Repository navigation
Refresh dependencies and fix first-completion cancellation races - #31
Conversation
Bumps [pytest](https://github.com/pytest-dev/pytest) from 8.4.2 to 9.0.3. - [Release notes](https://github.com/pytest-dev/pytest/releases) - [Changelog](https://github.com/pytest-dev/pytest/blob/main/CHANGELOG.rst) - [Commits](pytest-dev/pytest@8.4.2...9.0.3) --- updated-dependencies: - dependency-name: pytest dependency-version: 9.0.3 dependency-type: direct:development ... Signed-off-by: dependabot[bot] <support@github.com>
dependabot
Bot
added
dependencies
![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: 2038ac71ad
ℹ️ About Codex in GitHub
Codex has been enabled to automatically 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 👍.
When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".
| "ruff>=0.15.12", | ||
| "build>=1.5.0", | ||
| "pytest>=8.3.3,<9", # pytest-codeblocks 0.17.0 still uses the pre-pytest-9 collect hook signature. | ||
| "pytest>=8.3.3,<10", # pytest-codeblocks 0.17.0 still uses the pre-pytest-9 collect hook signature. |
There was a problem hiding this comment.
When this dependency group is resolved from pyproject.toml without the checked-in lockfile, or under a lowest/min-version resolution, it now permits pytest==9.x together with pytest-codeblocks==0.17.0; the existing comment says that plugin version still uses the pre-pytest-9 collect hook signature, and the repo’s pytest configuration always enables --codeblocks, so README/docs collection can abort before tests run. Either keep pytest below 9 or raise the pytest-codeblocks floor to a pytest-9-compatible release.
AGENTS.md reference: AGENTS.md:L20-L23
Useful? React with 👍 / 👎.
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
|
Refreshed against current main. Local pytest 9 validation passes 140 real event-bus tests and collects 53 README examples. Current CI has 119 successful checks, but the Rust worst-case forwarding performance test exceeds its unchanged budget (0.613 ms/event versus 0.600 ms/event): https://github.com/ArchiveBox/abxbus/actions/runs/37531280965/job/112501055300 . Leaving this PR unmerged until that failure is resolved; no performance assertions or timeouts were weakened. |
pirate
changed the title
pirate
changed the title
1 participant
Refresh vulnerable test/build dependencies: pytest 9.0.3, Werkzeug 3.1.9, OpenTelemetry exporter 0.219.0 / SDK 2.8.0, tsx 4.23.15, brace-expansion 5.0.12, and tar 7.5.22. OpenTelemetry removes vulnerable protobufjs; tsx removes its older esbuild copy. Keep one effective tar override and remove the obsolete workspace pin. Limit pytest to <9.1 while pytest-codeblocks uses the removed collect-hook argument.
CI exposed a Go first-completion race: cancelling losing handlers could become visible before their result records were terminal, and first-result polling read those records without synchronization. Settle loser results before signalling cancellation, atomically prevent cancelled pending handlers from starting, and use the existing result snapshot for polling. A real-handler regression observes the serialized loser result at cancellation time.
Validation: the publication regression fails before the fix; focused cancellation tests pass 100 repetitions, and the targeted race-detector run passes 20 repetitions. The full Go suite, 140 Python event-bus tests, 87 TypeScript tests, JS/type builds, 53 collected README examples, and pre-commit checks pass locally. pnpm audit reports zero vulnerabilities. No assertions, performance budgets, or timeout limits are weakened. Full hosted CI on the final commit is required before merge.
Final hosted CI passed: https://github.com/ArchiveBox/abxbus/actions/runs/37545168686 (commit 93cbdcb).