Computing Dynamic Slices of Concurrent Feature-Oriented Programs

This paper proposes a dynamic slicing algorithm for concurrent feature-oriented programs. The algorithm is named concurrent feature-oriented node-marking dynamic slicing (CFNMDS) algorithm. The proposed dynamic slicing technique uses a dependence based intermediate representation named concurrent co...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Arabian journal for science and engineering (2011) Ročník 44; číslo 11; s. 9471 - 9497
Hlavní autoři: Sahu, Madhusmita, Mohapatra, Durga Prasad
Médium: Journal Article
Jazyk:angličtina
Vydáno: Berlin/Heidelberg Springer Berlin Heidelberg 01.11.2019
Springer Nature B.V
Témata:
Algorithms
Dependence
Engineering
Graphical representations
Humanities and Social Sciences
Marking
multidisciplinary
Nodes
Product lines
Research Article - Computer Engineering and Computer Science
Run time (computers)
Science
Slicing
Mixin layer
FeatureIDE
Concurrent composite feature dependence graph (CCFDG)
Jak
Feature-oriented programming (FOP)
Concurrent feature-oriented node-marking dynamic slicing (CFNMDS)
Refinements
Concurrent composite control flow graph (CCCFG)
ISSN:2193-567X, 1319-8025, 2191-4281
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:This paper proposes a dynamic slicing algorithm for concurrent feature-oriented programs. The algorithm is named concurrent feature-oriented node-marking dynamic slicing (CFNMDS) algorithm. The proposed dynamic slicing technique uses a dependence based intermediate representation named concurrent composite feature dependence graph (CCFDG). It is based on marking and unmarking of the executed nodes in CCFDG appropriately during runtime. Ten standard open source product lines have been taken to experiment the proposed CFNMDS algorithm. Jak codes and feature-oriented models have been developed for these product lines. The advantage of the proposed approach is that no trace file is used to store execution history of the program. Also, no extra nodes are created during runtime in the proposed approach. Creation of extra nodes leads to take more time for marking and unmarking process. Use of any extra file also leads to higher runtime overhead for input/output operations. This results in taking more time to compute the slices. So, it is necessary to compute dynamic slices in less time without using any trace file and extra nodes. Using the proposed approach, slices can be extracted in O(1) i.e., in constant time. The slice computation time for various combinations of features for each product line has also been measured.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2193-567X
1319-8025
2191-4281
DOI:10.1007/s13369-019-04091-3