v1.2.3

Hot-fix

This release updates the project to 1.2.3 and fixes release automation so the GitHub Actions publish flow uses the tag as the source of truth.

Fixed

• Bumped all versioned project metadata to 1.2.3.
• Updated the Python publish workflow to sync release metadata from the pushed Git tag before building.
• Tightened artifact validation so the built wheel and sdist must match the tagged version.
• Normalized the release helper script so future releases can be driven by the tag version in CI.

Notes

• This change is primarily release/process-related.
• The conda recipe version was updated, but the source tarball sha256 still needs to be filled for the final conda-forge update once the v1.2.3 tag is available.

v1.2.2

Hot-fix - v1.2.2

This release fixes a packaging issue in the Python publish workflow where stale build artifacts could cause an older wheel to be uploaded instead of the new release.

Fixed

• Cleaned the build step before packaging so dist/, build/, and *.egg-info do not leak into the release job.
• Added a version check to ensure the built wheel and sdist match the version in pyproject.toml.
• Prevented stale artifacts such as vcf_rdfizer-1.2.0-py3-none-any.whl from being published for a newer tag.

Notes

• This change affects the release/publishing workflow only.
• No runtime or pipeline behavior changed.

v1.2.1

Hot-fix v1.2.1

This release fixes a bug affecting pip installations where the wrapper could try to build the Docker image from the installed package directory instead of pulling the published image.

Fixed

• pip-installed vcf-rdfizer now pulls the published Docker image when no local checkout Dockerfile is available.
• This avoids failures such as failed to read dockerfile: open Dockerfile: no such file or directory.

Notes

• Source checkouts still behave as before and can build locally when requested.
• No pipeline behavior changes were introduced beyond this install/runtime fix.

v1.2.0

VCF-RDFizer v1.2.0

Highlights

• Added a reusable release helper script to bump version metadata from a single semantic version input.
• Made the Docker publish workflow semver-aware so future releases do not require workflow edits for each new version line.
• Improved the README with a centered project logo and a direct link to the VCF-RDFizer vocabulary.
• Continued release metadata cleanup across packaging and citation files.

Details

• New scripts/release.py updates the version markers in the Python package, citation metadata, README, conda recipe, and release-oriented examples.
• The Docker GitHub Actions workflow now publishes multi-arch images for any semantic version tag and keeps latest updated automatically.
• README now includes the logo from assets/logo/logo.png and a reference to the vocabulary at https://w3id.org/vcf-rdfizer/vocab#.

Notes

• Release and packaging metadata were refreshed to support simpler future version bumps.
• This release is focused on packaging, release automation, and project presentation rather than pipeline behavior changes.

v1.1.0

Highlights

VCF-RDFizer v1.1.0 expands benchmarking and metrics support across the pipeline, improves RDF generation behavior, and refines packaging and release metadata.

What's New

• Added a TSV-only benchmarking mode to run only VCF-to-TSV conversion and collect timing/resource metrics.
• Expanded per-run metrics capture for TSV conversion, RDF conversion, and compression, including raw metrics artifacts and improved summary outputs.
• Fixed incorrect aggregate compression wall-time reporting by using wrapper-observed elapsed runtime for wall-clock metrics.
• Improved batch and aggregate RDF handling, including duplicate part protection and safer cleanup behavior.
• Added helper TSV generation for sample calls and FORMAT values to support richer mapping behavior.
• Updated the default mapping IRIs to the file-based file://{vcfFilePath}... style.
• Improved failure reporting, cross-platform coverage, citation metadata, third-party notices, and packaging/release workflow support.

Packaging

• PyPI publication is triggered automatically by pushing the v1.1.0 tag.
• Docker publication is also triggered for the v1.1.0 release tag.
• Conda-forge will be updated in the feedstock