Program slicing based on specification

Chung, In Sang; Lee, W. K.; Yoon, Gwang Sik; Kwon, Yong Rae

doi:10.1145/372202.372784

Cited by 14 publications

(24 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unlike in our approach, loop invariants are required and only groups of instructions that are at consecutive program locations can be removed. This approach improves and combines older approaches [9,8], an implementation also exists [10].…”

Section: Related Workmentioning

confidence: 97%

Using Relational Verification for Program Slicing

Beckert

Bormer

Gocht

et al. 2019

Software Engineering and Formal Methods

View full text Add to dashboard Cite

Program slicing is the process of removing statements from a program such that defined aspects of its behavior are retained. For producing precise slices, i.e., slices that are minimal in size, the program's semantics must be considered. Existing approaches that go beyond a syntactical analysis and do take the semantics into account are not fully automatic and require auxiliary specifications from the user. In this paper, we adapt relational verification to check whether a slice candidate obtained by removing some instructions from a program is indeed a valid slice. Based on this, we propose a framework for precise and automatic program slicing. As part of this framework, we present three strategies for the generation of slice candidates, and we show how dynamic slicing approaches-that interweave generating and checking slice candidatescan be used for this purpose. The framework can easily be extended with other strategies for generating slice candidates. We discuss the strengths and weaknesses of slicing approaches that use our framework.

show abstract

Section: Related Workmentioning

confidence: 97%

Using Relational Verification for Program Slicing

Beckert

Bormer

Gocht

et al. 2019

Software Engineering and Formal Methods

View full text Add to dashboard Cite

show abstract

“…Comuzzi et al [5] and Chung et al [4] have provided algorithms for code analysis enabling to identify spurious commands (commands that do not contribute to the validity of the postcondition). The former paper presents a variant of program slicing, called p-slice or predicate slice, using Dijkstra's weakest preconditions to determine which statements will affect a specific predicate.…”

Section: Theoretical Backgroundmentioning

confidence: 99%

“…The combination of slicing techniques with program verification has been proposed [11,5,4] with a double goal: on one hand, slicing may help in optimizing the verification process, allowing one to delete the statements that do not contribute to the validity of a given specification (the resulting slice is smaller and thus easier to verify than the original program). On the other hand, verification inspires a novel, much more powerful way of slicing programs (with respect to traditional, syntax-based slicing), based on their axiomatic semantics.…”

Section: Introductionmentioning

confidence: 99%

“…In this paper we introduce GamaSlicer, a tool that includes both traditional program verification functionality (it is a Verification Conditions Generator) and also a highly parameterizable semantic slicer. It includes all the published algorithms (precondition-based slicing [4], postcondition-based slicing [5], specification-based slicing [4] and their variants) as well as a new, more effective algorithm that we have developed. The tool is extremely useful for comparing the effectiveness of slicing algorithms; it also allows users to perform diverse workflows such as for instance…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

GamaSlicer

Cruz

Henriques

Pinto

2010

Proceedings of the Tenth Workshop on Language Descriptions, Tools and Applications

View full text Add to dashboard Cite

In this paper we present the GamaSlicer tool, which is primarily a semantics-based program slicer that also offers formal verification (generation of verification conditions) and program visualization functionality. The tool allows users to obtain slices using a number of different families of slicing algorithms (precondition-based, postcondition-based, and specification-based), from a correct software component annotated with pre and postconditions (contracts written in JML-annotated Java). Each family in turn contains algorithms of different precision (with more precise algorithms being asymptotically slower). A novelty of our work at the theoretical level is the inclusion of a new, much more effective algorithm for specification-based slicing, and in fact other current work at this level is being progressively incorporated in the tool. The tool also generates (in a step-by-step fashion) a set of verification conditions (as formulas written in the SMT-lib language, which enables the use of different automatic SMT provers). This allows to establish the initial correctness of the code with respect to their contracts.

show abstract

“…To verify if the component invocations respect both the precondition and postcondition we will resort to slicing techniques. Traditional program slicing is a syntactic technique, but semantic forms have also been studied for over 10 years now, which combine slicing techniques with program annotation and verification, to identify synergies and take advantage of good practices on both sides [4], [5], [6].…”

Section: Safety Reuse Of Software Componentsmentioning

confidence: 99%

Contract-Based Slicing Helps on Safety Reuse

Areias

Cruz

Pinto

2010

2010 IEEE 18th International Conference on Program Comprehension

View full text Add to dashboard Cite

In this poster we describe a work in progress aimed at using a variant of specification-based slicing to improve the reuse of annotated software components, developed under the so called design-by-contract approach. We have named this variant as contract-based because we use the annotations, more precisely the pre and post-conditions, to slice programs intra and inter-procedures. The idea, expressed in the poster, is to take the precondition of the reused annotated component as slicing criterion, and slice backward the program where the component is called. In that way, we can isolate the statements that have influence on the variables involved on the precondition and check if it is preserved by that invocation, or not.

show abstract

Program slicing based on specification

Cited by 14 publications

References 9 publications

Using Relational Verification for Program Slicing

Using Relational Verification for Program Slicing

GamaSlicer

Contract-Based Slicing Helps on Safety Reuse

Contact Info

Product

Resources

About