[joamag]

Summary

• reportlab 5.1.0 (released 2026-10-22) bumped __min_python_version__ to (3,10) in its __init__.py, but its package metadata still declares requires_python >=3.9. Pip on 3.9 installs it and import reportlab then raises ImportError: reportlab requires 3.10+, breaking test collection for test/node.py and test/printing/pdf.py (e.g. run 40146745150).
• Adds a reportlab<5.1.0 range for Python 3.9 in both requirements.txt and setup.py, following the existing per-version caps, and moves the uncapped entry to >= 3.10.

Testing

Ran the CI install and ADAPTER=tiny pytest steps in containers:

Image	reportlab	Result
python:3.9	5.0.1	289 passed, 5 skipped
pypy:3.9	5.0.1	289 passed, 5 skipped
python:3.10	5.1.0	289 passed, 5 skipped

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-26T08:39:16.290000Z	5630b98	PR opened

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" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[coderabbitai]

Caution

Review failed

The pull request is closed.

Finishing Touches

📝 Generate docstrings

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[claude]

Claude finished @joamag's task in 1m 45s —— View job

---

Review of PR #48

• Read the diff
• Read AGENTS.md and the files around the change
• Check line endings, the other places reportlab is pinned, the CHANGELOG
• Post findings

I found no bugs or regressions. There's one documentation nit, posted inline.

Findings by severity

• Nit (docs): CHANGELOG.md:20 says "the latest reportlab release". That stops being accurate once reportlab ships another version. It should name 5.1.0 and link fix: cap reportlab below 5.1.0 on Python 3.9 #48, like the entry for the earlier caps at line 172. I left a suggestion inline.

What I checked (confidence: high)

• Markers, requirements.txt: every Python version matches exactly one reportlab line:
  • 3.6–3.8 → <4.4.3
  • 3.9 → <5.1.0
  • 3.10+ → uncapped
  • 2.7 uses requirements.py2.txt, which this PR doesn't change.
• Markers, setup.py: same coverage:
  • 2.7 → uncapped (the same as before)
  • 3.9 → <5.1.0
  • 3.10+ → uncapped
• Comparing "3.9" with "3.10": these markers are compared as versions, not as text, by the pip versions this matrix uses. That includes the pip<21 of the Windows XP build on 2.7. The < "3.10" bound is safe.
• CI coverage: the Main Workflow tests CPython 3.9 and PyPy 3.9 with pip install -r requirements.txt, so CI checks the new range. The Windows builds (3.14.8 and 2.7.18), the E2E run (3.12) and the Docker image (hivesolutions/python:latest) don't hit the 3.9 marker, so they're unaffected.
• AGENTS.md rules:
  • The version in setup.py wasn't bumped.
  • Line endings are kept: CRLF in setup.py and requirements.txt, LF in CHANGELOG.md, as on master.
  • The entry is in the Unreleased section.
  • No Python source changed, so the black, 2.7/3.13 and comment-style rules don't apply.
• Tests and security: this is a dependency-only change with nothing new to test, and it touches no secrets, payloads or file paths.

What I couldn't check

• I couldn't reach PyPI from the runner, so I didn't confirm the 5.1.0 metadata (requires_python >=3.9 while the code needs 3.10) myself. I relied on the linked failing run and the test table in the PR description.
• The Main, E2E, Windows and Docker workflows for this PR were still running when I finished. Only the Admin Workflow had finished (it passed).

[joamag]

LGTM

[claude]

Nit (docs): "the latest reportlab release" won't make sense once reportlab ships another version, and someone reading this later won't know which release broke 3.9. The entry for the earlier caps (line 172) names the failure and links the reference. Something like this would match it:

Suggested change

	* PDF support on Python 3.9 after the latest reportlab release
	* PDF generation failing on Python 3.9 with reportlab 5.1.0, which requires Python 3.10 - [#48](https://github.com/hivesolutions/colony-print/pull/48)