Release v0.9.1

• Improve Java FileSystem support. Allow use of FileSystem implementations that do not support FileChannel, also allows wider compatability if memory mapped file access is not possible. This improves the ability to use jHDF with file systems like S3. Thanks to @tbrunsch for this contribution.
• Some improvements to test infrastructure. Also thanks to @tbrunsch
• Build updates

Release v0.9.0

• Breaking API change Fix typo WritiableDataset to WritableDataset so you will need to make code updates if you are using writing. Thanks to @jshook
• Allow files contain datatype version 0 to be read. Note a warning will be logged as this is out of spec. #524
• Add support for reading the time datatype. This is not commonly used as it appears to be poorly specified, but it will be read as byte[] to be interpreted. #523
• Build and dependency updates

Release v0.8.6

• Add support for reading implicit index chunked datasets #651 #655
• Test fixes for fixed-array index datasets
• Dependency updates

Release v0.8.5

• Fix issue writing string datasets containing non-ASCII characters #656
• Add support for reading chunked datasets using fixed-array paging. #622
• Allow maven publish tasks to complete without signing, if no signing keys are available. Makes local builds easier and allows building on jitpack.io https://jitpack.io/#jamesmudd/jhdf making a rolling jar release available.
• Dependency updates

Release v0.8.4

• Fix incorrectly written string attributes. #641
• Dependency updates

Release v0.8.3

• Add support for accessing decompressed chunks individually. Thanks to @marcobitplane #626
• Fix OSGi headers, and autogenerate them during the build. Thanks to @mailaender #625 #632
• Delete temporary file when closing a file read from an input stream. Thanks to @ivanwick #262 #636
• Build and dependency updates

Release v0.8.2

• Add support for writing boolean datasets and attributes as Bitfield

Release v0.8.1

• Add support for writing String datasets and attributes
• Add _jHDF default attribute to root group
• Build and dependency updates

Release v0.8.0

• Major writing support improvements
  • Attributes can now be written #552
  • Full support for byte, short, int, long, float, double and wrapper classes as datasets and attributes #587
  • Support for scalar datasets and attributes
  • Much more complete API on writable objects, allowing introspection of data type and data layout and data space etc.
• Many test improvements for writing support
• Build and dependency updates
• Note: This may be the last release supporting Java 8

Release v0.7.0

• Release adding HDF5 writing support! #354. Special thanks to @thadguidry for sponsoring this work. See WriteHdf5.java for example usage. Supports
  • Groups
  • n-dimensional byte, int and double datasets
• Fix UTF-8 groups names #539
• Java 21 now officially supported
• Adds h5dump to CI builds to perform compatability tests.
• Build and dependency updates