Releases: pydicom/pyjpegls
Releases · pydicom/pyjpegls
v1.5.1
v1.5.0
- Fix inconsistent byte ordering when encoding or decoding on big endian systems
- Drop support for Python 3.8 and add support for 3.13
- Require NumPy >= 2.0
v1.4.0
Adds support for NumPy v2.0.
The wheels for Python 3.8 still use NumPy 1.26, while those for Python 3.9-3.12 use NumPy 2.0.
v1.3.0
- Allow the use of 16-bit input when the precision is <= 8 with
encode_buffer()
v1.2.0
- Fixes decoding multi-component images with an interleave mode of 0
- Fixes encoding multi-component images
- Overhauls the interface to make it more flexible:
- Adds
jlswrite()
,encode_array()
andencode_buffer()
- Adds
jlsread()
anddecode_buffer()
- Adds
- The old interface is deprecated and will be removed in v2.0
- This includes
read()
,write()
,encode()
anddecode()
- This includes
v1.1.0
- Updates CharLS to v2.4.2
- Adds support for encoding multi-component (such as RGB) images
- Adds
encode_to_buffer
for encoding from an ndarray to a buffer-like - Adds
decode_from_buffer
for decoding from a buffer-like to a buffer-like - Adds test coverage
- Removes support for Python 3.7 and adds it for Python 3.12
This will likely be the last release that supports backwards compatibility with the original CharPyLS as the next version of CharLS removes the parts of the API that CharPyLS depended on.
Initial release 1.0.4
As this is a fork of PyCharLS, which has the latest version at 1.0.3, we start at 1.0.4.
This has an added release workflow to build wheels and upload them to PyPi.