Interfacing with Multi-Model Data and Queries (Master Project, Finished)
Author
Description
Polypheny is a versatile system capable of handling a verity of data and workloads. The newly added Protobuf-based query interface provides a scalable and feature-rich interface supporting both the relational and the document model. However, since Polypheny also supports the LPG model, the interface yet misses support for a major data model of Polypheny.
With Polypheny being designed as a middleware between the applications and the data, an important aspect of such a PolyDBMS is to provide adapters for various types of data, including static data sources like files. Two important data formats that are not yet supported by Polypheny are JSON and XML.
This Masters's project consists of two parts: The first part is to add support for the LPG data model to the proto query interface and the second part is to implement a data source adapter that maps static JSON and XML files as collections in Polypheny.
In more detail, the objectives of this project are:
- Add support for the LPG data model in the proto interface and the JDBC driver.
- Benchmark the performance of the proto interface and improve where necessary.
- Automate the generation of the proto interface documentation.
- Improve the handling of large results, especially binary results.
- Creating a reference implementation of a driver for C++.
- Implement a data source adapter for JSON and XML files.
- Improve the existing CSV adapter and proto interface where necessary.
If time allows, the project can be extended by adding support for additional data formats to the data source adapter (e.g., YAML, BSON), implementing drivers for additional programming languages and frameworks, or even implementing a second data source adapter.
Start / End Dates
2024/03/01 - 2024/08/31