Metamodel-Driven Evolution with Grammar Inference
Barrett R. Bryant, Qichao Liu, Marjan Mernik, Swapan Paruya, Samarjit Kar, Suchismita Roy · AIP conference proceedings · 2010
Domain‐specific modeling (DSM) has become one of the most popular techniques for incorporating model‐driven engineering (MDE) into software engineering. In DSM, domain experts define metamodels to describe the essential problems in a domain. A model conforms to a schema definition represented by a metamodel in a similar manner to a programming language conforms to a grammar. Metamodel‐driven evolution is when a metamodel undergoes evolutions to incorporate new concerns in the domain. However, this results in losing the ability to use existing model instances. Grammar inference is the problem of inferring a grammar from sample strings which the grammar should generate. This paper describes our work in solving the problem of metamodel‐driven evolution with grammar inference, by inferring the metamodel from model instances.