Release hygiene patch v2.3.1 - #6
Conversation
- Fix snapshot dates in README (2026-04-23 -> 2026-04-26, 2 places). - Fix BPMN-patterns count in archive description (8 -> 10). - Fix recommended best practices count in current sections (10 -> 11). - Rename 'Как работает внутри (9 шагов)' to '(11 шагов)'. - Fix typo 'не все ли узлы' -> 'все ли узлы' in Troubleshooting. - Backfill v2.3.0 release date (27 апреля 2026 / 2026-04-27). - Add 6 release-tests in test_metadata.py covering README<->source drift. - No changes to BPMN generation, Wizard, reuse-ID or Excel behavior.
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 1c98a76797
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
|
|
||
| # Pattern 1: degraded mode XML comment in Requirements section | ||
| comment_matches = list(re.finditer( | ||
| r"snapshot v1\.0 \((\d{4}-\d{2}-\d{2})\)", |
There was a problem hiding this comment.
The new snapshot-date test hard-codes v1.0 (snapshot v1.0 (...) and версия 1.0 от ...) instead of reading snapshot_version from SKILL.md, so a legitimate future snapshot version bump (for example to 1.1) will fail release tests even when README and frontmatter are fully consistent. This turns the guardrail into a false blocker for future releases.
Useful? React with 👍 / 👎.
| cutoff_match = re.search(r"^### v1\.", readme, re.MULTILINE) | ||
| current_part = readme[: cutoff_match.start()] if cutoff_match else readme |
There was a problem hiding this comment.
The cutoff ^### v1\. is too narrow: it leaves all v2.x changelog entries inside current_part, so this test will enforce the current best-practices count on historical v2.* release notes. When the checklist count changes in a future release, CI will fail unless old changelog text is rewritten, which is a regression from the intent to ignore historic sections.
Useful? React with 👍 / 👎.
1 participant
Summary
Tests