Title Details

Jean-Claude Royer is Full Professor at EMN, Nantes, France.

Hugo Arboleda is Associate Professor at University of Icesi, Columbia.

Chapter 1 Introduction 1

1.1 Software product line engineering 2

1.2 Model-driven engineering 3

1.3 Merging model-driven and software product line engineering 5

1.4 The FieSta framework 8

1.5 Book structure 11

Chapter 2 Software Product Line Engineering Basics 17

2.1 Introduction to product line engineering 17

2.2 Brief history 21

2.3 Application example: Smart-Home systems 24

2.3.1 Smart-Home system’s domain 24

2.3.2 Requirements of the application example 26

2.4 Software product line engineering 30

2.5 Domain engineering 34

2.5.1 Component-based software engineering 36

2.6 Variability management 37

2.6.1 Feature modeling 40

2.7 Application engineering 43

2.7.1 Product configuration 44

2.7.2 Product derivation 46

2.8 Benefits and drawbacks 48

2.9 Issues in product line 49

2.9.1 Variability management 50

2.9.2 Product derivation 50

2.9.3 Testing 51

2.9.4 Traceability 52

2.9.5 Product line evolution 53

2.9.6 Tool support 55

2.10 Summary 56

Chapter 3 Model-Driven Engineering 59

3.1 Introduction 59

3.2 Models and metamodels 60

3.2.1 The 4-level metamodeling framework 65

3.2.2 The nature of models 67

3.3 UML class diagrams and OCL 68

3.4 Model transformations 74

3.4.1 Scheduling of transformation rules 76

3.4.2 Model transformation patterns 78

3.4.3 Classification of model transformations 79

3.4.4 Vertical model transformations 80

3.4.5 Horizontal model transformations 81

3.4.6 Model composition or model weaving 81

3.5 Modeling framework 83

3.5.1 The eclipse modeling framework 83

3.5.2 The topcased toolkit 86

3.6 Model transformation languages 86

3.6.1 QVT 87

3.6.2 ATL 89

3.6.3 The openArchitectureWare framework 90

3.6.4 The Xtend language 92

3.7 Benefits and challenges for SPLE 96

3.8 Summary 98

Chapter 4 Model-Driven and Software Product Line Engineering 101

4.1 Introduction 102

4.2 Problem space issues 107

4.2.1 Separating points of views 107

4.2.2 Capturing variability and configuring products 108

4.2.3 Relating several points of view 109

4.2.4 Configuring products in a multi-staged process 110

4.3 Solution space issues 111

4.4 Developing core assets 112

4.4.1 Developing decision models and deriving products 112

4.5 Variability expression and product configuration 113

4.5.1 Metamodels 114

4.5.2 Feature models 120

4.6 Core asset development and product derivation 126

4.6.1 Transformation rules in the Smart-Home systems SPL 127

4.6.2 Creating and using decision models 132

4.7 Summary 138

Chapter 5 The FieSta Framework: Fine-Grained Derivation and Configuration 139

5.1 Introduction 139

5.1.1 Coarse-grained and fine-grained variations 140

5.2 Binding models and constraint models 142

5.2.1 Binding models 142

5.2.2 Constraint models 143

5.2.3 The cardinality property 146

5.2.4 The structural dependency property 147

5.2.5 The constraint metamodel and the binding metamodel 148

5.2.6 Validating binding models against constraint models 150

5.3.Deriving products based on constraint models and binding models 152

5.3.1 The extended decision metamodel 155

5.3.2 Creating executable model transformation workflows from decision models and constraint models 156

5.4 Identified limitations 157

5.4.1 Features combinatorial 157

5.4.2 Features interaction 158

5.4.3 Bindings interaction 159

5.5 Summary 160

Chapter 6 Tools Support 161

6.1 Introduction 161

6.2 The FieSta process 162

6.3 The SPL of Smart-Home systems 163

6.4 Variability expression and product configuration 170

6.4.1 MD-SPL project creation 170

6.4.2 Metamodels and feature models creation 170

6.4.3 Constraint models creation 173

6.4.4 Domain models and binding models creation 178

6.5 Completing and running the product derivation 184

6.5.1 Transformation rules creation 184

6.5.2 Decision models creation 186

6.5.3 Generation and execution of model transformation workflows 188

6.6 Summary 190

Chapter 7 A Second Comprehensive Application Example 191

7.1 Domain of the collection manager system 191

7.2 Requirements of the application example 192

7.2.1 Kernel commonalities 193

7.2.2 GUI commonalities 193

7.2.3 Kernel and GUI variability 193

7.3 The overall process 196

7.3.1 Domain engineering 196

7.3.2 Application engineering 197

7.4 Variability expression and product configuration 198

7.4.1 Metamodels 198

7.4.2 The feature model 202

7.4.3 The constraint model 204

7.4.4 Binding models 205

7.5 Core assets development and product derivation 207

7.5.1 Rule transformations in the SPL of the collection manager systems 207

7.5.2 Decision models 209

7.6 Summary 211

Chapter 8 Further Reading 213

8.1 Northop and Clements’ book 213

8.2 Pohl, Böckle and Van der Linden’s book 214

8.3 Gomaa’s book 214

8.4 Van der Linden, Schmid, and Rommes’ book 215

8.5 Stahl, Voelter, and Czarnecki book 216

8.6 AMPLE book 216

8.7 Feature modeling notations 218

8.8 Decision models 218

8.9 Model-driven software product lines 220

8.9.1 The Czarnecki and Antkiewicz’s approach 222

8.9.2 The Wagelaar’s approach 224

8.9.3 Loughran et al.’s approach 229

8.9.4 Voelter and Groher’s approach 232

8.9.5 Comparison table 235

8.10 Dynamic variability 236

8.11 Domain specific languages 238

8.12 Additional references 240

8.13 Summary 242

Chapter 9 Conclusion 243

9.1 Book summary 244

9.2 MD-SPL engineering 247

9.2.1 Metamodeling and feature modeling 248

9.2.2 Multi-staged configuration of products 249

9.2.3 Coarse and fine-grained variations and configurations 249

9.2.4 Core assets development and decision models 250

9.2.5 Product derivation 251

9.2.6 Comparison table 251

9.2.7 Perspectives 253

Bibliography 257

Index 271