ATL (ATLAS Transformation Language)

ATL (ATLAS Transformation Language)
Name	ATL (ATLAS Transformation Language)
Paradigm	Declarative, Model Transformation
Developer	ATLAS Group, INRIA
First appeared	2003
Typing	Static/Dynamic (model-driven)
License	Eclipse Public License

ATL (ATLAS Transformation Language) is a model transformation language designed for defining and executing model-to-model transformations within model-driven engineering environments. It was developed by a consortium including research institutions and industrial partners and is closely associated with model transformation standards promoted by organizations and projects in Europe and worldwide. ATL is used to transform structured models between different metamodels and to automate repetitive mapping tasks in software engineering projects.

Overview

ATL originated from collaborations among research entities and laboratories focused on model-driven engineering, and it has been promoted in international venues and workshops alongside projects and initiatives from institutions such as INRIA, Eclipse Foundation, OMG, European Commission, University of Grenoble Alpes, University of Rennes, École Normale Supérieure, IBM, Siemens, Thales and SAP. Its design reflects influences from earlier transformation efforts and standards that surfaced in conferences like ICSE, MODELS Conference, ETAPS, ECOOP, SLE Workshop and OOPSLA. ATL integrates with ecosystem tools that stem from collaborative projects funded by the European Union and research grants managed by national agencies like ANR and industrial consortia including Object Management Group partners.

Language Concepts and Syntax

ATL employs a hybrid declarative-imperative style inspired by academic work at institutions such as University of Twente, Technische Universität Berlin, Imperial College London, Politecnico di Milano and École Polytechnique. Its core concepts—rules, matched rules, lazy rules, helpers and modules—map instances of source metamodel elements to instances of target metamodel elements in a manner comparable to approaches discussed by researchers affiliated with University of Oxford, University of Cambridge, ETH Zurich, Stanford University and MIT. The syntax is concise and declarative, borrowing ideas related to pattern matching and template-based generation that were also explored at Carnegie Mellon University, University of Toronto, Harvard University and Princeton University. ATL distinguishes between in-place and out-place transformations, permitting transformations between metamodels such as those defined by UML, EMF, Ecore, SysML, BPMN and DSLs devised in academic and industrial settings like Cisco, Microsoft Research, Google Research and Oracle.

Execution and Tooling

Execution of ATL transformations typically occurs within integrated development environments and platforms that have been extended by tool providers and projects including Eclipse Foundation, EMF, Acceleo, Papyrus, QVT Operational, MagicDraw, IBM Rational and Sirius. Runtimes and engines for ATL have been implemented by teams linked to INRIA, LORIA, University of Rennes 1, CNR and IMT Atlantique, while continuous integration setups integrate with services and infrastructures from Jenkins, Travis CI, GitHub, GitLab and Azure DevOps. Profiling, debugging and visualization tools that support ATL use patterns and techniques also present in environments produced by JetBrains, Red Hat, Intel Corporation and Nokia.

Use Cases and Applications

ATL has been applied across scenarios investigated by research groups at TU Munich, RWTH Aachen, Delft University of Technology, University of Stuttgart and Chalmers University of Technology. Examples include transforming architectural models in projects with Airbus, Boeing, Renault, Volkswagen and Bosch; migrating legacy models in initiatives involving Siemens and ABB; and automating code generation chains used by teams at Ericsson, HP, Schneider Electric and Hitachi. ATL is also used in academic case studies comparing transformation approaches in work produced at University of Pennsylvania, Columbia University, University of California, Berkeley, National University of Singapore and Tsinghua University.

Design and Implementation

The language design reflects contributions from academic research groups at INRIA, University of Grenoble Alpes, University of Rennes 1, LORIA and industrial partners such as IBM and Eclipse Foundation members. Its implementation leverages the Eclipse Modeling Framework and components from projects maintained by teams at OBEO, Thales Research & Technology, CeaTech and other engineering organizations. The ATL virtual machine and compilers have been discussed in workshops and journal articles authored by researchers affiliated with Utrecht University, University of Twente, Katholieke Universiteit Leuven and KU Leuven.

Community and Adoption

ATL’s user and developer community spans academic groups, industry practitioners and open source contributors affiliated with institutions and companies like INRIA, Eclipse Foundation, IBM, Siemens, Thales, SAP, Airbus, Bosch and various universities cited above. Community interaction occurs in conferences such as MODELS Conference, SLE Workshop, ICSE and through mailing lists, code repositories and collaborative platforms hosted by organizations like GitHub, Eclipse Foundation and ResearchGate. Adoption has been documented in theses and reports produced at University of Pisa, Politecnico di Torino, University of Bologna, Università di Roma La Sapienza and Université Paris-Saclay.

Category:Model transformation languages