[dependabot]

Bumps github.com/posit-dev/go-python-packaging from 0.7.0 to 0.12.0.

Release notes

Sourced from github.com/posit-dev/go-python-packaging's releases.

v0.12.0

What's Changed

• Add Metadata 2.5/2.6 support, fix License-File multiplicity by @​nodivbyzero in posit-dev/go-python-packaging#64

New Contributors

• @​nodivbyzero made their first contribution in posit-dev/go-python-packaging#64

Full Changelog: posit-dev/go-python-packaging@v0.11.0...v0.12.0

Changelog

Sourced from github.com/posit-dev/go-python-packaging's changelog.

Changelog

All notable changes to this module are documented here.

The format is based on Keep a Changelog. While the major version is 0, breaking changes ship in a minor bump rather than a major one, and are always listed under a Breaking heading so they are not mistaken for a safe patch upgrade.

[Unreleased]

Breaking

• The module now requires Go 1.27.2 (was 1.25). Consumers on an older Go toolchain must upgrade, and go mod tidy will raise their own go directive to at least 1.27.2.
• marker, requirement, reqtxt: non-ASCII extra marker literals now normalize like Python's str.lower() (see the extras entry under Fixed). This changes String() output and Evaluate results: extra == "Ä" now renders as extra == "ä". extra == "ΑΣ.Β" no longer matches the active extra ΑΣ_Β (ασ-β vs ας-β). Invalid UTF-8 bytes no longer collapse to U+FFFD, so "\xff" no longer matches "\xfe". (#25)

Added

• packagename: new package with Normalize, the PEP 503 entry point for project-name normalization. It returns exactly what extras.Normalize returns; both now delegate to one shared internal implementation. (#25)
• distribution: Metadata 2.5 (PEP 794) and 2.6 (PEP 808) are now recognized. Metadata 2.5 adds Import-Name and Import-Namespace fields; 2.6 has no new fields (behavior change for Dynamic only).

Fixed

• extras: Normalize now lowercases non-ASCII letters the way Python's str.lower() does, matching pypa/packaging's canonicalize_name: Ä becomes ä, İ becomes i plus U+0307, and Σ becomes ς at the end of a word. Previously only ASCII letters were lowercased. Bytes that are not valid UTF-8 now pass through unchanged; previously each was replaced with U+FFFD. Valid PEP 508 names are ASCII, so their results do not change. (#25)
• distribution: License-File is now parsed as a multi-value field, matching the PEP 639 spec. Previously it was incorrectly treated as single-value.

[0.11.0] - 2026-09-25

Breaking

• marker, requirement, reqtxt: marker string literals now decode Python escape sequences, and malformed escapes are rejected. "line\nbreak" compares as a real newline, not a backslash and n, matching pypa/packaging. A truncated

... (truncated)

Commits

• 0facb58 Add Metadata 2.5/2.6 support, fix License-File multiplicity (#64)
• 063bc51 chore(release): add missing entries and date 0.11.0 (#63)
• 08a3aa5 Add Apache-2.0 §4(b)/(c) provenance blocks and normalize upstream citations (...
• fbcb5d3 marker/extras: port pypa conformance tables + uv-pep508 carets (rstudio/packa...
• c899b98 distribution: default-CI parse coverage + archive_reader leak fix (rstudio/pa...
• 52ec9ed pep508: bound the URL token at horizontal whitespace (#59)
• 273595c pep508: decode Python string escapes in marker literals (#58)
• 68c7c5c chore(release): date 0.10.0 (#57)
• b8ef6ef feat(wheelname): add ParseTags, which does not require a parseable version (#56)
• a9c99ee chore(release): date 0.9.0
• Additional commits viewable in compare view

[dependabot]

Superseded by #78.