A Lightweight Data Sharing System Based on Bidirectional Transformations.

Adrien Duchêne,Hugues Marchal,Zhenjiang Hu,Pierre-Yves Schobbens

DOI: https://doi.org/10.1145/3191697.3191722

2018-01-01

Abstract:Although the data sharing and synchronizing problems have been raised many years ago, they remain major issues in the database community. Still, some tools are provided to end-users in order to answer some of their needs. Yet, those platforms are most likely very complicated to handle notably because they ask the user to have very much knowledge, the user sometimes being the developer. Also, most of those systems do not really insure data consistency. Our approach based on bidirectional transformations (BXs) resolves collaboration between companies having their own data structure in an easier way, guaranteeing data consistency thanks to BXs. All this means that the user does not need to know databases structure other than his and the shared mappings, and will also never be asked to use pure code or database knowledge, limiting then the complexity. In addition to this, the system profits the bidirectional transformations properties to authorize or not editing the shared data. The bidirectional functions coded in BiGUL have indeed the power to grant or not any other user in the sharing group to edit the data. Moreover, the system is extensible in the way that the user can easily join a sharing group, after providing to the bidirectional functions a GLAV mapping table matching his local structure with the shared one.