Release v1.1.0

What's Changed

• Timing + Optimisations + Minor fixes by @bnlawrence in #191 - an excellent contribution towards serious optimization!
• Add support for pre-commit with mypy and Ruff and start adding type hints by @valeriupredoi in #173
• [JOSS REVIEW PR] Harmonize depenedncies from environment.yml and pyproject.toml by @valeriupredoi in #174
• pep639 license by @valeriupredoi in #180
• add CONTRIBUTING.md file (and contributing guidelines, of course) by @valeriupredoi in #181
• Documentation consistency fixes by @Zeitsperre in #175
• Allow negative slices when indexing chunked data by @davidhassell in #170
• Overhaul .gitignore file by @valeriupredoi in #179
• add link to Pyfive GitHub repository on doc landing page by @valeriupredoi in #185
• Skip tests if ceda-jasmin s3 storage is offline by @valeriupredoi in #187
• Add orthogonal_indexing flag by @davidhassell in #182
• [pre-commit.ci] pre-commit autoupdate by @pre-commit-ci[bot] in #188
• Replace python-lzf with python-neo-lzf by @valeriupredoi in #189

New Contributors

• @Zeitsperre made their first contribution in #175 as part of the JOSS review, also @dostuffthatmatters contributed towards the JOSS paper review, and gave us good advice on how to improve bits - thank you, gents 🍻
• @pre-commit-ci[bot] made their first contribution in #188 as auto-updater for the pre-commit versions

Full Changelog: v1.0.1...v1.1.0

Release v1.0.1

What's Changed

• Set __version__ attribute dynamically in __init__.py by @valeriupredoi in #152
• pin numpy>=2 by @valeriupredoi in #157
• add a pip check by @valeriupredoi in #159
• Test case for corner case file (buffer too small) by @kmuehlbauer in #160
• update setuptools pins by @valeriupredoi in #165
• Fix display of p5dump global attributes by @zequihg50 in #163

Special mentions

With very many thanks to @mo-gill for highlighting to us the corner case, and to @kmuehlbauer for finding an excellent fix to it in #160

Many thanks to all contributors 🍻

Full Changelog: v1.0.0...v1.1.0

Release v1.0.0

What's Changed

• fix fletcher32, add tests by @kmuehlbauer in #133
• add lzf decompress filter by @kmuehlbauer in #136
• introduce new H5 types to replace current type-tuples by @kmuehlbauer in #122
• mark test_hdf5_filters as flaky by @valeriupredoi in #141
• add pytest reruns plugin pytest-rerunfailures and minimal settings by @valeriupredoi in #142
• Optimise when we get access to b-tree by providing lazier view of datasets, access to b-tree location, and new p5dump by @bnlawrence in #138
• Added btree v2 test but skipping it (#137) by @zequihg50 in #143
• p5dump examples by @davidhassell in #147
• Milestone 1p0 by @bnlawrence in #148
• Proposal for is_cloud_optimized by @zequihg50 in #145

Full Changelog: v0.9.0...v1.0.0

Note

With very many thanks to all who contributed towards v1.0.0, and are not mentioned in the Changelog: @jjhelmus @bmaranville @synaptic @woutdenolf

Release v0.9.0

What's Changed

This is a maintenance release ahead of the major v1.0.0 release (TBA)

• use pytest temporary dir factory to write some of the test hdf5 files and move all fixed hdf5 sample data files to tests/data by @valeriupredoi in #117
• Install netcdf4 from conda-forge and pin netcdf4<1.7.3 in pyproject.toml by @valeriupredoi in #124
• Support Python 3.14 by @valeriupredoi in #125
• remove pin on netcdf4 by @valeriupredoi in #126
• update changelog for v0.8.0 and v0.9.0 (to be released tomorrow 18 Oct 2025, last before big v1.0) by @valeriupredoi in #127

All reviewed by @kmuehlbauer and @davidhassell 🍻

Full Changelog: v0.8.0...v0.9.0

Release v0.8.0

What's Changed

• Support for Opaque datasets by @bnlawrence in #114

with review from @kmuehlbauer

Full Changelog: v0.7.0...v0.8.0

Release v0.7.0

What's Changed

• add joss paper pdf conversion via gha by @valeriupredoi in #97
• fix changelog to include Kai as contributor to v0.6.0 and change Brian L -> Bryan L (typos) by @valeriupredoi in #96
• New logo by @valeriupredoi in #98
• fix Enum and Empty attributes by @kmuehlbauer in #102
• Fix user datatypes (enum, compound) by @kmuehlbauer in #105
• Added partial support for compact datasets. by @zequihg50 in #107
• fix handling of uninitialized vlen strings by @kmuehlbauer in #110
• add dataobjects.dtype to DatasetMeta by @kmuehlbauer in #112

All changes with review from @bnlawrence @davidhassell and @bmaranville - very many thanks 🍻

Full Changelog: v0.6.0...v0.7.0

Release v0.6.0

Version 0.6.0

2025-09-16

• Enumeration Support (#85 by
  Bryan Lawrence and Kai Mühlbauer)

Release by @valeriupredoi containing contributions by @bnlawrence and @kmuehlbauer 🍻

Release v0.5.1

Version 0.5.1

2025-08-22

• Add a Changelog #87 by David Hassell
• Improved documentation #84 by Kai Mühlbauer
• When getting object by address, do not fully instantiate when iterating #83 by Kai Mühlbauer
• Add documentation for Pyfive #81 by Valeriu Predoi
• Setup documentation builds on Readthedocs #80 by Valeriu Predoi

Issue and authors

• Closes #91
• Authors@ @davidhassell @kmuehlbauer @valeriupredoi

Release v0.5.0

Pyfive v0.5.0 release

Closes #75

Main highlights:

• Functionality enhancements to address lazy loading of chunked data, variable length strings, and other minor bug fixes: a completely new approach to handling thread-safe loading of HDF5 files by @bnlawrence and @davidhassell in #68
• Various improvements towards the pyfive Python package by @valeriupredoi #69 #72 #73 #74 #76 #79

Notes:

• the repository has now moved permanently to NCAS-CMS https://github.com/NCAS-CMS and will be maintained by NCAS-CMS
• a release cycle will be set up in the future, for now we are releasing on a "whim" ie when major updates are merged in main

Special thanks:

• many thanks to @jjhelmus for all his help passing the Pyfive baton
• many thanks to @kmuehlbauer for willing to help out here

🍻

pyfive v0.4.0

pyfive v0.4.0