test [1]. In this phase, we can know whether the software is having error or not. Many automated & Debolina Ghosh [email protected] Jagannath Singh [email protected] 1

School of Computer Engineering, KIIT Deemed to be University, Bhubaneswar, India

techniques are present to check the software is working as expected or not. In the second phase, i.e. software debugging the faulty software is carried out and the root cause of the error is finding out and rectified. Few work has been done to develop automated software debugging tools [3]. Hence, there is a need of developing automated tools for software debugging. In this paper, we review some of the existing software debugging techniques and find that most of the existing techniques are manual. In this paper, we present a brief survey of available software debugging techniques and identify their broad classification [4–8]. In the literature, we find that there is a scarcity of automated techniques for fault localization and debugging. The breakpoint-based debugging is very popular software debugging technique, but it must be done by the tester manually setting the breakpoints [9]. Similarly, the algorithmic and omniscient debugging is also manual [8–10]. In the study we find that the spectrum-based software debugging is the technique that can be automated [11, 12]. Few works have been carried out in the field of software debugging using spectrum-base formulas. In the paper, we present a study of some spectrum-based techniques and develop an automatic software debugging tool for JAVA programs.

2 Background Wong and Qi [2] presented an augmentation method a