Skip to content

Latest commit

 

History

History
248 lines (210 loc) · 12.1 KB

CHANGES.md

File metadata and controls

248 lines (210 loc) · 12.1 KB
Raw

jHDF Change Log

v0.7.0 - May 2024

  • 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

v0.7.0-alpha

  • Add initial HDF5 writing support! #354. Special thanks to @thadguidry for sponsoring this work. See WriteHdf5.java for example usage.
  • Build and dependency updates

v0.6.10

  • Add support for files containing superblock extensions #462
  • Build and dependency updates

v0.6.9

  • Add support for LZ4 compressed datasets #415
  • Improve BitShuffle performance and reduce memory usage
  • Add CI on ARM64 architecture
  • Build and dependency updates

v0.6.8

  • Add support for getting flat data Dataset#getDataFlat() #397
  • Add support for dereferencing addresses/ids HdfFile.getNodeByAddress(long address) #316
  • Dependency updates

v0.6.7

  • Add support for Bitshuffle filter #366
  • Add ability to get filter data Dataset#getFilters(); #378
  • Dependency and CI updates

v0.6.6

  • Add support for slicing of contiguous datasets. This adds a new method Dataset#getData(long[] sliceOffset, int[] sliceDimensions) allowing you to read sections of a dataset that would otherwise be too large in memory. Note: chunked dataset slicing support is still missing. #52 #361
  • Fix OSGi Export-Package header resulting in API access restriction when running in OSGi. #365 #367
  • Dependency and CI updates

v0.6.5

  • Add support for array type data in multi-dimensional datasets #341
  • Fix issue reading compound type attributes #338
  • Dependency updates

v0.6.4

  • Fix issue with byte shuffle filter when data length is not a multiple of element length. #318
  • Improve testing of byte shuffle and deflate filters
  • Add validation running on Java 17

v0.6.3

  • Improve support for NIO Path. Allows jHDF to open files on non-default file systems such as zip files or remote storage systems. Thanks, @tbrunsch for this contribution #304
  • Fix accessing a missing fill value could cause an exception #307
  • Dependency updates
  • CI and release process improvements

v0.6.2

  • Breaking API change Dataset#getMaxSize now returns long[] allowing files with max sizes larger than int max to be opened. #283
  • Add support for opaque datatype #264
  • Improve chunked dataset read performance with default logging #267
  • Dependency updates
  • Add GitHub Actions CI
  • Switch away from Bintray #250

v0.6.1

  • Add support for committed datatypes #255
  • Add support for attributes with shared datatype
  • Switch dependencies repository to Maven Central #250
  • Code cleanup

v0.6.0

  • Adds support for reading in-memory files from byte[] or ByteBuffers #245
  • Breaking API change To support in-memory files HdfFile#getHdfChannel is replaced by HdfFile#getHdfBackingStorage which now returns a HdfBackingStorage. Internally the new interface replaces the use of HdfFileChannel
  • Fix #247 reading empty arrays in variable length datasets
  • Dependency updates
  • Update Gradle

v0.5.11

  • Add LZF compression support allowing LZF datasets to be read. #239
  • Test dependency updates

v0.5.10

  • Add checksum validation, with "Jenkins Lookup 3 Hash". Will help to detect file corruption.
  • Add support for opening a HDF5 file from an InputStream. Many Java API provide InputStreams so this improves integration possibilities.
  • Test and coverage improvements
  • Test and build dependency updates

v0.5.9

  • Add support for v1 and v2 Data Layout Messages Fix #216
  • Add support for Old Object Modification Time Message - Improves compatibility with older files
  • Fix issue if compact datasets are read multiple times
  • Improve handling of empty contiguous datasets
  • Test and coverage improvements
  • Test dependency updates

v0.5.8

  • Breaking API change Dataset#getDiskSize is renamed Dataset#getSizeInBytes and Attribute#getDiskSize is renamed Attribute#getSizeInBytes
  • New API method Dataset#getStorageInBytes which returns the total storage size of the dataset. By comparison with Dataset#getSizeInBytes allows the compression ratio to be obtained
  • Fixes an issue when reading empty datasets with no allocated storage #162
  • Code quality improvements and cleanup
  • Dependency updates
  • CI and build improvements

v0.5.7

  • Fix #177 Reading null or padded strings of zero length
  • Fix #182 Typo in Dataset.isVariableLength. This is an breaking API change replace calls to isVariableLentgh() with isVariableLength()
  • Add initial support for reading large attributes #183
  • Dependency updates
  • CI and build improvements

v0.5.6

  • Add support for reading half precision (16 bit) floats
  • Add support for getting the ByteBuffer backing contiguous datasets and attributes
  • Memory usage and performance improvements
  • Test coverage improvements
  • CI and build improvements

v0.5.5

v0.5.4

  • Add support for variable length datasets #123
  • Add support for Compound datatype v3 messages allowing more compound datasets to be read
  • Fix #139 bug accessing chunked v4 string datasets
  • Fix #143 bug traversing links
  • Code cleanup
  • Upgrade to Gradle 6.1
  • Update dependencies

v0.5.3

  • Add support for chunked v4 datasets with b-tree chunk indexing
  • Improve exceptions for unsupported b-tree records
  • Improve test coverage
  • Upgrade to Gradle 6.0.1

v0.5.2

  • Fix #124 String padding not handled correctly.
  • Fix #132 Multi dimensional fixed length string datasets read incorrectly.

v0.5.1

  • Fix bug in chunked v4 datasets (added in v0.5.0) where incorrect data was returned if, fixed array or extensible array indexing was used and the dataset dimensions were not a multiple of the chunk dimensions.
  • Adds support for enum datasets (which are returned in string form) #121
  • Adds HdfFile convenience constructors for URI and Path
  • Fix #125
  • Update dependencies
  • Refactors test files to separate HDF5 files from scrips.
  • Improvements to test coverage.

v0.5.0

  • Adds support for some types (the most common) of chunked v4 datasets:
    • Single chunk
    • Fixed array
    • Extensible array
  • Fix #113 fixed length UTF8 datasets can now be read correctly.
  • Fix #112 multiple accesses to a global heap object now behave correctly.
  • Lots of code cleanup and minor improvements
  • Updates dependencies

v0.4.8

  • Add support for reference data type. Thanks to Gisa Meier and JCzogalla #106 #91
  • Creation order tracking is skipped allowing these files to be read
  • FileChannel can now be accessed allowing more low-level access to datasets
  • Add version logging when the library is used

v0.4.7

  • Fix #101
  • Add additional testing of attributes
  • Add attribute example

v0.4.6

  • Adds support for compound datasets
  • Adds support for array data type
  • Adds support for reading chunked datasets with Fletcher32 checksums. Note: the checksum is not verified.
  • Improved performance of Dataset.isEmpty method
  • Dependency updates

v0.4.5

  • Fix #49 - Big (>10x) performance improvement for chunked dataset reads. Chunks are now decompressed in parallel and the resulting data copies are a large as possible.
  • Update Gradle to 5.5
  • Update test dependencies

v0.4.4

  • Fix #88 error when running on Java 8
  • Improvements to IDE support
  • Improvements to exceptions in currently unsupported cases

v0.4.3

  • Initial work for #49 slow chunked dataset reads
  • Lots of typos cleaned up
  • Add additional build data to MANIFEST.MF

v0.4.2

  • Add support for byte shuffle filter
  • Many filter management improvements including support for dynamically loaded filters
  • Add support for reading dataset fill values #74
  • Checkstyle added to improve code consistency - not full code formatting yet...
  • Update Gradle to 5.4
  • Update commons-lang3 to 3.9 (Java 8)
  • Update mockito-core to 2.27.+

v0.4.1

  • Add support for broken links
  • Add support for attribute and link creation order tracking #70
  • Allow superblock v1 files to be loaded
  • Improve exceptions thrown when lazy loading fails
  • Fix bug to allow non-cached groups to be loaded
  • Improvement to documentation
  • Update Gradle
  • Update test dependencies
  • Code base cleanup
  • Improvements to CI builds and PR validation

v0.4.0

  • Add support for accessing attributes (see Attribute.java)
  • Add support for scalar datasets
  • Add support for empty datasets
  • Add support for files with user blocks
  • Fix bug where "old" style groups containing soft links could not be opened
  • Fix bug reading unsigned numbers from "awkward" buffer sizes
  • Lots of minor code cleanup and refactoring
  • Improvements to tests and coverage

v0.3.2

  • Fix bug when fixed size string datasets contain strings of exactly that size.
  • Fix bug where >1D fixed size datasets could not be read
  • Add more JavaDoc
  • Minor refactoring

v0.3.1

  • Add support for String datasets
  • Remove Dataset.getDataBuffer - Not all datasets can reasonably support accessing the backing buffer
  • Dataset.getMaxSize now always returns a result previously returned Optional if no max size was in the file now it returns the dataset size if no max size is present.
  • Remove dependency on org.slf4j.slf4j-simple, now just depends on slf4j-api
  • Update SLF4J to 1.8.0-beta4
  • Update to Gradle 5.2.1 and Gradle plugins

v0.3.0

  • First release to support reading chunked datasets. (note: v1.8 files only)
  • Initial support for compressed datasets, GZIP only at the moment.

Pre 0.3.0

Lots of initial development towards being able to read HDF5 files in pure Java. See Git history if your interested.