| Sign up

PDF (9.9 MB)

Cite

Collect

Submit Manuscript

Show Outline

Figures (11)

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 8

Fig. 9

Fig. 10

Tables (6)

Table 1

Table 2

Table 3

Table 4

Table 5

Open Access

CCDive: A Deep Dive into Code Clone Detection Using Local Sequence Alignment

Yasir Glani^¹(), Luo Ping^¹, Syed Asad Shah^², Lin Ke^¹

1School of Software, Tsinghua University, Beijing 100084, China

2Department of Computer Engineering, Bilkent University, Ankara 06800, Turkey

Show Author Information

Abstract

The rapid evolution of software development has accentuated the deficiencies of prevailing code clone detection techniques. As modern applications become more complex, traditional cloning tools often struggle to detect general and large-gap clones that undergo regular modification. Such challenges pose threats to software integrity, emphasizing the critical need for improved code cloning techniques. Observing the prevailing gap, we propose an innovative code clone dive (CCDive) code cloning technique, which is designed to detect an extensive range of clones, from direct clones to the often challenging large-gap clones, thoroughly covering different categories, such as very strongly Type-III, strongly Type-III, and moderate Type-III clones. In CCDive, the fusion of a level-by-level abstraction and an innovative similarity matching algorithm ensures the recognition of clones even when nearly half the original code in the chunk has been modified. Furthermore, by integrating the Smith–Waterman local sequence alignment, the capability of CCDive to spot exact code transformation locations can be enhanced. In a comprehensive evaluation, CCDive was compared with well-known code cloning techniques. The efficacy of CCDive was measured using precision, recall, F1-score, accuracy, and efficiency. CCDive consistently surpassed other techniques in the precision, recall, F1-score, and accuracy metrics for both file-based and function-based clone detection. The robust performance of CCDive emphasizes its effectiveness, reliability, accuracy, and efficiency, making it well-suited for practical applications in the real world.

Keywords

clone detection software maintenance software security software reuse

References

[1]

Q. U. Ain, W. H. Butt, M. W. Anwar, F. Azam, and B. Maqbool, A systematic review on code clone detection, IEEE Access, vol. 7, pp. 86121–86144, 2019.

Crossref Google Scholar

[2]

H. Zhang and K. Sakurai, A survey of software clone detection from security perspective, IEEE Access, vol. 9, pp. 48157–48173, 2021.

Crossref Google Scholar

[3]

S. Kim and H. Lee, Software systems at risk: An empirical study of cloned vulnerabilities in practice, Computers & Security, vol. 77, pp. 720–736, 2018.

Crossref Google Scholar

[4]

Y. Golubev, V. Poletansky, N. Povarov, and T. Bryksin, Multi-threshold token-based code clone detection, in Proc. IEEE Int. Conf. Softw. Anal., Evol. Reengineering (SANER ), pp. 496–500, 2021.

Clone type	Granularity
Type-I	Exact clone
Type-II	Renamed clone
VST-III	90%−100% similarity
ST- III	70%−90% similarity
MT- III	50%−70% similarity

Dataset	Number of files	Number of methods	Number of functions	Number of lines
D4JDataset	85173	278110	675696	10631318
IJADAtaset	100000	252022	961655	7164932
GTHDataset	100000	172307	364495	7768509

Technique	Dataset	Type-III					Type-III LG
Technique	Dataset	Precision	Recall	F1-score	Accuracy	D.Time (s)	Precision	Recall	F1-score	Accuracy	D.Time (s)
CCDive	D4JDataset	0.99	0.79	0.88	0.89	2.51	0.98	0.63	0.77	0.81	2.52
	IJADataset	0.98	0.89	0.93	0.94	3.55	0.97	0.71	0.82	0.84	3.63
	GTHDataset	0.95	0.82	0.88	0.89	1.37	0.93	0.56	0.70	0.76	1.37
DroidMD	D4JDataset	0.57	0.08	0.14	0.51	10.70	0.60	0.09	0.16	0.52	10.96
	IJADataset	0.33	0.04	0.07	0.48	9.05	0.43	0.06	0.11	0.49	9.68
	GTHDataset	0.82	0.23	0.36	0.59	5.02	0.81	0.22	0.35	0.58	4.68
PDG	D4JDataset	0.65	0.92	0.76	0.72	0.75	0.64	0.89	0.74	0.52	0.73
	IJADataset	0.47	0.50	0.49	0.47	1.99	0.47	0.49	0.48	0.49	1.80
	GTHDataset	0.51	0.59	0.55	0.52	1.44	0.50	0.57	0.54	0.58	1.70
VUDDY	D4JDataset	0.66	0.86	0.74	0.70	0.23	0.65	0.83	0.73	0.69	0.28
	IJADataset	0.47	0.49	0.48	0.47	0.33	0.46	0.48	0.47	0.46	0.30
	GTHDataset	0.51	0.58	0.54	0.51	1.09	0.50	0.57	0.54	0.51	1.16
LWH	D4JDataset	0.54	0.98	0.69	0.56	1.09	0.53	0.97	0.69	0.56	2.44
	IJADataset	0.50	0.97	0.66	0.49	13.13	0.49	0.94	0.63	0.48	12.93
	GTHDataset	0.56	0.98	0.72	0.61	16.79	0.56	0.98	0.72	0.61	16.79
IBEFT	D4JDataset	0.44	0.04	0.07	0.49	20.08	0.44	0.04	0.07	0.49	20.01
	IJADataset	0.60	0.03	0.06	0.51	15.71	0.60	0.03	0.06	0.51	18.35
	GTHDataset	0.85	0.04	0.08	0.52	9.12	0.75	0.03	0.06	0.51	8.49
AST	D4JDataset	0.43	0.57	0.49	0.41	1.26	0.44	0.58	0.50	0.41	1.22
	IJADataset	0.48	0.51	0.49	0.47	1.67	0.46	0.48	0.47	0.46	1.67
	GTHDataset	0.47	0.66	0.55	0.46	0.96	0.46	0.63	0.53	0.45	0.94

Technique	Dataset	Type- III					Type- III LG
Technique	Dataset	Precision	Recall	F1-score	Accuracy	D.Time (s)	Precision	Recall	F1-score	Accuracy	D.Time (s)
CCDive	D4JDataset	0.99	1.00	1.00	0.99	3.32	0.99	1.00	1.00	0.99	2.23
	IJADataset	0.98	0.98	0.98	0.98	3.61	0.99	0.96	0.97	0.97	3.76
	GTHDataset	0.97	0.91	0.94	0.94	1.25	0.97	0.83	0.89	0.90	1.28
DroidMD	D4JDataset	0.96	0.49	0.65	0.73	7.41	0.98	0.47	0.647	0.73	8.27
	IJADataset	0.89	0.40	0.55	0.68	12.88	0.90	0.36	0.51	0.66	15.02
	GTHDataset	0.96	0.52	0.68	0.75	2.91	0.86	0.46	0.62	0.72	3.22
PDG	D4JDataset	0.66	0.93	0.77	0.72	0.64	0.65	0.93	0.77	0.71	0.66
	IJADataset	0.60	0.80	0.69	0.64	2.50	0.60	0.78	0.68	0.62	2.58
	GTHDataset	0.60	0.80	0.69	0.64	1.84	0.61	0.80	0.69	0.65	2.00
VUDDY	D4JDataset	0.67	0.91	0.77	0.73	0.21	0.66	0.91	0.77	0.72	0.21
	IJADataset	0.60	0.75	0.67	0.62	0.33	0.60	0.74	0.66	0.62	0.35
	GTHDataset	0.59	0.73	0.65	0.61	1.92	0.60	0.72	0.65	0.61	1.79
LWH	D4JDataset	0.53	1.00	0.69	0.55	2.75	0.52	1.00	0.68	0.53	3.16
	IJADataset	0.50	0.97	0.66	0.49	9.04	0.49	0.67	0.66	0.49	2.73
	GTHDataset	0.53	0.99	0.69	0.55	6.28	0.53	0.99	0.69	0.55	6.31
IBEFT	D4JDataset	0.93	0.28	0.43	0.63	16.26	0.96	0.26	0.41	0.62	16.59
	IJADataset	0.962	0.22	0.35	0.60	26.45	0.91	0.20	0.33	0.59	30.67
	GTHDataset	1.00	0.33	0.50	0.67	5.97	1.00	0.30	0.46	0.65	6.17
AST	D4JDataset	0.55	0.73	0.63	0.56	1.18	0.52	0.76	0.62	0.54	1.09
	IJADataset	0.53	0.62	0.57	0.54	1.73	0.52	0.57	0.54	0.52	1.76
	GTHDataset	0.50	0.75	0.60	0.51	0.88	0.49	0.67	0.57	0.49	0.91