• DocumentCode
    600295
  • Title

    A comparison of database fault detection capabilities using mutation testing

  • Author

    McCormick, Donald W. ; Frakes, W.B. ; Anguswamy, R.

  • Author_Institution
    Comput. Sci., Virginia Tech., Falls Church, VA, USA
  • fYear
    2012
  • fDate
    20-21 Sept. 2012
  • Firstpage
    323
  • Lastpage
    326
  • Abstract
    Mutation testing involves systematically generating and introducing faults into an application to improve testing. A quasi-experimental study is reported comparing the fault-detection capabilities of realworld database application test suites to those of an SQL vendor test suite (NIST SQL) based on mutation scores. The higher the mutation score the more successful the test suite will be at detecting faults. The SQLMutation tool was used to generate query mutants from beginner-level sample schemas obtained from three popular real-world database test suite vendors - MySQL, SQL Server, and Oracle. Four SQLMutation operators were applied to both realworld and NIST SQL vendor compliance test suites - SQL Clause (SC), Operator Replacement (OR), NULL (NL) and Identifier Replacement (IR). Two mutation operators, SC and NL generated significantly lower mutation scores in real-world test suites than for those in the vendor test suite. The IR operator generated significantly higher mutation scores in real-world test suites than for those in the vendor test suite. The OR operator produced roughly the same mutation scores in both the real-world and vendor test suites.
  • Keywords
    SQL; database management systems; fault tolerant computing; program testing; MySQL; NIST SQL suite; NULL; Oracle; SQL Server; SQL clause; SQL vendor test suite; SQLMutation tool; Structured Query Language; database fault detection capability; identifier replacement; mutation score; mutation testing; operator replacement; Databases; ISO standards; NIST; Servers; Software; Testing; SQL; database testing; mutation testing; test adequacy criteria;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Empirical Software Engineering and Measurement (ESEM), 2012 ACM-IEEE International Symposium on
  • Conference_Location
    Lund
  • ISSN
    1938-6451
  • Print_ISBN
    978-1-4503-1056-7
  • Electronic_ISBN
    1938-6451
  • Type

    conf

  • DOI
    10.1145/2372251.2372310
  • Filename
    6475435