Design And Implementation Of The Software Product Line Variability Modeling Method

Software product line is an effective, systematic method of software reuse. It’svariability copes with the product’s unique functions and specific requirements, throughthe reuse of pre-defined, configurable components to customize the software. It’shelpful to improve software reuse, reduce costs, shorten development time and improvesoftware quality. So the variability modeling is very significant.The thesis’emphasis is to analyse and research the architecture and variability of the softwareproduct line, combine the variability theory with UML modeling, use the model framework of EMF,GEF and GMF, through Profile weak extension and MOF strong extension mechanism to design andrealize the variability modeling of software product line. This paper’s main contents include thefollowing aspects:1. Analysing and researching the architecture and variability of the softwareproduct line. identifing and defining systematically the reality system, then analyzingthe variability’s variable_points and variables, and the relationship between them.2. Combining the variability theory with UML modeling, to establish thevariability modeling extension theory based on UML, and define the all meta-models invariability modeling process.3. Studying the variability Profile weak extension and initially realizing variabilityProfile modeling. Using UML standard meta-model, combining with variabilitymodeling theory, establishing variability Profile view, and quoting the variability Profilein other UML views, to realize the corresponding variability modeling of UML views.4. Designing and realizing variability modeling based on the MOF strongextension mechanism. First, based on the MOF meta-metamodel, combining withvariability modeling theory, using the EMF model framework to design and define theindependent variability meta-model. Second, through GMF modeling framework torealize variability modeling, and using the GEF graphics framework and plug-intechnology to extend and perfect the variability modeling tools.