Cm24585587

Prof.D.M.Thakore, S.J.Sarde / International Journal of Engineering Research and Applications
(IJERA) ISSN: 2248-9622 www.ijera.com
Vol. 2, Issue 4, July-August 2012, pp.585-587

Assessing the Software Complexity and Security metrics from UML
Class diagram
1
Prof.D.M.Thakore, 2S.J.Sarde (M.Tech Student)
1
Department of Computer Engineering
Bharati vidyapeeth Deemed University College of Engineering, Pune-43, Maharashtra, India
2
Department of Computer Engineering
Bharati vidyapeeth Deemed University College of Engineering, Pune-43, Maharashtra, India

Abstract-As the standard definition of software software quality metrics for re-engineering process is
engineering it is the development of sound done manually or with some tool but it can take more
engineering principle in order to achieve the time due to complexity and unambiguous nature of
software which is effective, efficient, understandable, source code. If some user require particular metrics
and further it can be run on any real time machine. urgently to re-engineer the existing software system,
In software engineering, Software metrics are very because such metrics are vital or necessary for effective
useful for a forward engineering of re-engineering re-engineer process. If such metrics are extracted
process of existing software system. Also, they are manually then it should take more time and also,
absolutely necessary in re-engineering process. They budget. That is the derive force behind the development
show exactness, clear picture and understanding of of such a system which does it automatically without
the existing software system. It should be the first any manual efforts
step in effective re-engineering process. Consider e.g. Both users (developer and system
Quality of software systems heavily depends analyzer) should review the source code for complexity
on their structure, which affects maintainability and and security metrics manually. It should be good for
readability. However, the ability of humans to deal small software system, but if software system is large
with the complexity and security of large software then it is quite cumbersome because large software
systems is limited. In this paper we are proposing system may contain large no. of source code (LOC)
system methodology approach, that should measure with more complexity and less security. Also, it should
coupling, cohesion (Complexity), Data Access and require more time for calculation of complexity and
Operation Access Metric (Security) which dependent security and also wastage of resources.
on the assumption that the attributes, methods, Therefore, to develop such suitable Software
relationship and classes of Object-Oriented systems Reverse Engineering Tool (SRET) which could help
are connected in more than one way. system analyzer and developer to review the source
code and calculate the source code complexity and
Index term: Class diagram, object oriented language, security with automatic framework using UML Class
software quality metrics-complexity and security, Diagram.
source code, Unified Modelling language etc.
II Existing tools and their limitations
I Introduction Software engineering developer and system
Software engineers generally use indirect analyst should be based on the tools for implementing
measures that lead to metrics which provide a these metrics to support them in quality evaluation and
quantitative basis for understanding the underlying ensure tasks to allow to measure software quality and to
information in software development processes. deliver the information needed as input for their
Software metrics have always been important for decision making and engineering processes. Currently a
software developers to assure the quality of some large body of software metrics tools exists. But these
representation of software and organizations are are not the tools which have been used to evaluate the
achieving promising results through their use. software metrics.
Therefore, to develop suitable software metrics models
for user (developer) who urgently need them for re- 1) Analyst4j : It is based on the Eclipse platform and
engineering of existing software system. Therefore, this available as a stand-alone Rich Client Application or as
proposed methodology approach “source code analysis an Eclipse IDE plug-in. It features search, metrics,
for software quality metrics” proposes that the analyzing quality and report generation for Java
evaluation of metrics as part of reverse engineering in programs [2].
re-engineering process i.e. Software Reverse
Engineering Tool (SRET). SRET should be developed 2) VizzAnalyzer: It is a tool for quality analysis. It
under two categories: 1) Complexity and 2) Security reads or parses software code and other design
Source code requirement analysis is for extracting specifications as well as documentation and performs a

585 | P a g e

number of quality analyses. The VizzAnalyzer is a particular, to propose security metrics to measure Data
framework or environment for analyses and Encapsulation (accessibility) and Cohesion
visualizations of existing software system. The (interactions) of a given object-oriented class from the
VizzAnalyzer is a framework designed to help point of view of potential information such as source
programmers or developer or system analyst in software code. Defining another security metrics which cover the
engineering activities like re-engineering [2,3]. entire source code of existing software including
coupling, inheritance, and cohesion [6].
Shortcomings with this technique:
1) It has high coupling values and less cohesion values. IV Proposed Work and system architecture
2) Also it ignores security metrics such as Data and
Operation Access Metrics.
3) It analysis only few source code language file such as
for java.

III Literature Survey
Software metrics are very useful in a forward
engineering of re-engineering process of existing
software system. Also, they are absolutely credential in
re-engineering process. They show exactness, clear
picture and concise understanding of the existing
software system. It should be the first step in effective
re-engineering process.
Use of the powerful and practical metrics like
cohesion, coupling of existing systems and develop or
generate new metrics and then add them to the new
complexity and security categories to enhance the
quality of the software re-engineering process which
leads to the enhance quality of the software which is
under the process of re-engineering[4].
Software reengineering is an expensive process due
to the complexity of the software, we cannot
emphasise on the area where the re-engineering work is
Figure 1: Proposed Architecture of System
required,. Coupling and Cohesion metrics are
complexity metrics out of particularly cohesion metrics
have the potential to help in this identification and to Here take the input from user (developer), a
measure progress. The most extensive work on such document which contains source code. Convert this
metrics is with cohesion metrics. It should use of source code document into Core Prototype Model
dependence information that make them an excellent (class diagram) specifies the entities and relations that
choice for cohesion measurement. can and should be extracted immediately from source
It should be raise the most important question such code. Then process this output for Complete Model as
as does a software developer or analyst which could be input specifies Object, Property, Entity and Association
access to complexity metric values for the program do a are made available to handle the extensibility
better job of restructuring the program? [5] requirement. On the output of Complete Model, then
Security metric is not considered as much as other apply our rules on the basis of which Security
quality attributes such as complexity metrics. Also, Accessibility and complexity metrics are extracted.
most security studies concentrate on the level of In the case of in information extraction:
individual program statements. Such type of approach
makes it hard and expensive to discover and fix DAM = No. of private (protected) attributes/total no.
vulnerabilities caused by design errors in the existing of attributes in class
system. OPM = No. of public methods/ total no. of methods
Therefore, in this paper we should also focus on the in a class
security design of an existing object oriented Cohesion = No. of methods interactions with
application and define security metrics. These metrics attributes in the program code / maximum
allow designers (developer or system analyzer) to find no. of methods interactions with attributes
out and fix security vulnerabilities at an early stage of Coupling = Access frequency of attributes of one
the re-engineering process which will help to reduce the class/sum of frequency of all attributes
cost of software reengineering as it reduces the rework
and consumption of resources which helps the designer 1. Core Prototype Model
to review the security metrics to make particular The Core Prototype Model states the entities
decision about security into re-engineering approach. In and relations that can and should be extracted

586 | P a g e

immediately from source code. The core model consists [2] “Comparing Software Metrics Tools”
of the main Object Oriented entities such as Classes, RudigerLincke, Jonas Lundberg and Welf Lowe
Methods, Attributes and Inheritance Definition. For the Software Technology Group School of
reengineering we require the other two ideas namely the Mathematic Mathematics and Systems
associations Invocation and Access. An Invocation EngineeringVaxjoUniversity,Sweden{rudiger.lin
presents the definition of a Method calling another cke|jonas.lundberg|welf.lowe}@vxu.se
Method and an Access presents a Method accessing an [3] A Qualitative Evaluation of a Software
Attribute. Development and Re-Engineering Project
In automated software modelling, the Source Code Thomas Panas,RudigerLincke, Jonas
as software requirement specification is translated to the Lundberg,Welf Lowe Software Technology
formal specifications such as Template Information Group MSI, University of Vaxjo, Sweden
with CDIF Format. [4] “Development and Application of Reverse
Engineering Measures” in a Re-engineering Tool
2. Complete-Meta-Model S. Zhou, H. Yang and P. Luker William C. Chu
Objects, Property, Entity and Association are made Department of Computer Science Department of
accessible to handle the prerequisite. For specifying Information Engineering De Montfort University
language plug-ins, it is authorized to define language Feng Chia University England Taiwan
precise classes and to add language precise attributes to [5] “An Empirical Study of Slice-Based Cohesion
existing Objects. Tool prototypes are more limited in and Coupling Metrics” Timothy M. Meyers and
extending the model: they can define tool precise David Binkley Loyola College in Maryland
properties for and can add attributes to existing Objects. Baltimore, Maryland 21210-2699, USA
They are, however, not authorized to extend the {tmeyers,binkley}@cs.loyola.edu
repertoire of entities and associations. [6] Alshammari, Bandar and Fidge, Colin J. and
Corney, Diane (2009) “Security metrics for
3. CDIF Information Exchange Format object-oriented class designs”. In: QSIC 2009
CDIF is standard for transferring models or Proceedings of : Ninth International Conference
standard for information exchange with different tools. on Quality Software , August 24-25, 2009, Jeju,
Key issue in the reengineering of large scale Korea. (In Press)
object-oriented systems is due to the heterogeneity in [7] “New Conceptual Coupling and Cohesion
today’s object-oriented programming languages. Metrics for Object-Oriented
Proposed system also generates Template Information Systems”BélaÚjházi, Rudolf Ferenc,
into CDIF Form for these programming constraints. TiborGyimóthy University of Szeged, Hungary
This is also added facility provided in the Department of Software Engineering
proposed system as compared to current existing work. ujhazi.bela@stud.u-szeged.hu, {ferenc,
gyimi}@inf.u-szeged.hu and Denys Poshyvanyk
V Conclusion The College of William and Mary, USA
In this paper we have emphasise on the Computer Science Department
software quality metrics complexity and security via denys@cs.wm.edu
analyzing UML class diagram which is obtained as an [8] “Reverse Engineering Component Models for
input from the source code and the document Quality Predictions” Steffen Becker, Michael
specification. The proposed work is fully automated Hauck, and MirceaTrifu FZI Research Center
eliminating the manual effort required from the Software Engineering Karlsruhe, Germany Klaus
developer and analyzer, further because of the Krogmann Karlsruhe Institute of Technolgy
elimination of manual work these system is effective, Software Design and Quality Karlsruhe,
efficient for the reengineering of the software which Germany Jan Kofroˇn Charles University in
already in existence with effective utilization of the key Prague Distributed Systems Research Group
resources . Prague, Czech Republic
[9] “An Exchange Model for Reengineering Tools”
VI References Sander Tichelaar and Serge Demeyer, Software
[1] “Beyond Language Independent Composition Group, University of Berne,
ObjectOrientedMetrics:Model Independent Switzerland, {demeyer,tichel}@iam.unibe.ch
Metrics” Michele Lanzalanza@iam.unibe.ch [10] “A Visual Analysis and Design Tool for
Software Composition Group Universit á di Planning Software Reengineerings” Martin Beck,
Berna, Svizzera and Jonas Tr ümper and JürgenDöllner
StéphaneDucasseducasse@iam.unibe.ch {martin.beck}, {jonas.truemper},
Software Composition Group Universit é de {juergen.doellner}@hpi.uni-potsdam.deHasso-
Berne, Suisse Plattner-Institute – University of Potsdam,
Germany

587 | P a g e

Cm24585587

More Related Content

What's hot (17)

Viewers also liked (20)

Similar to Cm24585587 (20)

Cm24585587