Configuration Generation

#Configuration Generation This page explains how the configurations are generated to test the transformation depending on the framed features. This solution is a simplified method inspired by [1]. It uses requirements that can be seen in the table Requirements below. The requirements are tied to a configuration of its dependant feature. Features on which the require dont depend are not shown in the table. An algorithm goes through a list of the requirements and constructs complete configurations of requirements that dont contradict each other. This is done until every requirement is used once. The algorithm uses a set of a reduced number of features. This list can be seen in the table Reduced feature list. The value of the column Index indicates the position of the feature in the string representation which is used in the source code.

##Reduced feature list

Index	Feature
0	Role_Properties/ Role_Behavior
1	Role_Inheritance
2	Compartments
3	Dates
4	Role_Implication
5	Role_Prohibtion
6	Role_Equivalence
7	Group_Constraints
8	Occurence_Constraints
9	Relationship
10	Relationship_Cardinality
11	Intra_Relationship_Constraints
12	Inter_Relationship_Constraints
13	Compartment_Types
14	Compartment_Properties/ Compartment_Behavior
15	Compartment_Inheritance
16	Playable_By_Defining_Compartment
17	Data_Types
18	Data_Type_Inheritance

##Requirements

Requirement	Description	Dependat Configuration
R1	Role properties and operation are transformed	Role_Properties/ Role_Behavior
R2	Role properties and operation are not transformed	!(Role_Properties/ Role_Behavior)

[1] K.D. Scheidemann, Optimizing the selection of representative configurations in verification of evolving product lines of distributed embedded systems, in: Proceedings of the 10th International Software Product Line Conference (SPLC’06), 2006, pp. 75–84.