It's been a while since the last release, and we have a lot of great changes, fixes, and improvements! It feels about time to bump the minor patch version. Say hello to v1.1.0! A huge thank you to all the contributors to the repo! 🎉

What's Changed

• write: fix/support BigEndian PixelData writes, pad odd byte lengths with extra byte by @ducquangkstn in #302
• Remove mocks by @suyashkumar in #312
• Update to go 1.22 and go mod tidy by @suyashkumar in #317
• Use innolitics DICOM Standard JSON dump to generate tags. by @lnogueir in #316
• Improve comment about magic word not present. by @lnogueir in #318
• Deflate transfer syntax support (for parse) by @suyashkumar in #321
• Remove Reader interface in favor of using the struct directly, use writer struct pointer for consistency by @suyashkumar in #324
• Read elements with VR=UN and undefined VL as SQ by @suyashkumar in #331
• Add WriteOption to override missing transfer syntax in dataset by @suyashkumar in #332
• Parse: If transfer syntax is missing, attempt to infer it by peeking next 100 bytes. by @suyashkumar in #330
• Improve error annotations in read.go, rename interface{} to any by @suyashkumar in #333
• Improve NativeFrame memory footprint by storing data in a flat slice and supporting generic integer data representation. by @suyashkumar in #315
• Fix typos found by codespell by @DimitriPapadopoulos in #284
• tag.Add: allows users to register custom or missing tags into the tags registry by @flicaflow in #342
• Allow unknown specific character set to default to UTF-8 encoding when parsing by @johnmarkli in #346
• Rename testable examples to follow Go convention by @suyashkumar in #351
• Parse undefined length OB/OW elements by @patrickwhite256 in #350

New Contributors

• @lnogueir made their first contribution in #316
• @flicaflow made their first contribution in #342
• @johnmarkli made their first contribution in #346
• @patrickwhite256 made their first contribution in #350

Full Changelog: v1.0.7...v1.1.0

dicom v1.1.0 Latest

Latest

Compare

Choose a tag to compare

What's Changed

• fix Application Entity (AE) value parsing by @bench in #276
• Ignore missing metadata group length field by @faustoespinal-philips in #269
• Add darwin arm64 GOARCH to dicomutil standard release by @suyashkumar in #265
• Add basic Parser.Next based benchmark by @suyashkumar in #266
• Log and return to ensure defers are still called in example dicomutil. by @suyashkumar in #278
• Add Dataset & Element equality methods, greatly speeding up tests that rely on dataset comparisons. by @suyashkumar in #280

New Contributors

• @bench made their first contribution in #276
• @faustoespinal-philips made their first contribution in #269

Full Changelog: v1.0.6...v1.0.7

dicom v1.0.7

Compare

Choose a tag to compare

Notable Changes

• New options to skip pixel data processing or skip pixel data entirely (#256, #255) to save memory or CPU (#264) for those who don't need the PixelData.
• Option to tolerate VL and PixelData mismatches #245.
• Support padding byte with odd number of pixel data bytes #233

What's Changed

• Treat Unknown Tags with defined VL as OW by @suyashkumar in #232
• fix issue with odd number of pixel bytes by @jabillings in #233
• Update go to 1.18 by @suyashkumar in #244
• Allow option to tolerate PixelData VL and expected length mismatches. by @ducquangkstn in #245
• Fix typos found by codespell by @DimitriPapadopoulos in #247
• Documentation: http → https by @DimitriPapadopoulos in #249
• Add Fuzz test by @jesslatimer in #252
• Refactor read* methods into a lightweight reader struct. by @suyashkumar in #254
• Introduce option to SkipPixelData by @suyashkumar in #255
• Add option to skip Element Value processing of PixelData, while preserving roundtrip read/write by @suyashkumar in #256
• Add write support for SkipPixelData option, w/ roundtrip tests. by @suyashkumar in #258
• Bump golang.org/x/text from 0.3.7 to 0.3.8 by @dependabot in #262
• Preset buffer size in benchmark setup, update local benchmark settings. by @suyashkumar in #261
• Benchmark across Parse options, add memory benchmarking. by @suyashkumar in #264

New Contributors

• @jabillings made their first contribution in #233
• @ducquangkstn made their first contribution in #245
• @DimitriPapadopoulos made their first contribution in #247
• @jesslatimer made their first contribution in #252
• @dependabot made their first contribution in #262

Thank you again to all contributors!

Full Changelog: v1.0.5...v1.0.6

dicom v1.0.6

Compare

Choose a tag to compare

What's Changed

• Add initial support for bitsAllocated=1 by @suyashkumar in #212
• Introduce element-by-element Writer API, add SkipMetadataReadOnNewParserInit Parser Option, add ParseOptions to Parser API. by @kristianvalind in #208
• Interpret "" DICOM Character Set as iso-8859-1 by @marineotter in #219

New Contributors

• @marineotter made their first contribution in #219

Full Changelog: v1.0.4...v1.0.5

dicom v1.0.5

Compare

Choose a tag to compare

• In this release, we properly return sentinel errors on dicoms with unsupported BitsAllocated values instead of parsing them incorrectly (#210).

• We also have added some debug logging that will only print when in a binary build with -tags debug (#209).

• Additions and improvements to pkg/ (TODO(suyashkumar): describe these)

dicom v1.0.4

Compare

Choose a tag to compare

v1.0.2 of this library comes with some notable improvements:

• Significant Native PixelData read performance improvements (15-20% from #157 and ~50% from #163)
• vrraw package to hold raw string VR consts

Other useful updates:

• Comprehensive benchmarks and enhanced testing for NativePixelData read/write
• GitHub action benchmark support (PRs show diff against base ref, otherwise shows diff against HEAD~1)

dicom v1.0.2

Compare

Choose a tag to compare

In this release:

• dicomutil now supports a --version flag
• AT Tag support
• Some bugfixes (#155)

dicom v1.0.1

Compare

Choose a tag to compare

🎉 v1.0 is here!

After almost a year of work, we are finally ready to cut the official dicom v1.0 release!

With this release, almost the entire library has been rewritten (from the API to the internals) to be more performant, more maintainable, easier to use, more canonical Go, and better tested. There were so many changes internally and at the API layer, so I'll focus on just 5 to call out below.

A selection of 5 interesting changes:

• A new and more canonical data structure to represent DICOM Dataset and Element in a Go world without generics (inspired by protocol buffer's oneof).
• Easy to use Flat and Nested element iterators to examine the DICOM Dataset tree.
• Simpler Parse and Write APIs (while still retaining some advanced options like channel-based streaming of frames, or an element-by-element read API).
• Several bugfixes discovered along the way (support for icons, support for dicoms without proper headers, and more).
• Completely rewritten and redesigned internal logic for reading and writing dicoms (more piecewise, modular), better handling of reading and writing sequence elements, and new per-commit benchmarks to go with this.

Thanks to all the contributors, and a special thanks to Segmed who logged issues, helped test against terrabytes of public dicoms, and helped with pieces of the tool!

There are still improvements to make, and new features to add on the roadmap, but it's still great to get to this point!

v1.0.0 dicom

Compare

Choose a tag to compare

v1.0.0-alpha.2 Alpha Release Pre-release

Pre-release

Compare

Choose a tag to compare