Updated to latest C-Blosc2 2.14.3. This was necessary to fix
several CVEs in the C-Blosc2 library. Also, the new version adds
protection when platforms have just one CPU. This caused the
internal number of threads to be 0, producing a division by zero.
For more info, you can have a look at the release notes in:
Python-Blosc2 comes with NDArray support, which is a new Python class that
allows to create multi-dimensional arrays of compressed data. NDArray
follows the same (or very similar) API as NumPy arrays, so it can be used
as a drop-in replacement. See the documentation for more details:
https://www.blosc.org/python-blosc2/reference/ndarray_api.html
Sources repository
------------------
The sources and documentation are managed through github services at: