(3.235.139.152) 您好!臺灣時間:2021/05/11 05:16
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:李明錡
研究生(外文):Lee Ming-Chi
論文名稱:個體導向系統設計之測試方法及複雜度分析
論文名稱(外文):Software Complexity Measurement and object-Oriented System Design Testing
指導教授:莊淇銘莊淇銘引用關係
指導教授(外文):Chung Chi-Ming
學位類別:博士
校院名稱:淡江大學
系所名稱:資訊工程研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:1994
畢業學年度:82
語文別:中文
論文頁數:140
中文關鍵詞:個體導向系統軟體測試軟體衡量遺傳
外文關鍵詞:Object-Oriented SystemSoftware TestingSoftware MetricsInheritance
相關次數:
  • 被引用被引用:0
  • 點閱點閱:88
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1
在1992年,美國的軟體成本已高達G.N.P.之12%,且正每年逐漸增加,而
其中有一半是花費在軟體錯誤之測試及維護上。然而實際上軟體開發完成
後,其軟體錯誤仍持續增加中。因此如何降低軟體開發成本及增加軟體正
確性對軟體科學家是一項重要研究課題。軟體測試及軟體衡量方法是兩項
用來確保軟體品質及軟體正確性的重要技巧。然而它們之間卻存在許多問
題。在本篇論文中,我們提出一系列方法及演算法來解決這些問題。這些
問題可分為三部份:我們第一部份是----目前傳統的測試方法中均存在有
許多缺點。其中有兩項共同缺點分別是 1.)大部份軟體測試方法均無法衡
量巢狀結構複雜度 2.)只提供一純量衡量,即使很小的程式也有很大的複
雜度數量,實在很難反應真實複雜度。因此我們提出一多項式衡量方法(
Polynomial Metric )來改善這些缺點。第二部份的問題----是缺乏物件
導向程式語言衡量方法,自從物件導向程式語言這幾年被重視以來,它已
成為1990年代軟體發展的主流。然而傳統軟體衡量方法並不適用於它。因
此我們提出一組以 OOP為主的衡量方法來衡量 OOD系統的複雜度。第三部
份和第二部份很類似, 因為傳統的測試方法並不適用於物件導向系統,
因此我們提出一套測試方法來測試物件導向系統錯誤。此測試方法是利用
自動機理論( Finite State Machine )來設計。
Software testing and software metrics are two important
techniques toassure the software quality and software
correctness . However, they are related to many problems. In
this thesis, we propse a series of methodologies and algorithms
to solve the problems. These problems can divided into three
parts. The first part is that there existed many weaknesses in
conventional software metrics. They had two common weaknesses:
(1) most of them cannot measure the complexity of nested
control flow structure, and (2) they provide only a
quantitative measure, even for a small program, most measure
have a huge number. We propose a {\it Polynomial metric} to
improve the srawbacks of traditional software metrics. The
second part of the problems is concerned with the lack of
object-oriented programming software metrics. Since object-
oriented programming was noticed for the past few years, it has
been the major tendancy of software development, including
object-oriented analysis (OOA), and object-oriented design
(OOD). However, the procedure-orientedware metrics could not be
appropriate for object-oriented programming. Therefore, we
present a family of software metricssed on OOP to measure the
complexity of an OOD system. The third part of the problems is
similar to the second problem. Because, traditional testing
methodologies could not bee for object-oriented system. We
propose a series of testing methodologies to test the object-
oriented design system.
Cover
Contents
Abstract
1 Introduction
2 Survey of Software Metrics and Testing
2.1 Background of Software Metrics
2.1.1 Prograin Size Based Metrics
2.1.2 Control Flow Based Metrics
2.1.3 Data Flow Based Metrcis
2.2 Background of Software Testing
2.2.1 The Testing Hierarchies
2.2.2 Control Flow Based Testing Methodologies
2.2.3 Data Flow Based Testing Methodologies
2.2.4 White and Cohen''s Domain Testing Methodology
3 Polynomial Metric
3.1 liitroduction
3.2 Evaluating Software Complexity Measures
3.2.1 Introduction
3.2.2 Definition
3.2.3 Desirable Properties of Complexity Measures
3.3 Metric Design
3.4 Algorithm for Polynomial Metric
3.5 Comparison and Analysis
3.5.1 Analysis with Halstead''s Software Science
3.5.2 Analysis with McCabe''s Measure
3.5.3 Analysis with Polynomial Metric
3.5.4 Compound Predicate Can Reduce Software Complexity
3.5.5 Evaluate Polynomial Metric by Weyuker''s Properties
3.6 Experiments
4 Object-Oriented Design Complexity Measurement
4.1 Introduction
4.2 Towards a Metrics suite for Object-Oriented Design
4.3 A family of Inheritance-Based Metrics
4.4 Restrict Inheritance-Based Metric to Weyuker''s Properties
4.5 URI Metrics
4.5.1 Definition and Background
4.5.2 Algorithm for Finding URIs
4.5.3 Metric Design
5 Object-Oriented Design Testing
5.1 Testing Object-Oriented Design Modeled by Finite State Machine
5.1.1 Introduction
5.1.2 Classification of Object-oriented Design
5.1.3 Object-Chart Design
5.1.4 Testing OOD modeled by Finite State Machine
5.2 Redundant Inheritance Detection for OOD by Transitive Closure Matrix
5.2.1 Introduction
5.2.2 Redundant Inheritance and Relative Works
5.2.3 Acyclic Inheritance and Transitive Closure Matrix
5.2.4 Redundant Inheritance Elimination
5.2.5 Summary
6 Conclusion and Future Researches
6.1 Conclusions and Contributions
6.2 Future Researches
Reference
Appendix
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔