Toward Bridging the Gap between Formal Semantics and Implementation of Triple Graph Grammars

Abstract

Bibliografische Information der Deutschen NationalbibliothekDie Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar. Abstract. The correctness of model transformations is a crucial element for the model-driven engineering of high quality software. A prerequisite to verify model transformations at the level of the model transformation specification is that an unambiguous formal semantics exists and that the employed implementation of the model transformation language adheres to this semantics. However, for existing relational model transformation approaches it is usually not really clear under which constraints particular implementations are really conform to the formal semantics. In this paper, we will bridge this gap for the formal semantics of triple graph grammars (TGG) and an existing efficient implementation. Whereas the formal semantics assumes backtracking and ignores non-determinism, practical implementations do not support backtracking, require rule sets that ensure determinism, and include further optimizations. Therefore, we capture how the considered TGG implementation realizes the transformation by means of operational rules, define required criteria and show conformance to the formal semantics if these criteria are fulfilled. We further outline how static analysis can be employed to guarantee these criteria. Universitätsverlag

A Static Analysis of Non-confluent Triple Graph Grammars for Efficient Model Transformation

“…All other critical pairs (conflicts) must either be manually checked to be confluent, or removed by adjusting the TGG rules as required for confluence. More restrictively, Giese et al [6] require confluence without filtering backtracking paths automatically. Although OCL constraints can be used to resolve critical pairs, this is a manual process required for both forward and backward transformations and it remains unclear how such OCL constraints must be restricted to guarantee formal properties.…”

“…Although confluence avoids backtracking (i.e., is used to show efficiency), solves completeness problems, and can be statically checked (cf. [8,6,7]), it can be too restrictive in practical scenarios (as in our running example) as it forces a TGG to be a bijection (a function in both directions).…”

Bidirectional Model Transformation with Precedence Triple Graph Grammars

“…(I.b) Functional Behavior: Demanding functional behavior [GHL10,HGO10] guarantees that the algorithm can choose freely between applicable rules at every decision point and will always get the same result without backtracking. Although functional behavior might be suitable for fully automatic integrations, our experience with industrial partners [RLSS11] shows that user interaction or similar guidance (e.g., configuration files) of the integration process is required and leads naturally to non-functional sets of rules with certain degrees of freedom.…”

“…This is especially the case for bidirectional model transformation, where defining a precise semantics for the automatic manipulation and synchronization of models with a corresponding efficient tool support is quite challenging [CFH + 09]. Amongst the numerous bidirectional model transformation approaches surveyed in [Ste08], the concept of Triple Graph Grammars (TGGs) features not only solid formal foundations [EEE + 07, KLKS10] but also various tool implementations [GHL10,KRW04,KLKS10].…”

“…If supported TGGs are required to exhibit functional behaviour [GHL10,HGO10], which can be enforced statically via a critical pair analysis [EEPT06], the control algorithm cannot make wrong choices enabling an efficient implementation. However, as shown in [K05,RLSS11], transformations required for real-world applications are often non-functional.…”