Analysing Microsoft Access Projects: Building a model in a Partially Observable Domain - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Analysing Microsoft Access Projects: Building a model in a Partially Observable Domain

Résumé

Due to the technology evolution, every IT Company migrates their software systems at least once. Reengineering tools build system models which are used for running software analysis. These models are traditionally built from source code analysis and information accessible by data extractors (that we call such information observable). In this article we present the case of Microsoft Access projects and how this kind of project is partially observable due to proprietary storing formats. We propose a novel approach for building models that allows us to overcome this problem by reverse engineering the development environment runtime through the usage of Microsoft COM interface. We validate our approach and implementation by fully replicating 10 projects, 8 of them industrial, based only on our model information. We measure the repli-cation performance by measuring the errors during the process and completeness of the product. We measure the replication error, by tracking replication operations. We used the scope and completeness measure to enact this error. Completeness is measured by the instrumentation of a simple and scoped diff based on a third source of information. We present extensive results and interpretations. We discuss the threats to validity, the possibility of other approaches and the technological restrictions of our solution.
Fichier principal
Vignette du fichier
article.icsr.pdf (804.64 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02966146 , version 1 (15-10-2020)

Identifiants

  • HAL Id : hal-02966146 , version 1

Citer

Santiago Bragagnolo, Nicolas Anquetil, Stéphane Ducasse, Abderrahmane Seriai, Mustapha Derras. Analysing Microsoft Access Projects: Building a model in a Partially Observable Domain. ICSR 2020, Dec 2020, Hammamet, Tunisia. ⟨hal-02966146⟩
88 Consultations
369 Téléchargements

Partager

Gmail Facebook X LinkedIn More