Identifying Semantic Differences in AspectJ Programs

1. Identifying Semantic Differences in AspectJ Programs Martin Görg and Jianjun Zhao Computer Science Department, Shanghai Jiao Tong University

2. Outline • Motivation and Background • Difference Analysis Algorithm • Evaluation of Quality and Feasibility • Conclusions

3. Motivation and Background

4. Static Semantic Difference Analysis • static: source code analysis at compile time • semantic: differences in behavior modified P P’

5. Why solve the Problem? • Motivation • Reduce testing costs • Produce correct software • Possible applications • Debugging support • Regression test selection • Program slicing

6. AOP and AspectJ • AOP encapsulates crosscutting concerns • AspectJ • implementation of AOP • extension to Java Aspect Code Base Code aspect A { double C.d; before() : … { } …. } public class C { int i; void m1() { } … } Introduce Advise

7. AspectJ Example • aspect Constraints { • public boolean Shape.immovable = false; • void around(Shape s) :execution (public Shape+.set*(..))&& target(s) { • if (!s.immovable) {proceed( ) ; } } } ITD around advice

8. Hammocks • Single entry • Single exit • For any directed graph S E

9. Motivational Example 1 public class Point extends Shape { 2 private int x, y; 3 public void setX(int i){ 4 x = i; } 5 public void setY(int i){ 6 y = i; } 1 public class Point extends Shape { 2 private int x, y; 3 public void setX(int i){ 4 x = i; } 5 void setY(int i){ 6 y = i; } a change in visibility alters program execution

10. Difference Analysis Algorithm for AspectJ Programs

11. Algorithm Outline • For every module in P find a matching module in P’ (module-level matching) • Build extended CFGs for all modules in P and P’ and identify hammocks • Perform node-by-node comparison on every pair of hammock graphs(node-level matching)

12. Matching at Module Level • Signature matching • Disjunctive matching • Obtain best match from multiple candidates P: P’: P: P’: public void p1.C1.add(int, Object) boolean p1.C1.add(int, Object) public void p1.C1.add(int, Object) boolean p1.C1.add(double, Object)

13. 1. Matching at Module Level • Problem: Not every AspectJ construct has a signature (most importantly: advices) • Solution: • Define new AspectJ signatures(e.g.[strictfp] before (Formals) : [throws TypeList] : Pointcut {Body}) • Define disjunctive patterns for these signatures

14. 2. Build CFGs and Hammocks

15. 3. Matching at Node Level • Simultaneous graph traversal • Node-by-node comparison • Recursive • Two user inputs • Similarity threshold (S) • Maximum lookahead (LH)

16. 3. Matching at Node Level Similarity Threshold S = 0.5; Lookahead LH = 1 P P’ X X P P’ S S H H U U V V’ E E e e Y Y

17. Tests and Evaluation Quality and Feasibility

18. Signature definitions and disjunctive matching • Minimal change with maximal effect • Deficits: some combinations and swapped statements

19. S < 0.6: LH has only minor impact • LH  20: within one minute • S  0.6: slow for LH > 20, but not needed

20. Conclusions New Findings and Open Tasks

21. What we did • New signatures for AspectJ modules • Disjunctive matching • for AspectJ and Java modules • a solution for modified signatures • Application of hammock algorithm • from OO to AO • Evaluations using a tool implementation

22. Conclusions • Disjunctive matching is a good idea • modules are correctly matched • more work for node-level matching • eliminates work of type-level matching • replaces user interaction • Type-level matching is not required • Hammock graph matching can be applied given: • correctly matched modules • appropriately modeled and labeled CFGs

23. Future Work • Improve disjunctive matching patterns • Extend CFG representations • Solve the swapping problem • Handle dynamic pointcuts

24. Thank You for listening Questions?