International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011 1
ISSN 2229-5518
Metrics for Component Based
Measurement Tools
P. Edith Linda, V. Manju Bashini, S. Gomathi
Abstract— The main aim of this paper is to integrate the different object oriented metric tools and make them available as a single add-on. The first part of this paper analyzed five different tools and they are migrated into one to make use of those tools in efficient manner.
Index Terms— Metrics, component, de-compiler, SLOC, LOC, NOS, NOP, Cyclomatic complexity
—————————— • ——————————
HE emerging Object oriented programming is most widely used by the programmers to develop the software. The object oriented programs are used in the IT world which is used for secured transactions and mainly convenient. Even though they are widely used, the com- plex and confusing structure may arise if the program is not properly measured and not properly structured. To overcome this problem many proposed varieties of metric tools to measure the criteria of Object oriented program- ming are used. Again the problem arises that some metric tools measure the particular criteria of the object oriented programs. In order to overcome the individuality of those
tools, they should be migrated.
Component based measurement tool is the concept of migrating many tools available widely. The main aim is to search some important tools and making them as a single add-on to provide the user friendly environment.
Each individual tool will measure some parameter to
measure the java program. But a single tool will not
measure all the parameters.
Thus there exist a need of collection of tools to measure the java programs which will satisfy all the pa- rameters to be measured.
Software metrics are used to measure the software quality to check whether it satisfies the requirements. Metrics are defined as “Quantifiable measures that could be used
————————————————
• P. Edith Linda, Assistant Professor, School of IT & Science, Dr. G. R. Da- modaran college of science, Coimbatore-641014 p.lindavinod@gmail.com
• V. Manju Bashini,, Assistant Professor, School of IT and Science, Dr. G. R. Damodaran college of science, Coimbatore – 641014 manjubashi- ni@gmail.com
• S. Gomathi, Research scholor, Dr. G. R. Damodaran college of science, Coim- batore- 641014. Gomathisrinivasn88@gmail.com
to measure characteristics of a software system or the software development process.” Software metrics are essential to plan, predict, monitor, control, evaluate, products and processes. The main goal of the software metrics is to reduce costs, Improve quality, Con- trol/Monitor schedule, small testing effort, many reusable fragments, to better understand the quality of the product and the program.
Metrics can be categorized into three Kinds, measures and size.
Two kinds of software metrics they are Product metrics: quantify characteristics of the product being de- veloped to measure the size, reliability and Process me- trics: quantify characteristics of the process being used to develop the software to measure the efficiency of fault detection.
Types of Measures are Direct Measures (internal attributes): to measure the Cost, effort, LOC, speed, memory and Indirect Measures (external attributes): to measure the Functionality, quality, complexity, efficiency, reliability, maintainability.
Size-oriented metrics are LOC - Lines Of Code, KLOC - 1000 Lines Of Code, LOC – Statement Lines of Code (ignore whitespace).
Components are the collections of many pre programmed tools which is used as the add-on page which is to make use of those tools. There are various tools available to measure the Java source code. Those tools are developed
IJSER © 2011
International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011 2
ISSN 2229-5518
and modeled based on the developer ’s point of view. Some tools will measure some parameters to be measured in java program. Initially the tools which are used to measure the java object oriented programs are searched and analyzed individually to make it as a component. The component based tools will be executed separately but available in a same location and some tools will provide a chart for the results when the program is executed.
In 1990’s, software systems for e-business transactions were built of components across multiple platforms, pro- gramming language and network protocols. No tools were available for detecting “bugs” or to measure the efficiencies in the programs (object oriented programs). Initially the errors are printed using the output statement and the number of lines [1] , number of packages and most important concepts of object oriented programs with the critical programs
Those programs which are used to measure the
OOPS programs seem to be critical and they are imple- mented with the large number of coding. In 1997, Inter- metrics, a software engineering company , proposed to develop a debugging tool for Component- based software systems. It only identifies the errors and yields the error free program.
But finding the error won’t give the remedy. Thus
the problem to find the efficiency of the program is solved with the help of the component based tools which effi- ciently provides the charts for the results and measures the program.
The main advantage of component based tool is
that the user can select their own tool to measure their program according to their needs. In this, the user can also see the links from where the tools are available, that is the user can also download or know more about the tool from the direct link.
The collection of more tools which will provide more flexibility and choice to the user or the programmer and he can know more about tools. Those tools will dis- play an error if the program is not correct.
Some tools will also provide the charts and warn-
ing message if the program is not structured and pro- grammed in a proper format based on the measurement criteria.
All the tools won’t meet the user required criteria. Some may be useful for generating report and some may be
used just for measuring the program. Some tool may
measure the oldest method of measuring the program with some limited parameters which are not in current use and which may be efficient to measure. All the listed tools will measure the common parameter like SLOC (Source Line Of Coding) and LOC (Lines Of Coding). The need depends on the programmer or the user who uses the component metrics. Integrating the various tools have many advantages and also some disadvantages. Those are listed.
Advantages on integrating components
1. Flexible
2. User friendly
3. Compatibility
4. Comparability
Disadvantages on integrating components
1. Confusion may arise about which tool to be used
2. Some may not meet the user requirement
3. Some tools are only trial versions
4. All the tools which are listed may have some common criteria which is not useful.
The paper by Amandeep Kaur et. al. [7] to formulate a framework of metrics representing the attributes of object oriented system. And the suggestion is that the impact of OO design on software quality characteristics made an experimental validation and a rework in defect programs. Encapsulation, inheritance, polymorphism, reusability, Data hiding and message-passing are the major attribute of an Object Oriented system and are the indicators of the object oriented programs. The well known quantifiable approach to express any attribute is a metrics. Empirical Data is collected from three different projects based on object oriented paradigms to calculate the metrics.
Lakshmi Narasimhan and Bayu Hendradjaya[8] defined a two suites which cover the static and dynamic aspects of the component assembly. The triangular metric which is used to classify the type and nature of the application, is formed with the complexity and criticality metrics. Dur- ing the runtime of the complete application the dyamic metrics are collected. This paper is mainly deals with the collection of some metrics and evaluated the results.
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011 3
ISSN 2229-5518
Shyam R. Chidamber and Chris F. Kemerer [11] stated that the metrics is must and needed for a new organiza- tion which is adopting a new technology for whom the proper practice is needed to evaluate the program size and other paradigms. To practice and to evaluate the pro- grammer need to implement the a new metric suite for Object Oriented programs for OO design. The empirical validation of a theoretical metrics of OO design and the development is the main idea of the paper.
Dr. Linda H. Rosenberg [12] stated the development of object oriented programs is very much needed in today’s life cycle. A different approach of design and implementa- tion, and also a different approach to software metrics is required by Object oriented development. Object oriented technology uses the objects as its fundamental blocks and not algorithms. The object oriented programs needs a special metrics to evaluate. Cyclomatic complexity is standard for traditional functional and procedural pro- grams. There are many proposed object oriented metrics for evaluating the object oriented program.
Dr K. Anbumani, K.P. Srinivasan [13] stated use of soft- ware metrics in evaluating the quality of software prod- ucts and the software development process. There are currently three sets of design metrics for object oriented software which are suggested by different groups of au- thors, Chidamber and kemerer metrics.
Lakshmi Narasimhan et. al.[14] of Department of Com- puter Science from East Carolina University, Greenville, NC, USA suggested that during the execution phase, there exist lack of metrics which will reduce the mainten- ance costs and defined metrics whose values are col- lected. Metrics are useful for increase the re-usability and assessing the maintenance cost of individual components and that of the application in which the component is integrated.
Radu Marinescu [15] given that the object oriented para- digm has influenced the software engineers for ease of program writing and implementing. The impact of mi- sunderstanding the object oriented designs made the de- sign complex and more tedious to program and poor OO design. Object oriented metrics have the benefits with the high degree of code reuse, higher maintainability and flexibility, etc., Now they can improve the quality of the design using the object oriented metrics.
Lakshmi Narasimhan and Bayu Hendradjaya[16] defined a two suites which cover the static and dynamic aspects of
the component assembly. The triangular metric which is used to classify the type and nature of the application, is formed with the complexity and criticality metrics. Dur- ing the runtime of the complete application the dyamic metrics are collected. This paper is mainly deals with the collection of some metrics and evaluated the results.
Magnus Andersson Patrik Vestergren[17] targeted to eva- luate whether the software metrics can be used to deter- mine the object-oriented design quality of a software sys- tem. There are several metrics to evaluate the object oriented design and programs. To validate each metric an experimental study was conducted. The final conclusion is software metrics can improve the system design quality such as complexity of methods/classes, package structure design and the level of abstraction in a system. Therefore metrics can be assured that the rules are followed and the program is properly written with the rules.
According to Saida Benlarbi et. al.[18] the basic premise behind the object oriented metrics which is an earlier pre- dictors of classes that contains error which may be found during the testing. In this paper they conclude that the empirical validation methods provide misleading conclu- sion to the validity of object oriented metrics. The result of the empirical validation is studied with the C++ which is an object oriented program.
The paper by Tieng Wei Koh et. al.[19] review the 12 ob- jected- oriented software metrics proposed in 90’s by Chi- damber, Kemerer and Li. The authors stated that the software cost estimation is mainly based on the size or predicting the volume of various kinds of software deli- verable items. Size and complexity is very critical and it influences the integral software development effort. Soft- ware cost estimation is criticsl because it is mainly based on the software size estimation.
Marco Scotto et. al.[20] metrics tools is to be updated for the standardization of software measures but it is not still standardize. Intermediate abstraction layer is the possible solution suggested here to decouple the extraction process from the use of the information. This paper presents the web metrics which is an automated tool for software metrics collection. This open source project is computed with the CK metric suite.
The tools which are analyzed in this paper are LOCC, JHAWK, CODE COUNTER. These tools are used to
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011 4
ISSN 2229-5518
measure the object oriented paradigms in the program. The main advantages of these tools are, it will generate the charts and will produce the reports about the pro- grams. The programmer should re-model or re-structure the program according to the report or the warning chart generated by those tools[9,10]. Another software called cavaj Decompiler which is used to convert the class files into java files for user convenience is also integrated into this component.
This is the tool to measure the Object oriented program to yield the LOC, SLOC, NOP, NOC, NOM, NOI in that par- ticular program which is given as input. The main draw- back of this metric tool is that it won’t generate the charts for the report or result generated by the tool [4]. LOCC metrics, which provides a grammar-based architecture and interface to the production and use of different size metrics. Developers can use the size metrics distributed with LOCC or can design their own metrics, which can be easily incorporated into LOCC. LOCC pays specific atten- tion to the problem of supporting incremental develop- ment, where a work product is not created all at once but rather through a sequence of small changes applied to previously developed programs.
This tool is used to measure the various parameters of CYCLOMATIC COMPLEXITY, COMMENT LINES, NOP in the given input program. The main advantage is that it will generate a chart for the program and also it will show the warning chart. It will mainly measures the looping and iterations which are present in the input program[6].
JHawk is a Stand alone, Eclipse plugin and
command line versions. JHawk Metric interchange format that allows you to record a snapshot of a code base and keep it for future use either in JHawk itself or in our new DataViewer product. JHawk Data Viewer, a standalone product that allows graphical and textual comparison of metrics over time.
The disadvantage is that the Halstead metrics are not used now for measuring the program. Some may need to measure those metrics.
This metrics is mainly used to measure the NOS, COM- MENT LINES, TOTAL of the two metrics. The main ad- vantage of this tool is, it will measure and generate the report for many languages like ASP, HTML, C++, C etc. The main advantage is it will measure many languages and will generate the three forms of reports according to
the user needs. The main drawback is it won’t cover the important parameters to be measured for the object oriented programs [4].
The main features of code counter are Easy to use with
just 3 steps, Quickly counts several types of source code - including Java, JSP, C or C++, VB, PHP, HTML, Delphi or Pascal, ASM, XML, COBOL etc., Smart Comment feature - knows which comment types are used by each language and counts accordingly, Save and load count profiles - no need to enter the same information over and over for the same source files to count, Supports unicode files and contents. Can also count unix type or windows type of text files (CRLF/CR only)
This tool is mainly used to measure the total lines of code (LOC), blank lines of code (BLOC), comment lines of code (CLOC), lines with both code and comments (C&SLOC), logical source lines of code (SLOC-L), McCabe VG com- plexity (MVG), and number of comment words (CWORDS). Physical executable source lines of code (SLOC-P) is calculated as the total lines of source code minus blank lines and comment lines. Counts are calcu- lated on a per file basis and accumulated for the entire project. Loc Metrics also generates a comment word his- togram[4].
SLOC Metrics measures the size of your source code based on the Physical Source Lines of Code metric rec- ommended by the Software Engineering Institute at Car- negie Mellon University (CMU/SEI-92-TR-019). Specifical- ly, the source lines that are included in the count are the lines that contain executable statements, declarations, and/or compiler directives. Comments, and blank lines are excluded from the count. When a line or statement contains more than one type, it is classified as the type with the highest precedence[5]. The order of precedence for the types is: executable, declaration, compiler direc- tive, comment and lastly, white space.
Main features of SLOC are Fast and efficient processing
of source files, Process any size of file, Process any num- ber of files. Flexible support for C/C++, C#, Java, HTML, Perl, Visual Basic and more. Ability to add definitions for other file types, Formatted and hyperlinked results in HTML file format. Project folders and files are sorted by their Source Lines of Code so that you can easily identify code intensive modules and files, Shows differences in counts of source lines of codes between two measure- ments by SLOCMetrics as DIFF.html in the output folder ,
Formatted results in CSV (Comma Delimited) file format,
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011 5
ISSN 2229-5518
Prepares results as a well-formed XML document , Mul- tiple SLOC Configurations can be saved as files and reused, can be run as a Console application us- ing sloccmd.exe present in the SLOCMetrics 3.0 installation directory. Configuration file path can be passed as a pa- rameter to sloccmd.exe, An easy to use, Windows-based graphical user interface.
Main drawback of the existing system is they are availa- ble individually. So the users may meet some difficulty in order to measure the program.In this proposed system, those tools are integrated and made as a single add-on for the users flexibility and user friendliness. All the tools are available in the single web page so that the user can access those tools efficiently. The java decompiler, an open source, also combined in this page which will convert the class files into java files.
The tools will measyre the different parameters
which need to be measured. The tools itself will generate the charts and reports according to the result. The tools also shows the warning message if the program have to restructure and also the error message if the given pro- gram is contain any error.
Metric tools | Parameters measured | Chart Genera- tion |
LOCC | loc, sloc, nop ,noc ,nom, noi | No |
JHAWK5 | cyclomatic complexity, comment lines, nop | Yes |
CODE COUNTER | nos, comment lines, total | Yes |
LOC ME- TRICS | loc, bloc, cloc, c&sloc, sloc-l, mvg, cwords, sloc-p | Yes |
SLOC METRICS | comments, blank lines, executable declarations, compiler directive, white space | yes xml, xsl, html, notepad |
As there are wide applications of association rules in data mining, it is important to provide good per-
formance. In view of this, we proposed a new algorithm which considers the significance of the item also but not only support of the item. The results through experiment shows that the computational cost of the link-based mod- el is reasonable. At the expense of three or four additional database scans, we can acquire results different from those obtained by traditional counting-based models. Particularly for sparse data sets, some significant item sets that are not so frequent can be found in the link-based model. Through comparison, we found that our model and method address emphasis on high-quality transac- tions. In this paper, a brief discussion of a number of al- gorithms was presented along with a comparative study of a few significant ones based on their performance and memory usage.
[1] http://www.atp.nist.gov/gems/97-06-0038gem.pdf
[2] An Analysis of Quality of Service Metrics and Frameworks in Grid Computing Environment, RussWakefield, Colorado State Uni- versity, Ft. Collins, Colorado
[3] Russ Wakefield, Colorado State University, Ft.Collins,Colorado, “An Analysis of Quality of Service Metrics and Frameworks in a Grid Computing Environment”.
[4] http://www.locmetrics.com/ [5] http://microguru.com/sloc/ [6]http://www.downloadpipe.com/free-trial/JHawk-5-Personal- License-download1326333.html
[7] Amandeep Kaur, Satwinder Singh, Dr. K. S. Kahlon, “A Metric
Framework for Analysis of Quality of Object Oriented Design”.
[8] V. Lakshmi Narasimhan, and Bayu Hendradjaya “Theoretical
Considerations for Software Component Metrics”
[9] Boehm, Barry Dr.; Software Engineering Economics; Prentice
Hall, Englewood Cliffs, NJ; 1981; 900 pages.
[10] Galorath, Daniel D. & Evans, Michael W.; Software Sizing, Es- timation, and Risk Management: When Performance is Measured Performance Improves; Auerbach, Philadelphia, AP; ISBN 10-
0849335930; 2006; 576 pages.
[11] Shyam R. Chidamber and Chris F. Kemerer, “A METRICS SUITE FOR OBJECT ORIENTED DESIGN”
[12] Dr. Linda H. Rosenberg “Applying and Interpreting Object
Oriented Metrics”.
[13] Dr. K. Anbumani, K. P. Srinivasan, “A Set of Objected Oriented design Metrics”.
[14] V. Lakshmi Narasimhan, P. T. Parthasarathy, and M. Das “Evalua-
tion of a Suite of Metrics for Component Based Software Engineering
(CBSE)”.
[15] Radu Marinescu, “Using Object-Oriented Metrics for Automatic Design Flaws Detection in Large Scale Systems, How to Use Object- Oriented Metrics in the Early Stages of a Redesign Operation”
[16] V. Lakshmi Narasimhan, and Bayu Hendradjaya “Theoretical
Considerations for Software Component Metrics”
[17] Magnus Andersson Patrik Vestergren “Object-Oriented Design
Quality Metrics “
[18] Saida, Benlarbi, Khaled El Emam, Nishith Goel “Issues in Vali- dating Object-Oriented Metrics for Early Risk Prediction”
IJSER © 2011 http://www.ijser.org
International Journal of Scientific & Engineering Research Volume 2, Issue 5, May-2011 6
ISSN 222S-5518
(http://books.google.com)
[19] Tieng Wei Koh, Mohd Hasan Selamat, Abdul Azim Abdul Gha ni, Rusli Abdullah "Review of Complexity Metrics for Object Oriented Software Products
[20] Marco Scotto, Alberto Sillitti, Giancarlo Succi, Tullio Vernazza
"A relational approach to software metrics"