DOCSLIB.ORG

Lessons Learned from Building Model-Driven Development Tools

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Lessons Learned from Building Model-Driven Development Tools Noname manuscript No. (will be inserted by the editor) Lessons Learned from Building Model-Driven Development Tools Richard F. Paige1 and D´anielVarr´o2 Department of Computer Science, University of York, UK. e-mail: [email protected] Department of Measurement and Information Systems Budapest University of Technology and Economics, Hungary. e-mail: [email protected] The date of receipt and acceptance will be inserted by the editor Abstract Tools to support modelling in system and of MDD tools. The principles will be distilled from an software engineering are widespread, and have reached analysis of the development of two MDD tools that are a degree of maturity where their use and availability used in practice: VIATRA and Epsilon. These tools are are accepted. Tools to support Model-Driven Develop- used in industry, on real projects, and have developed ment (MDD) { where models are manipulated and man- in very different ways. Arguably, some of the lessons aged throughout the system/software engineering lifecy- learned from the development of these tools can inform cle { have, over the last ten years, seen much research the development of new MDD tools, and can also be and development attention. Over the last ten years, we used to support the evolution of existing tools. Our ob- have had significant experience in the design, develop- jective is not to propose the \ideal" MDD tool; such a ment and deployment of MDD tools in practical settings. tool is unlikely to exist. However, the principles under- In this paper, we distill some of the important lessons pinning the development of MDD tools that are widely we have learned in developing and deploying two MDD used in practice can help in assessing existing tools, and tools: Epsilon and VIATRA. In doing so, we aim to iden- in improving them. tify some of the key principles of developing successful To elicit and present the lessons learned, the paper MDD tools, as well as some hints of the pitfalls and risks. will summarise the development of Epsilon and VIA- TRA, starting from initial use cases and system require- ments, leading to an initial tool architecture. Evolution- ary steps (e.g., refactorings) will be discussed and pre- 1 Introduction sented, along with the triggers { which included both re- quests for new functionality, as well as requests related to The last ten years have seen the development of nu- non-functional characteristics such as performance and merous tools for supporting Model-Driven Development scalability { that led to these refactorings. (MDD): the manipulation and analysis of structured de- The paper is structured as follows. Section 2 de- scriptions. MDD tools such as ATL [23], xText [42], Ep- scribes the development and evolution of Epsilon and silon [16], VIATRA [5], ATOM3 [15], KerMeta [29], Ac- VIATRA, focusing on initial motivation and scenarios, celeo [1], FUJABA [13], MOFLON [3], GReAT [2] and as well as some of the most important evolutionary steps. many others have been developed and deployed in a vari- Section 3 distills some of the key take-home messages ety of software and systems engineering contexts. Many from these developments. We conclude in Section 4. of these tools are the result of research and development on specific industrial use cases; others are the result of theoretical and conceptual research. Some are now be- 2 Evolution of MDD Tools ing applied on large-scale software engineering projects and are being turned into products. As such, the field of This section presents the development and evolution of MDD tool development has reached a sufficient level of two MDD tools: Epsilon and VIATRA. Both Epsilon and maturity for its results to be assessed, and distillation of VIATRA are predominantly textual tools, featuring tex- lessons learned from the development of these tools can tual languages for specifying and executing MDD tasks. take place. Each part describes the tool as it currently exists, then This paper aims to identify some of the principles talks about its development cycle. The development of underpinning the design, implementation and evolution Epsilon starts with an initial use case, and then moves to 2 Paige and Varr´o a discussion of refactoring stages, including both techni- [35]) and unit testing of model management operations cal refactorings (i.e., to improve performance) and con- (EUnit). ceptual refactorings. The development of VIATRA starts The core language in Epsilon is EOL; all other lan- from model transformation requirements, including sig- guages in Epsilon reuse EOL in some way, as it pro- nificant requirements for performance, and then broad- vides the conceptually common features required for ma- ens its scope to include other MDD tasks. nipulating and analysing models. In particular, it sup- ports navigation of models (via OCL-like expressions and queries), modification of models (via assignment 2.1 The design and evolution of Epsilon statements), and multiple-model access capabilities. How- ever, EOL was { at least conceptually { not the first lan- guage to be developed; it emerged from the development 2.1.1 Epsilon Epsilon [16] is a tool for model manage- of a different Epsilon language, as we now discuss. ment, specifically for analysing and manipulating models in specific ways. More precisely, it is a set of textual task- 2.1.2 Initial use case and architecture The use case that specific model management languages, and a framework triggered the development of Epsilon was one of model for implementing new model management languages by merging. Model merging is the process of combining two exploiting the existing ones. It is a component of the or more source models (possibly from different modelling Eclipse Model Framework Technology (EMFT) project. languages) into a target model (which may conform to From a user's perspective, Epsilon provides a set of a completely different metamodel than the source mod- inter-related and largely interoperable languages that els). The model merging scenarios that were of interest can be used to implement and execute various model were two-fold: management tasks. The current architecture of Epsilon is illustrated in Figure 1. { support for version control on models, including com- bining different versions of models, identifying differ- ences, etc. { support for merging behavioural models (e.g., state Model Migration (Flock) Unit Testing (EUnit) machines, sequence diagrams). Our research suggested that the process of model merg- Model Refactoring (EWL) Model Validation (EVL) ing could be separated into four stages: 1.A comparison phase, where correspondences between Model Comparison (ECL) Model Merging (EML) equivalent elements of source models are identified; 2.A conformance checking phase, where corresponding M2M Transformation (ETL) M2T Transformation (EGL) elements identified in the previous phase are exam- ined to identify conflicts that may render merging impossible. This phase was particularly important Epsilon Object Language for version control, i.e., when merging models of the same metamodel. Epsilon Model Connectivity 3.A merging phase, where corresponding and conform- ing elements are combined; and EMF driver MDR driver XML driver Z driver 4.A reconciliation phase, where any inconsistencies in- troduced in the merging phase are resolved. What is revealing about this separation is that each Fig. 1 Epsilon model management platform phase { comparison, `checking', merging { is an oper- ation that we might want to apply to models. Indeed, it was this observation that suggested that model merging Epsilon consists of a set of textual languages (each itself was a composite operation that can be applied to including parsers, editing tools, and interpreters/virtual models { that is, merging was constructed from other machines), a connectivity framework (more on that later), model management operations. and some additional tools to help ease development. Each This raised the question what did model management language has further development tools (e.g., syntax- operations have in common? To this end, we investi- aware editors). Each language aims to support a par- gated model transformation, merging, constraint check- ticular model management task. More specifically, there ing (e.g., via OCL) and model-to-text transformation, is a language for direct manipulation of models (EOL) and identified a small set of features that all these oper- [26], as well as languages for model merging (EML) [24], ations have in common: model comparison (ECL) [25], model-to-model transfor- mation (ETL) [27], model validation (EVL) [28], model- { navigating models; to-text transformation (EGL) [36], model migration (Flock) { accessing multiple models simultaneously; Lessons Learned from Building Model-Driven Development Tools 3 { evaluating expressions on models; 2.1.3 Conceptual refactoring After designing and im- { modifying models plement EOL and EML, and fulfilling our initial require- We also noticed that different MDD tools tended to ments and use cases, we took a step back and studied reimplement many of these common features using non- the core features of EML. In particular, we observed two standard syntaxes, thus rendering interoperability of tools recurring patterns in the logic of the merging programs more difficult (in part because syntax transformation that we were writing in EML: would have to take place) We argued [26] that there was { Significant parts of the EML programs focused on substantial value in encapsulating these features so that writing the expressions that would compare the mod- they could be reused in different operations that required els. In many places, these expressions would be re- them. This was the genesis of the core language of Ep- peated. A reusable operation concept was added to silon { EOL { and the conceptual architecture depicted EOL to simplify EML programs, but even with this in Figure 1. there was still substantial opportunities for reuse of In parallel with this conceptual process, we obtained comparison logic that was not supported by EOL (or (from our collaborators who provided the initial use case) EML).
Load more

Recommended publications
  • The Model Transformation Language of the VIATRA2 Framework
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Science of Computer Programming 68 (2007) 214–234 www.elsevier.com/locate/scico The model transformation language of the VIATRA2 framework Daniel´ Varro´ ∗, Andras´ Balogh Budapest University of Technology and Economics, Department of Measurement and Information Systems, H-1117, Magyar tudosok krt. 2., Budapest, Hungary OptXware Research and Development LLC, Budapest, Hungary Received 15 August 2006; received in revised form 17 April 2007; accepted 14 May 2007 Available online 5 July 2007 Abstract We present the model transformation language of the VIATRA2 framework, which provides a rule- and pattern-based transformation language for manipulating graph models by combining graph transformation and abstract state machines into a single specification paradigm. This language offers advanced constructs for querying (e.g. recursive graph patterns) and manipulating models (e.g. generic transformation and meta-transformation rules) in unidirectional model transformations frequently used in formal model analysis to carry out powerful abstractions. c 2007 Elsevier B.V. All rights reserved. Keywords: Model transformation; Graph transformation; Abstract state machines 1. Introduction The crucial role of model transformation (MT) languages and tools for the overall success of model-driven system development has been revealed in many surveys and papers during the recent years. To provide a standardized support for capturing queries, views and transformations between modeling languages defined by their standard MOF metamodels, the Object Management Group (OMG) has recently issued the QVT standard. QVT provides an intuitive, pattern-based, bidirectional model transformation language, which is especially useful for synchronization kind of transformations between semantically equivalent modeling languages.
    [Show full text]
  • Model-Driven Engineering, the System Under Development Is Described and Analyzed Using Models
    X perf =1.00 X loss =0.01 SDSoftware Design and Quality A Declarative Language for Bidirectional Model Consistency Master’s Thesis of Dominik Werle at the Department of Informatics Institute for Program Structures and Data Organization (IPD) Reviewer: Prof. Dr. Ralf H. Reussner Second reviewer: Jun.-Prof. Dr.-Ing. Anne Koziolek Advisor: Dipl.-Inform. Max E. Kramer Second advisor: M.Sc. Michael Langhammer 09. October 2015 – 08. April 2016 Karlsruher Institut für Technologie Fakultät für Informatik Postfach 6980 76128 Karlsruhe I declare that I have developed and written the enclosed thesis completely by myself, and have not used sources or means without declaration in the text. Karlsruhe, 08. April 2016 .................................... (Dominik Werle) Abstract In model-driven engineering, the system under development is described and analyzed using models. Dierent models provide dierent abstractions of the system for dierent purposes. If multiple modeling languages are used, their models can contain overlapping informa- tion about the system. Then, they can become inconsistent after changes and need to be synchronized. The complexity of synchronizing models can be reduced by specifying consistency relationships in specialized languages and by automating the synchronization based on this specication. The languages can hide complexity that is not specic to the domain of the modeling languages, for example by deriving the operations needed for propagating changes from a model to another model for all pairs of models instead of requiring explicit specication for each pair and direction. In the course of this thesis, we designed the mapping language for specifying these consistency relationships and implemented a framework that maintains consistency based on the specication.
    [Show full text]
  • Research Report
    RZ 3621 (# 99631) 07/18/2005 Computer Science 10 pages Research Report A Systematic Approach to Designing Model Transformations Jochen M. Kuster,¨ Ksenia Ryndina and Rainer Hauser IBM Research GmbH Zurich Research Laboratory 8803 Ruschlikon¨ Switzerland {jku,ryn,rfh}@zurich.ibm.com LIMITED DISTRIBUTION NOTICE This report will be distributed outside of IBM up to one year after the IBM publication date. Some reports are available at http://domino.watson.ibm.com/library/Cyberdig.nsf/home. Research IBM Almaden · Austin · Beijing · Delhi · Haifa · T.J. Watson · Tokyo · Zurich A Systematic Approach to Designing Model Transformations Jochen M. Kuster¨ , Ksenia Ryndina and Rainer Hauser IBM Zurich Research Laboratory 8803 R¨uschlikon Switzerland jku,ryn,rfh ¡ @zurich.ibm.com Abstract forming them into a formal language [9]. Currently, a great deal of research is focused on the expression of model transformation Design and implementation of model transformations is one designs and appropriate tool support for model transformation de- key prerequisite for the vision of model driven development to be- velopers. This has led to the Query/Views/Transformation (QVT) come true. However, model transformations are quite different proposal by the Object Management Group (OMG) [21], which from other software artifacts since a model transformation design aims to provide a standardized framework for model transforma- typically consists of a set of transformation rules instead of object- tion development. oriented models such as class diagrams and statecharts. There- fore traditional software engineering approaches are not directly Many existing model transformation approaches favor that applicable for the development of model transformations. Never- transformations are captured in a rule-based way by a set of trans- theless, in the line of existing software engineering practice, model formation rules (see VIATRA [5], GReAT [13], UMLX [26], transformations must be developed in a systematic way in order to BOTL [4] and work by Milicev [17]).
    [Show full text]
  • Expressive and Efficient Model Transformation with an Internal DSL
    Expressive and Efficient Model Transformation with an Internal DSL of Xtend Artur Boronat Department of Informatics, University of Leicester, UK [email protected] ABSTRACT 1 INTRODUCTION Model transformation (MT) of very large models (VLMs), with mil- Over the past decade the application of MDE technology in indus- lions of elements, is a challenging cornerstone for applying Model- try has led to the emergence of very large models (VLMs), with Driven Engineering (MDE) technology in industry. Recent research millions of elements, that may need to be updated using model efforts that tackle this problem have been directed at distributing transformation (MT). This situation is characterized by important MT on the Cloud, either directly, by managing clusters explicitly, or challenges [26], starting from model persistence using XMI serial- indirectly, via external NoSQL data stores. In this paper, we draw at- ization with the Eclipse Modeling Framework (EMF) [31] to scalable tention back to improving efficiency of model transformations that approaches to apply model updates. use EMF natively and that run on non-distributed environments, An important research effort to tackle these challenges was showing that substantial performance gains can still be reaped on performed in the European project MONDO [19] by scaling out that ground. efficient model query and MT engines, such as VIATRA[3] and We present Yet Another Model Transformation Language (YAMTL), ATL/EMFTVM [46], building on Cloud technology either directly, a new internal domain-specific language (DSL) of Xtend for defin- by using distributed clusters of servers explicitly, or indirectly, by ing declarative MT, and its execution engine.
    [Show full text]
  • Formal Validation and Model Generationfor Domain-Specific
    Budapest University of Technology and Economics Faculty of Electrical Engineering and Informatics Department of Measurement and Information Systems Formal Validation and Model Generation for Domain-Specific Languages by Logic Solvers Thesis booklet Oszkár Semeráth Thesis supervisor: Prof. Dániel Varró Budapest, 2019 1. Introduction 1 Introduction 1.1 Domain-specific modeling languages My thesis is motivated by the challenges in the development of complex, safety- critical systems such as automotive, avionics or cyber-physical systems, which is characterized by a long development time and strict safety standards (like DO-178C or DO-330). Model-Based System Engineering (MBSE) is a widely used technique in those application domains [WHR14], which facilitates the use of models in different phases of design and on various levels of abstraction. Fur- thermore, MBSE promotes the use of Domain-Specific (Modeling) Languages (DSLs) to precisely capture the main features of a target domain, thus en- abling the engineers to model their solutions with the concepts of the problem domain. Additionally, advanced modeling environments can automate several development steps, with a particular emphasis on verification and validation (V&V). Thus they can significantly improve the overall productivity of the de- velopment and quality of the product. A complex industrial modeling environment supports the development of models by continuously evaluating consistency constraints to ensure that the models satisfy the design rules of the domain. There is already efficient sup- port for automatically validating constraints and design rules over large model instances of the DSL using tools like Eclipse OCL [Wil12; KPP09] or Viatra [Ber+11; Ber+10]. Modeling environments often incorporate the automated generation of various artifacts such as source code, task specific views, or doc- umentation by using code generation or model transformation techniques.
    [Show full text]
  • Efficient ATL Incremental Transformations
    Journal of Object Technology Published by AITO | Association Internationale pour les Technologies Objets http://www.jot.fm/ Efficient ATL Incremental Transformations Th´eoLe Calvara Fr´ed´ericJouaultb Fabien Chhelb Mickael Clavreulb a. LERIA, University of Angers, France b. ERIS Team, Groupe ESEO, France Abstract Incrementally executing model transformations offers several benefits such as updating target models in-place (instead of creating a new copy), as well as generally propagating changes faster (compared with complete re-execution). Active operations have been shown to of- fer performant OCL-based model transformation incrementality with use- ful properties like fine-grained change propagation, and the preservation of collection ordering. However, active operations have so far only been available as a Java library. This compels users to program at a relatively low level of abstraction, where most technical details are still present. Writing transformations at this level of abstraction is a tedious and error prone work. Using languages like Xtend alleviates some but not all issues. In order to provide active operation users with a more user-friendly front-end, we have worked on compiling ATL code to Java code using the active operations library. Our compiler can handle a significant subset of ATL, and we show that the code it generates provides similar performance to hand-written Java or Xtend code. Furthermore, this compiler also enables new possibilities like defining derived properties by leveraging the ATL refining mode. Keywords Incremental model transformation Active operations ATL. 1 Introduction Incremental model transformation execution offers many advantages over batch (i.e., non-incremental) execution, such as faster change propagation.
    [Show full text]
  • The Train Benchmark: Cross-Technology Performance Evaluation of Continuous Model Queries
    Softw Syst Model DOI 10.1007/s10270-016-0571-8 REGULAR PAPER The Train Benchmark: cross-technology performance evaluation of continuous model queries Gábor Szárnyas1,2,3 · Benedek Izsó1 · István Ráth1,4 · Dániel Varró1,2,3 Received: 30 October 2015 / Revised: 27 March 2016 / Accepted: 3 October 2016 © The Author(s) 2017. This article is published with open access at Springerlink.com Abstract In model-driven development of safety-critical tion rules is challenging for the currently available tools. systems (like automotive, avionics or railways), well- Checking well-formedness constraints can be captured by formedness of models is repeatedly validated in order to declarative queries over graph models, while model update detect design flaws as early as possible. In many indus- operations can be specified as model transformations. This trial tools, validation rules are still often implemented by paper presents a benchmark for systematically assessing the a large amount of imperative model traversal code which scalability of validating and revalidating well-formedness makes those rule implementations complicated and hard constraints over large graph models. The benchmark defines to maintain. Additionally, as models are rapidly increas- well-formedness validation scenarios in the railway domain: ing in size and complexity, efficient execution of valida- a metamodel, an instance model generator and a set of well- formedness constraints captured by queries, fault injection Communicated by Prof. Lionel Briand. and repair operations (imitating the work of systems engi- neers by model transformations). The benchmark focuses This work was partially supported by the MONDO (EU ICT-611125) project, the MTA-BME Lendület Research Group on Cyber-Physical on the performance of query evaluation, i.e.
    [Show full text]
  • Model Transformation by Graph Transformation: a Comparative Study
    Model Transformation by Graph Transformation: A Comparative Study Gabriele Taentzer1, Karsten Ehrig1, Esther Guerra2, Juan de Lara3, Laszlo Lengyel4, Tihamer Levendovszky4, Ulrike Prange1, Daniel Varro4, and Szilvia Varro-Gyapay4 1 Technische Universit¨atBerlin, Germany {karstene,gabi,ullip}@cs.tu-berlin.de 2 Universidad Carlos III de Madrid, Spain [email protected] 3 Universidad Autonoma de Madrid, Spain [email protected] 4 Budapest University of Technology and Economics, Hungary {varro,gyapay}@mit.bme.hu {lengyel,tihamer}@aut.bme.hu Abstract. Graph transformation has been widely used for expressing model transformations. Especially transformations of visual models can be naturally formulated by graph transformations, since graphs are well suited to describe the underlying structures of models. Based on a com- mon sample model transformation, four different model transformation approaches are presented which all perform graph transformations. At first, a basic solution is presented and crucial points of model transforma- tions are indicated. Subsequent solutions focus mainly on the indicated problems. Finally, a first comparison of the chosen approaches to model transformation is presented where the main ingredients of each approach are summarized. 1 Introduction Raising the abstraction level from textual programming languages to visual mod- eling languages, model transformation techniques and tools have become more focused recently. Model transformation problems can be formulated as graph transformation problems, thus, a variety of tools choose this technique as the underlying mechanism for the transformation engine. This paper aims at com- paring four approaches to model transformation that apply graph transforma- tion techniques for model transformation. These can be characterized by their tool support. The presented tools are AGG [1], AToM3 [15], VIATRA2 [25] and VMTS [26] [16].
    [Show full text]
  • Model Transformations
    Model Transformations Modelling with UML, with semantics 271 What is a transformation? • A transformation is the automatic generation of a target model from a source model, according to a transformation definition. • A transformation definition is a set of transformation rules that together describe how a model in the source language can be transformed into a model in the target language. • A transformation rule is a description of how one or more constructs in the source language can be transformed into one or more constructs in the target language. • Unambiguous specifications of the way that (part of) one model can be used to create (part of) another model • Preferred characteristics of transformations • semantics-preserving Modelling with UML, with semantics 272 Model-to-model vs. Model-to-code • Model-to-model transformations Model ModelModel Meta-model • Transformations may be between different languages. In particular, between different languages defined by MOF Transformation Transformer Rules • Model-to-text transformations Model Meta-model • Special kind of model to model transformations optional, can be repeated • MDA TS to Grammar TS Code Transformer Generation Templates Manually Generated Code Written Code optional Modelling with UML, with semantics 273 Transformations as models • Treating everything as a model leads not only to conceptual simplicity and regular architecture, but also to implementation efficiency. • An implementation of a transformation language can be composed of a transformation virtual machine plus a metamodel-driven compiler. • The transformation VM allows uniform access to model and metamodel elements. MMa MMt MMb Ma Mt Mb Transformation Virtual Machine Modelling with UML, with semantics 274 Model transformation • Each model conforms to a metamodel.
    [Show full text]
  • Model Driven Software Engineering for Web Applications
    Model Driven Software Engineering for Web Applications Andreas Kraus Dissertation zur Erlangung des akademischen Grades des Doktors der Naturwissenschaften an der Fakultät für Mathematik, Informatik und Statistik der Ludwig-Maximilians-Universität München vorgelegt am 23.04.2007 Model Driven Software Engineering for Web Applications Tag der Einreichung: 23.04.2007 Tag des Rigorosums: 04.07.2007 Berichterstatter Prof. Dr. Rolf Hennicker (Ludwig-Maximilians-Universität, München) Prof. Dr. Antonio Vallecillo (Universidad de Malaga, Spanien) 2 Model Driven Software Engineering for Web Applications Summary Model driven software engineering (MDSE) is becoming a widely accepted approach for developing complex applications and it is on its way to be one of the most promising para- digms in software engineering. MDSE advocates the use of models as the key artifacts in all phases of the development process, from analysis to design, implementation and testing. The most promising approach to model driven engineering is the Model Driven Architec- ture (MDA) defined by the Object Management Group (OMG). Applications are modeled at a platform independent level and are transformed to (possibly several) platform specific implementations. Model driven Web engineering (MDWE) is the application of model driven engineering to the domain of Web application development where it might be par- ticularly helpful because of the continuous evolution of Web technologies and platforms. However, most current approaches for MDWE provide only a partial application of the MDA pattern. Further, metamodels and transformations are not always made explicit and metamodels are often too general or do not contain sufficient information for the automatic code generation. Thus, the main goal of this work is the complete application of the MDA pattern to the Web application domain from analysis to the generated implementation, with transformations playing an important role at every stage of the development process.
    [Show full text]
  • Domain Specific Language Tool Development
    University of Latvia ELĪNA KALNIŅA MODEL TRANSFORMATION DEVELOPMENT USING MOLA MAPPINGS AND TEMPLATE MOLA Thesis for the PhD Degree at the University of Latvia Field: Computer Science Section: Programming Languages and Systems Scientific Advisor: Prof., Dr. Habil. Sc. Comp. AUDRIS KALNINS Riga – 2011 This work has been supported by the European Social Fund within the project «Support for Doctoral Studies at University of Latvia». Scientific Advisor: Professor, Dr. Sc. Comp. Audris Kalniņš Latvijas Universitāte Referees: Professor, Dr. Sc. Comp. Guntis Bārzdiņš University of Latvia Professor, Dr. Sc. Ing. Oksana Ņikiforova Riga Technical University Professor, Dr. Olegas Vasilecas Vilnius Gediminas Technical University (Vilnius, Lithuania) The defence of the thesis will take place in an open session of the Council of Promotion in Computer Science of the University of Latvia, in the Institute of Mathematics and Computer Science of the University of Latvia (Room 413, Raina Boulevard 29, Riga, Latvia) on March 7, 2012 at 4 PM. The thesis and its summary are available at the library of the University of Latvia (Kalpaka Boulevard 4, Riga, Latvia). 3 4 ABSTRACT Model transformation development for three specific domains: Model-Driven Software Development (MDSD), DSL tool development and transformation synthesis has been studied in the thesis. It is concluded that transformation development in domain- specific transformation languages is more straightforward and faster compared to traditional transformation languages. A domain-specific model transformation language has been developed for each studied domain. Two of them are based on mappings. In both cases it was concluded that mappings better fit for typical tasks and transformations better fit for non-standard tasks.
    [Show full text]
  • Visual Models Transformation in Metalanguage System Alexander O
    Advances in Information Science and Applications - Volume II Visual Models Transformation in MetaLanguage System Alexander O. Sukhov, Lyudmila N. Lyadova (DSMLs, DSLs), intended to solve a particular class of Abstract — At the process of creation and maintenance of problems in the specific domain. Unlike general purpose information systems the model-based approach to the software modeling languages, DSMLs are more expressive, simple in development is increasingly used. This approach allows to move the applying and easy to understand for different categories of focus from writing of the program code with using general purpose users as they operate with domain terms. To support the language to the models development with automatic generation of data structures and source code of applications. However at usage of process of development and maintenance of DSMLs the this approach it is necessary to transform models constructed by special type of software – language workbench (DSM- various categories of users at different stages of system creation with platform) – is used. usage of various modeling languages. An approach to models The various categories of specialists (programmers, system transformation in DSM platform MetaLanguage is considered. This analysts, database designers, domain experts, business approach allows fulfilling vertical and horizontal transformations of analysts, etc.) are involved in the process of information the designed models. The Metalanguage system support “model-text” and “model-model” types of transformations. The component of systems creation and maintenance. Often they need transformations is based on graph grammars described by production modification of modeling language description to customize rules. Transformations of model in Entity-Relationship notation are and adapt DSML to new conditions, requests of business and presented as example.
    [Show full text]