Up to [cvs.NetBSD.org] / pkgsrc / math / py-tables
Request diff between arbitrary revisions
Keyword substitution: kv
Default branch: MAIN
py-tables: fix PLIST for setuptools 84.2 and depend on it Bump PKGREVISION.
py-tables: updated to 3.11.1 Changes from 3.11.0 to 3.11.1 * Fix blosc2 loading (:issue:`1305`, :issue:`1306`). Changes from 3.10.2 to 3.11.0 * Add support for Python 3.14. * Set Cython "freethreading_compatible" directive. Please refer to https://www.pytables.org/cookbook/threading.html for details about threading support. * Python wheel packages are now generated exploiting limited API and stable ABI (`abi3`).
Fix PLIST after py-setuptools update; bump depends and revision
py-tables: update to 3.10.1.
3.10.1
Bugfixes
Fix version constraints for the numpy runtime requirements
(#1204).
For a mistake it didn't allow to use PyTables with numpy 2.x.
Fix compatibility with PyPy (:issue:1205), Thanks to Michał Górny.
Improvements
Enforce numpy >= 2 as build constraint (see discussion in #1200).
Use tuple of plain ints for chunk info coordinates.
Thanks to Ivan Vilata-i-Balaguer.
Enable faulthandler in tables.tests.test_all. Thanks to Eric Larson.
3.10.0
New features
New direct chunking API which allows access to raw chunk data skipping the HDF5 filter pipeline (cutting overhead, see "Optimization tips" in User's Guide), as well as querying chunk information (#1187). Thanks to Ivan Vilata and Francesc Alted. This development was funded by a NumFOCUS grant.
Improvements
This release is finally compatible with NumPy 2 with wheels being built against it so that they are still binary-compatible with NumPy 1 installations (#1176, #1183, #1184, #1192, #1195, #1160, #1172, #1185). NumPy >= 1.20 is required now. Thanks to Antonio Valentino, Maximilian Linhoff and Eric Larson.
Fix compatibility with Python 3.13 (#1166), Python >= 3.10 required. Cython 3.0.10 is required for building. Thanks to Antonio Valentino.
Add type hints to atom.py (#1079). This also narrows some types, only allowing bytes to be stored in VLStringAtom and only str in VLUnicodeAtom. Thanks to Nils Carlson.
Add type hints to (hopefully) the complete PyTables API (#1119, #1120, #1121, #1123, #1124, #1125, #1125, #1126, #1128, #1129, #1130, #1131, #1132, #1133, #1135, #1136, #1137, #1138, #1139, #1140, #1141, #1142, #1143, #1145, #1146, #1147, #1148, #1150, #1151, #1152). Thanks to Ko Stehner.
Reduce impact of CPU information gathering by caching in local file (#1091, #1118, #1081). Thanks to Antti Mäkinen and Maximilian Linhoff.
Bugfixes
Fix Windows AMD64 build issues with Bzip2 and C-Blosc2 libraries (#1188). Thanks to Antonio Valentino and Eric Larson.
Fix typos and may other language errors in docstrings (#1122). Thanks to Ko Stehner.
Fix Blosc2 filter not setting dparams.schunk on decompression (#1110 and #1109). Thanks to Tom Birch.
Fix using B2ND optimizations when Blosc2 is not the only enabled filter; move Fletcher32 compression to end of pipeline when enabled (#1191 and #1162). Thanks to Ivan Vilata and Alex Laslavic.
Fix broken internal passing of createparents argument in Leaf.copy (#1127 and #1125). Thanks to Ko Stehner.
Re-enable relative paths in ExternalLink class (#1095). Thanks to erikdl-zeiss.
Fix using prefix in heavy tests methods of test_queries (#1169). Thanks to Miro Hrončok.
Fix TypeError when computing Blosc2 search paths with missing library (#1188 and #1100). Thanks to martinowitsch, Padraic Calpin and Eric Larson.
Avoid overflow RuntimeWarning on NumPy expectedrows value (#1010). Thanks to wony-zheng and Ivan Vilata.
Other changes
Add wheels for macOS ARM64 (Apple Silicon), set MACOSX_DEPLOYMENT_TARGET in Docker (#1050 and #1165). Thanks to Clemens Brunner, Antonio Valentino, Maximilian Linhoff and Eric Larson.
Avoid illegal hardware instruction under macOS on M1/M2 with Rosetta and AMD64 wheels (#1195 and #1186). Thanks to Antonio Valentino and Jon Peirce.
Produce nightly wheels (with HDF5 1.14.4), also uploaded to Scientific Python Anaconda repo. Wheels are also produced for PR workflows. Thanks to Antonio Valentino and Eric Larson (#1175).
Wheels are no longer linked with the LZO library to avoid licensing issues (#1195). Thanks to Antonio Valentino.
Hash-pin dependencies on wheel workflows to increase build procedure security, with support for Dependabot and Renovatebot updates (#1085 and #1015). Thanks to Joyce Brum and Diogo Teles Sant'Anna.
Hash-pin GitHub action versions in wheels workflow. Thanks to Antonio Valentino.
Update ReadTheDocs configuration to version 2 (#1092). Thanks to Maximilian Linhoff.
Assorted fixes to b2nd benchmark, with new results. Thanks to Ivan Vilata.
Point users to example code to handle "Selection lists cannot have repeated values" exception (#1161 and #1149). Thanks to Joshua Albert.
Remove unused getLibrary C code. Thanks to Antonio Valentino.
Update included C-Blosc to 1.21.6 (#1193). Thanks to Ivan Vilata.
Update included HDF5-Blosc filter to 1.0.1 (#1194). Thanks to Ivan Vilata.
py-tables: update to 3.9.2. Changes from 3.9.1 to 3.9.2 =========================== Bugfixes -------- - Fix the assembly of returned slice data in Blosc2 NDim optimized slice reads by using Blosc2's `b2nd_copy_buffer` (:PR:`1078`). The bug only showed up when the chunk did not fully cover the innermost dimension. Add unit tests to ckeck for regressions, along with foreign-generated files, and enable and fix Blosc2 NDim tests which were not being run. Thanks to Ivan Vilata. Improvements ------------ - PyTables wheels now use a threadsafe build of the HDF5 library (:issue:`1075` and :PR:`1077`). Concurrent reads should be possible with no need for additional locking or monkey-patching of the file open function. Thanks to Kiet Pham. - Partial support for the future NumPy 2, with some tests still failing (:PR:`1068`). Thanks to Thomas Grainger. - Relax the reading of Blosc2 NDim to cope with datasets stored with other tools (:PR:`1072`), e.g. missing chunk rank/shape in filter values, having HDF5 chunks where the Blosc2 super-chunk contains more than one inner chunk, or chunks with data not padded to the full chunk size (example script and tests included). Also enhance checks, comments and logged messages. Thanks to Ivan Vilata. Other changes ------------- - Drop compatibility with the obsolete HDF5 1.8 API. PyTables now requires at least the 1.10 API (:PR:`1080`). Thanks to Antonio Valentino. - Require python-blosc2 >= 2.3.0 or c-blosc2 >= 2.11.0 (which adds support for the `b2nd_copy_buffer` function). - Use the main Conda Forge channel for Python 3.12 (:PR:`1066`). Thanks to Thomas Grainger. - Assorted fixes to the b2nd slicing benchmark. Thanks to Ivan Vilata. - Assorted fixes to b2nd slicing optimization tips (:PR:`1069`). Thanks to Ivan Vilata. Changes from 3.9.0 to 3.9.1 =========================== - Minimum supported version for Python is 3.9 (see :issue:`1062`). Changes from 3.8.0 to 3.9.0 =========================== New features ------------ - Apply optimized slice read to Blosc2-compressed `CArray` and `EArray`, with Blosc2 NDim 2-level partitioning for multidimensional arrays (:PR:`1056`). See "Multidimensional slicing and chunk/block sizes" in the User's Guide. Thanks to Marta Iborra and Ivan Vilata. This development was funded by a NumFOCUS grant. - Add basic API for column-level attributes as `Col._v_col_attrs` (:PR:`893` and :issue:`821`). Thanks to Jonathan Wheeler, Thorben Menne, Ezequiel Cimadevilla Alvarez, odidev, Sander Roet, Antonio Valentino, Munehiro Nishida, Zbigniew Jędrzejewski-Szmek, Laurent Repiton, xmatthias, Logan Kilpatrick. Other changes ------------- - Add support for the forthcoming Python 3.12 with binary wheels and automated testing. - Drop wheels and automated testing for Python 3.8; users or distributions may still build and test with Python 3.8 on their own (see :commit:`ae1e60e` and :commit:`47f5946`). - New benchmark for ERA5 climate data. Thanks to Óscar Guiñón. - New "100 trillion baby" benchmark. Thanks to Francesc Alted. - New benchmark for querying meteorologic data. Thanks to Francesc Alted. Improvements ------------ - Use `H5Dchunk_iter` (when available) to speed up walking over many chunks in a very large table, as well as with random reads (:issue:`991`, :PR:`997`, :PR:`999`). Thanks to Francesc Alted and Mark Kittisopikul. - Improve `setup.py` (now using `pyproject.toml` as per PEP 518) and `blosc2` discovery mechanism. Blosc2 may be used both via python-blosc2 or system c-blosc2 (:PR:`987`, :PR:`1000`, :issue:`998`, :PR:`1017`, :PR:`1045`). Thanks to Antonio Valentino, Ben Greiner, Iwo-KX, nega. - Enable compatibility with Cython 3 (:PR:`1008` and :issue:`1003`). Thanks to Matus Valo and Michał Górny. - Set GitHub workflow permissions to least privileges (:PR:`1007`). Thanks to Joyce Brum. - Add `SECURITY.md` with security policy (:PR:`1012` and :issue:`1011`). Thanks to Joyce Brum. - Handle py-cpuinfo missing in some platforms (:PR:`1013`). Thanks to Sam James. - Avoid NumPy >= 1.25 deprecations, use `numpy.all`, `numpy.any`, etc. instead. Thanks to Antonio Valentino. - Avoid C-related build warnings. Thanks to Antonio Valentino. - Streamline CI wheel building & testing with `cibuildwheel`, more clear distinctions between build and runtime dependencies. - Update included c-blosc to v1.21.5 (fixes SSE2/AVX build issue). - Require python-blosc2 >= 2.2.8 or c-blosc2 >= 2.10.4 (Python 3.12 support and assorted fixes). - Update external libraries for CI-based wheel builds (:PR:`1018` and :issue:`967`): * hdf5 v1.14.2 * lz4 v1.9.4 * zlib v1.2.13 Bugfixes -------- - Fix crash in Blosc2 optimized path with large tables (:issue:`995` and :PR:`996`). Thanks to Francesc Alted. - Fix compatibility with NumExpr v2.8.5 (:PR:`1046`). Thanks to Antonio Valentino. - Fix build errors on Windows ARM64 (:PR:`989`). Thanks to Cristoph Gohlke. - Fix `ptrepack` failures with external links (:issue:`938` and :PR:`990`). Thanks to Adrian Altenhoff. - Replace stderr messages with Python warnings (:issue:`992` and :PR:`993`). Thanks to Maximilian Linhoff. - Fixes to CI workflow and wheel building (:PR:`1009`, :PR:`1047`). Thanks to Antonio Valentino. - Fix garbled rendering of `File.get_node` docstring (:PR:`1021`). Thanks to Steffen Rehberg. - Fix open `extern "C"` block (:PR:`1026`). Thanks to Ivan Vilata. - Fix Cython slice indexing under Python 3.12 (:PR:`1033`). Thanks to Zbigniew Jędrzejewski-Szmek. - Fix unsafe temporary file creation in benchmark (:PR:`1053`). Thanks to Al Arafat Tanin (Project Alpha-Omega).
py-tables: update to 3.7.0. This is a minor version of PyTables. The main feature added is that compatibility with Python 3.10, numpy 1.21 and HDF5 1.12 has been improved, while support for Python 3.5 has been dropped. The CI infrastructure has been moved to GitHub Actions.
py-tables: updated to 3.6.1 Changes from 3.6.0 to 3.6.1 Maintenance release to fix packaging issues. No new features or bugfixes. Changes from 3.5.3 to 3.6.0 PyTables 3.6 no longer supports Python 2.7. Improvements - Full python 3.8 support. - On Windows PyTables wheels on PyPI are linked to `pytables_hdf5.dll` instead of `hdf5.dll` to prevent collisions with other packages/wheels that also vendor `hdf5.dll`. This should prevent problems that arise when a different version of a dll is imported that the version to which the program was linked to. This problem is known as "DLL Hell". With the renaming of the HDF5 DLL to `pytables_hdf5.dll` these problems should be solved. Bugfixes - Bugfix for HDF5 files/types with padding. For details see :issue:`734`. - More fixes for python 3.8 compatibility: Replace deprecated time.clock with time.perf_counter - Improvements in tests as well as clean up from dropping Python 2.7 support.
py-tables: updated to 3.5.1 Changes from 3.5.0 to 3.5.1 - Maintenance release to fix how PyPi repo is handling wheel versions. Changes from 3.4.4 to 3.5.0 Improvements - When copying data from native HDF5 files with padding in compound types, the padding is not removed now by default. This allows for better compatibility with existing HDF5 applications that expect the padding to stay there. Also, when the description is a NumPy struct array with padding, this is honored now. The previous behaviour (i.e. getting rid of paddings) can be replicated by passing the new allow_padding parameter when opening a file. For some examples, see the new examples/tables-with-padding.py and examples/attrs-with-padding.py. - Added a new flag --dont-allow-padding in ptrepack utility so as to replicate the previous behaviour of removing padding during file copies. The default is to honor the original padding in copies. - Improve compatibility with numpy 1.16. - Improve detection of the LZO2 library at build time. - Suppress several warnings. - Add AVX2 support for Windows.
py-tables: (ex py-pytables) updated to 3.4.4 Changes from 3.4.3 to 3.4.4 Improvements Environment variable to control the use of embedded libraries. Include citation in repository. gh-690. Bugs fixed Fixed import error with numexpr 2.6.5.dev0 gh-685. Fixed linter warnings. Fixed for re.split() is version detection. gh-687. Fixed test failures with Python 2.7 and NumPy 1.14.3 gh-688 & gh-689.