Representing and Generating Mutant Programs in Support of Measuring Test Suite Effectiveness

1
Representing and Generating Mutant Programs in
Support of Measuring Test Suite Effectiveness

• Ronald Finkbine, Ph.D.Indiana University
  SoutheastNew Albany, INrfinkbine_at_ius.edu

2
Introduction

• Definitions
• Purpose
• Background
• Mutation Operators
• Sample
• Database Flow
• Summary

3
Definitions

• Source program
• Test suite
• Mutation operator (s)
• Mutant
• Seed with a fault

4
Purpose

• Inject common syntax changes
• Induce branch execution
• Measure ratio of mutants killed by test suite
• Between 0 and 1, higher is better
• Strength of test suite
• Examine equivalent mutants (better code?)

5
Background

• Fortran, Mothra, (Ada)
• MuJava (JMutation)
• C, Proteum
• Java Objects, Sofya
• GAMUT uses genetic algorithms
• Parasoft Insure

6
Mutation Operators

• Scalars integer, floats, char data
• Increment, decrement, set constant
• Operators math, logic, relational
• Collections empty, remove, add, mutate element,
  reorder

7
Sample Token Set
import java.io. class test public static void
main (String args) System.out.println
("hello world") //main //class
test Figure 1
Token 1 1 1 import Token 2 1 8 java Token 3 1
12 . Token 4 1 13 io Token 5 1 15 . Token 6 1 16
Token 7 1 17 Figure 2
8
Token_Tbl and MutantOp_Tbl

• Number
• Column
• Line
• Value
• Spelling

• Number
• TokenNumber
• MutOperator
• CompilesStatus

9
MutantTokens_Tbl

• Number
• Column
• Line
• Value
• Spelling

10
MutantExecution_Tbl

• Number
• PgmOrMutant
• RunTimeExpected
• TestCaseNumber
• TestCaseResult

11
TestCase_Tbl

• Number
• ControlFileName
• InputFileName
• OutputFileName

12
Database Flow
13
Summary

• Definitions
• Purpose
• Background
• Mutation Operators
• Sample
• Database Flow