I’m happy to announce the first stable release of HDF5-UDF. HDF5-UDF is a tool that enables the procedural generation of datasets using user-defined functions (UDFs). It is possible to access the network and connect to sensors, web services, and more – and expose that data as HDF5 datasets that regular tools can readily consume.
This first release provides support for UDFs in Python, C/C++, and Lua – all of which execute in a sandboxed environment that restricts access to system resources. Moreover, HDF5-UDF is shipped as a filter, which means that no modifications are needed to applications that read from HDF5 files.
The project comes with several examples that should help you get started. It’s even possible to embed the classic Doom game as an HDF5 dataset, for instance!
Please refer to the installation page for details on how to install it from the binary package or to build it from the source code.
I hope you enjoy it!