Apollo3zehn/PureHDF
A pure .NET library that makes reading and writing of HDF5 files (groups, datasets, attributes, ...) very easy.
C#MIT
Issues
- 0
- 5
Integer overflow in `SelectionHelper.DecodeStream`
#124 opened by audunska - 9
- 0
Add tests for very large files
#125 opened by Apollo3zehn - 4
Divide encoding and decoding infos into separate classes (one per type) and make them replaceable by the user? Like with the JsonSerializer?
#117 opened by Apollo3zehn - 7
- 0
Is ISA-L compiled with AVX2 support? It may be required to pass `-mavx2` to the compiler like for Bitshuffle.PInvoke
#120 opened by Apollo3zehn - 2
Accessing Enumeration nested within Compound Type
#119 opened by rtran89 - 4
Cannot read opaque datasets (created with PureHDF)
#105 opened by Blackclaws - 0
- 0
Improve docs: `ReadRaw`; see also `CanRead_Raw_ByteArray`, `CanRead_Raw_Memory`, `CanRead_Raw_Span`
#93 opened by Apollo3zehn - 0
- 0
DefaultReadingChunkCacheFactory (and write factory): should the input parameter be the dataset itself? Otherwise it is hard to determine a proper chunk size
#116 opened by Apollo3zehn - 3
Honor H5Name for Compound Writing
#85 opened by Blackclaws - 0
- 0
Improve docs: How to write data with nullable value types? Answer: Use `new H5Attribute<int?>(data);`
#92 opened by Apollo3zehn - 0
Add documentation about how to create soft links
#112 opened by Apollo3zehn - 1
Should dataset_access become part of dataset instead of `dataset.Read` methods? It is like that in the C lib
#104 opened by Apollo3zehn - 2
- 0
- 0
Create tests for shuffle16_tiled_avx2 - probably the shuffle mask needs to be reversed
#75 opened by Apollo3zehn - 0
H5NativeWriter: there are nullability warnings (memoryData, `InternalWriteDataset<T, TElement>`)
#97 opened by Apollo3zehn - 0
Are the fixed array and extensible array caches (e.g. `private FixedArrayHeader? _header;` and other variables) thread safe?
#114 opened by Apollo3zehn - 0
Reading: Detect if chunks are well defined (all dimensions except the first one are equal to the dataset dimension) and then bulk copy the contents
#113 opened by Apollo3zehn - 0
Run CI pipeline also for pull requests
#111 opened by Apollo3zehn - 0
Add no-cache test to show that PureHDF uses a suboptimal algorithm to distribute data in simple cases
#110 opened by Apollo3zehn - 0
Add chunk optimized algorithm: distribute whole chunk data, then next chunk`s data
#109 opened by Apollo3zehn - 0
Add option to ChunkCache to throw Exception when cache is being flushed (which means it is too small)
#107 opened by Apollo3zehn - 6
- 0
Support reading big-endian datasets
#102 opened by Apollo3zehn - 0
Write support: be able to pass raw byte array with specific generic type (e.g. double[])
#100 opened by Apollo3zehn - 8
Support for writing H5 references
#95 opened by W-Sebastian - 0
convert `H5OpaqueInfo` to something like `H5ObjectReference`, i.e. a wrapper for the `H5Dataset`?
#96 opened by Apollo3zehn - 3
- 0
Use `RuntimeHelpers.IsReferenceOrContainsReferences<T>()` instead of `DataUtils.IsReferenceOrContainsReferences` using preprocessor directives
#91 opened by Apollo3zehn - 11
Unable to open file written with PureHDF
#88 opened by Blackclaws - 1
- 5
Stackoverflow exception when writing data set
#86 opened by lellid - 2
Over size in DataLayoutMessage4.Create
#84 opened by marklam - 7
Intermittent Write Failure
#82 opened by Blackclaws - 3
Source Generator for Writing
#77 opened by Blackclaws - 0
- 0
Use `h5dump -pH` to also output storage layout information (contiguous, chunked, etc)
#80 opened by Apollo3zehn - 2
Large opaque datasets produce overflow
#78 opened by Blackclaws - 2
- 7
Write Opaque Datatype
#76 opened by Blackclaws - 3
Blosc decompression error at compression level 0
#74 opened by marklam - 1
Shuffle filter breaks roundtripping (seen in 3d array of variable-length lists)
#73 opened by marklam - 4
Merging content from multiple files
#69 opened by Blackclaws - 1