Blosc2

* New compatibility with MinGW32/64. See 302.

* Improved support for AArch64 (aka ARM64), ARMv7l, ARMv6l and powerpc64le.
See 306. Thanks to Alexandr Romanenko, Mark Kittisopikul and Mosè Giordano
from the Julia packaging team for their help.

* BloscLZ updated to 2.4.0. Aligned access in ARM has been enabled, as well
as other performance improvements. Expect much better performance,
specially on ARM platforms (like Apple Silicon).

* Zstd sources updated to 1.5.0.

* zlib-ng sources updated to 2.0.3.

=======================================

* [API change] `blosc2_decompress_ctx()` gets a new `srcsize`
parameter to ensure that it does not read past the end
of the provided buffer. See 144. Thanks to Nathan Moinvaziri
(nmoinvaz).

* [BREAKING CHANGE] The format for frames has changed and
BLOSC2_VERSION_FRAME_FORMAT is now set to 2. There is no attempt to support
previous formats, but there will probably be backward compatibility support
starting from version 2 on.

* New functionality for updating, inserting and deleting chunks in a super-chunk.

* Support for special values. Large sequences of repeated values can be represented
with an efficient, simple and fast run-length representation, without the need to use
regular codecs.

* Internal Zstd sources updated to 1.4.9.

* Internal LZ4 sources updated to 1.9.3.

* Internal zlib support is provided now by new zlib-ng 2.0.2 (replacing miniz).

* The support for Snappy codec has been completely removed. Snappy is a C++
library, which is not good for a library that aims to be fully pure C.
Snappy was removed from sources in C-Blosc(1) some years ago, so there
should not be a lot of data compressed with Blosc/Snappy out there (and
for the existing ones, a transcoding is always possible using C-Blosc(1)).

* The Lizard codec has been removed. Lizard is a pretty good one, but it
looks like it is not timely maintained. Zstd/Zlib can cover its place pretty
nicely.

* The split of blocks only happens for BLOSCLZ and LZ4 codecs. All
the rest are not split at all.

* Public APIs for frames have been removed. Frames should be considered an
storage detail, so having them accessible publicly should only bring
unnecessary cognitive load. Care have been carried out so as to ensure
the same functionality via the super-chunk (schunk) API.

* [FORMAT] New *sparse frame* format for on-disk I/O. This allows for storing
data chunks in super-chunks in separate files inside a directory. The way
to generate sparse frames is via `storage.contiguous=false` and
`storage.urlpath=dirname`. See `README_SFRAME_FORMAT.rst` for details.

========================================

* The prefilter functionality has been introduced and declared stable.
With that, you can callback arbitrary functions previous to any filter.
This can be useful for performing (parallel) computations on chunks.
For an example of use, see `tests/test_prefilter.c`.

* New blosc2_set_maskout() function to avoid decompressing blocks. This
can be handy when it is not needed to decompress all the blocks in a
chunk. This should be always called before blosc2_decompress_ctx() and
its effect is reset to the default (decompress all blocks) after that.

* New ALTIVEC optimizations for Power architecture. These include support
for both shuffle and bitshuffle filters. For details, see https://github.com/Blosc/c-blosc2/pull/98.
Thanks to Jerome Kieffer and ESRF for making this happen.

* New blosc2_frame_from_sframe() function for getting a `blosc2_frame`
out of an in-memory serialized frame.

* Zstd codec updated to 1.4.4.