SEKE 2012 Proceedings - Knowledge Systems Institute

contains broadcast of Subject to Observer nor event notification
or data observation. Thus, in fact, DPRE identifies 7 instances,
and only 2 of which is TP, so the precision is 28.6%.
TABLE II.
OBSERVER INSTANCES IDENTIFIED BY JDP-DETECTOR
Subject/Observer Event Object Manual
1 Figure/FigureChangeListener FigureChangeEvent T
2 StandardDrawing/
StandardDrawingView
3 StandardDrawingView/
AbstractTool
4 StandardDrawing/
FigureChangeEventMulticaster
DrawingChangeEvent
MouseEvent
FigureChangeEvent
We download N. Tsantalis's tool DPD 4.5 [5], and recognize
3 Observer instances (StandardDrawing/Drawing-
ChangeListener, StandardDrawingView/Painter, Standard-
DrawingView/Figure) with it. The first one of their instances is
the same as instance 2 of Table II. The other two instances is
not included in our result. The second one of their instances is
excluded by structure analysis of JDP-Detector. For the third
instance, Notify method of StandardDrawingView passes its
parameter instead of its fields to Update methods of Figure.
The dependence path of the data transfer process is included in
the SDG calculated by JDP-Detector, but is neglected by JDP-
Detector.
Based on the comparison, we conclude that except for some
defects in implementation of JDP-Detector, our approach is
effective in revealing the data flow of the Observer pattern, and
it identifies an instance (item 3 in Table II) that is not identified
by DPD and DPRE.
D. Time Consumption of the Identification Process
Our experiment environment is as following: Intel
Core(TM)2 Quad CPU Q8400 @ 2.66GHz processor, 3GB
RAM, Windows XP operation system. Identification of Adapter,
Strategy, State, and Observer instances in JHotDraw 5.1 by
JDP-Detector uses 5.217s, 1.793s, 7.39s and 14.082s, respectively.
VI. CONCLUSION
In this paper, we propose an a pproach to identify design
pattern instances in source codes using dependency information.
Our approach improves the recognition precision (especially in
distinguishing State instances from Strategy instances), and
identifies the variables playing roles in design patterns that
cannot be identified by existing approaches.
Compared with the approaches using method call sequences,
our approach reflects the effect of method call but has less
constraint to the method call sequence, and is somewhat more
flexible.
T
T
F
Using data flow information, we can describe the characteristics
of design patterns more detailed, which brings high identification
precision, but may sometimes seems too strict. And
furthermore, our approach can only identify design patterns
that has obvious data flow characteristics.
REFERENCES
[1] E. Gamma, H. Richard, R. Johnson, and J, Vlissides, Design Patterns:
Elements of Reusable Object-Oriented Software. Boston: Addison-
Wesley Longman, 1995.
[2] X. Zhou, J. Q ian, L. Chen, and B. Xu, "Identification of centrally
managed aggregations in design patterns using shape analysis,". Journal
of Software, Vol.21, No. 11, pp. 2725−2737, November 2010.
[3] A. V. Aho, M. S. Lam, R. Sethi and J. D. Ullman, Compilers: Principles,
Techniques and Tools, 2nd ed., Addison-Wesley Longman Publishing
Co., Inc., Boston, MA, 2007.
[4] R. Vallee-Rai, L. Hendren, V. Sundaresan, P. Lam, E. Gagnon and P. Co,
"Soot — A Java optimization framework," In: IBM Centre for Advanced
Studies Conference(CASCON), November 1999.
[5] N. Tsantalis, A. Chatzigeorgiou and G. Stephanides, "Design pattern
detection using similarity scoring," IEEE Transactions on software
engineering, Vol. 32, Issue 11, pp. 896-909, November 2006.
[6] Y. Guéhéneuc, J. Guyomarc’h and H. Sahraoui, "Improving designpattern
identification: a new approach and an exploratory study," Journal
of Software Quality Control. Vol. 18, Issue 1, pp. 145–174, March,
2010.
[7] A. D. Lucia, V. Deufemia, C. Gravino and M. Risi, "Improving
behavioral design pattern detection through model checking," In: 14th
European Conference on Software Maintenance and Reengineering
(CSMR), pp. 176-185, March 2010.
[8] A. D. Lucia, V. Deufemia, C. Gravino and M. Risi, "Design pattern
recovery through visual language parsing and source code analysis,"
Journal of Systems and Software, Vol. 82, Issue 7, pp. 1177–1193, July
2009.
[9] N. Pettersson, Welf and J. Nivre, "Evaluation of a ccuracy in design
pattern occurrence detection," IEEE Trans. Software Eng. Vol. 36, No. 4,
pp. 575-590, July/August, 2010.
[10] S. Romano, G. Scanniello, M. Risi and C. Gravino, "Clustering and
lexical information support for the recovery of design pattern in source
code," In: 27th IEEE International Conference on Software Maintenance
(ICSM), pp. 500-503, November, 2011.
[11] J. Y. Guéhéneuc and G. Antoniol, "Identification of behavioural and
creational design motifs through dynamic analysis," Journal of Software
Maintenance and Evolution: Research and Practice, Vol. 22, Issue 8, pp.
597627, December 2010.
[12] G. Rasool, I. Philippow and P. Mäder, "Design pattern recovery based
on annotations," Advances in Engineering Software, Vol. 41, Issue 4, pp.
519–526, April 2010.
[13] M. V. Detten and D. Travkin, "An evaluation of the Reclipse tool suite
based on the static analysis of JHotDraw," Technical Report tr-ri-10-322,
Software Engineering Group, Heinz Nixdorf Institute, University of
Paderborn, Oct 2010.
[14] S. Horwitz, T. Reps and D. Binkley, "Interprocedural slicing using
dependence graphs," ACM Transactions on Programming Languages
and Systems (TOPLAS), vol. 12, Issue 1, pp. 26-60, January 1990.
[15] J. Ferrante, K. Ottenstein, and J. Warren, "The program dependence
graph and its use in optimization". ACM Trans. Program. Lang. Syst.
Vol. 9, Issue 3, pp. 319-349, July 1987.
292