Options
On the automation-supported derivation of domain-specific UML profiles considering static semantics
Kraas, Alexander (2021): On the automation-supported derivation of domain-specific UML profiles considering static semantics, in: Software and Systems Modeling, Berlin ; Heidelberg ; New York, NY, USA: Springer, Jg. 21, Nr. 1, S. 51–79, doi: 10.1007/s10270-021-00890-1.
Faculty/Chair:
Author:
Title of the Journal:
Software and Systems Modeling
ISSN:
1619-1374
1619-1366
Publisher Information:
Year of publication:
2021
Volume:
21
Issue:
1
Pages:
Language:
English
Abstract:
In the light of standardization, the model-driven engineering (MDE) is becoming increasingly important for the development of DSLs, in addition to traditional approaches based on grammar formalisms. Metamodels define the abstract syntax and static semantics of a DSL and can be created by using the language concepts of the Meta Object Facility (MOF) or by defining a UML profile.
Both metamodels and UML profiles are often provided for standardized DSLs, and the mappings of metamodels to UML profiles are usually specified informally in natural language, which also applies for the static semantics of metamodels and/or UML profiles, which has the disadvantage that ambiguities can occur, and that the static semantics must be manually translated into a machine-processable language.
To address these weaknesses, we propose a new automated approach for deriving a UML profile from the metamodel of a DSL. One novelty is that subsetting or redefining metaclass attributes are mapped to stereotype attributes whose values are computed at runtime via automatically created OCL expressions. The automatic transfer of the static semantics of a DSL to a UML profile is a further contribution of our approach. Our DSL Metamodeling and Derivation Toolchain (DSL-MeDeTo) implements all aspects of our proposed approach in Eclipse. This enabled us to successfully apply our approach to the two DSLs Test Description Language (TDL) and Specification and Description Language (SDL).
Both metamodels and UML profiles are often provided for standardized DSLs, and the mappings of metamodels to UML profiles are usually specified informally in natural language, which also applies for the static semantics of metamodels and/or UML profiles, which has the disadvantage that ambiguities can occur, and that the static semantics must be manually translated into a machine-processable language.
To address these weaknesses, we propose a new automated approach for deriving a UML profile from the metamodel of a DSL. One novelty is that subsetting or redefining metaclass attributes are mapped to stereotype attributes whose values are computed at runtime via automatically created OCL expressions. The automatic transfer of the static semantics of a DSL to a UML profile is a further contribution of our approach. Our DSL Metamodeling and Derivation Toolchain (DSL-MeDeTo) implements all aspects of our proposed approach in Eclipse. This enabled us to successfully apply our approach to the two DSLs Test Description Language (TDL) and Specification and Description Language (SDL).
GND Keywords: ;  ;  ; 
Domänenspezifische Programmiersprache
Modellgetriebene Entwicklung
UML
Metamodell
Keywords: ;  ;  ;  ; 
model-driven engineering (MDE)
DSL
Metamodels
UML profiles
machine-processable language
DDC Classification:
RVK Classification:
Peer Reviewed:
Yes:
International Distribution:
Yes:
Type:
Article
Activation date:
July 21, 2022
Versioning
Question on publication
Permalink
https://fis.uni-bamberg.de/handle/uniba/54846