Abstract
The article describes a solution to a problem of integration of distributed, heterogeneous and variously fragmented collections of data located in object-oriented databases. Basing on their investigation, authors present three methods of integration procedures for differently structured databases constituting virtual repositories in a data grid with a P2P architecture. The core of the described architecture is based on the Stack-Based Query Language (SBQL) and virtual updatable SBQL views. In the background, the system is based on a P2P architecture for data transporting and indexing. The article provides three short examples on how to apply the proposal in virtual repository. The procedures are implemented under the integrator prototype.