fix: re.sub backslash-escape crash in PR comment section replace

[JarbasAl]

🤖 Auto-generated by Claude Fable 5 (claude-fable-5) via Claude Code — NOT human-reviewed. Verify before acting.

Bug

scripts/update_pr_comment.py, insert_or_replace_section() (line 1135), called:

return re.sub(pattern, new_section, body, flags=re.DOTALL)

new_section is built by build_section() from arbitrary generated content (test output, coverage numbers, bus-message payloads, etc.) and passed to re.sub as a string replacement. re.sub interprets backslashes in a string replacement as escape/backreference syntax, so content containing sequences like \x or \1 raised:

re.error: bad escape \x at position N

This reddened CI in downstream repos whenever a check's generated section content happened to contain a backslash sequence. Observed deterministically on ovos-skill-volume PR #129's bus-coverage comment step, where the section content included raw \x byte sequences from the bus message dump.

Fix

Pass a callable (lambda _match: new_section) as the replacement instead of a raw string, so re.sub inserts the content verbatim without interpreting any backslash sequences in it. This was the only re.sub/re.subn call in the file with a dynamic replacement argument; audited the rest of the module and found no other instances.

Test evidence

Added TestInsertOrReplaceSectionBackslashSafety in test/test_update_pr_comment.py covering:

• replace-path content containing \x00, \x01, \x4F
• replace-path content containing backreference-like sequences (\1, \g<name>)
• append-path (no existing section) content containing backslashes

Red (before fix, on unmodified insert_or_replace_section):

FAILED test/test_update_pr_comment.py::TestInsertOrReplaceSectionBackslashSafety::test_replace_with_backslash_x_does_not_raise
FAILED test/test_update_pr_comment.py::TestInsertOrReplaceSectionBackslashSafety::test_replace_with_backreference_like_sequence_does_not_raise
re.error: bad escape \x at position 113 (line 6, column 11)
2 failed, 1 passed, 37 deselected in 0.22s

Green (after fix):

40 passed in 93.73s (0:01:33)

Full existing suite (40 tests) still passes after the fix.

Summary by CodeRabbit

• Bug Fixes

  • Fixed an issue where generated content containing backslashes or backreference-like sequences could fail during comment updates.
  • Section replacement and appending now preserve such content correctly.

• Tests

  • Added regression coverage for updating sections with special character sequences.

---

🤖 Independent review added by Claude Sonnet 5 (claude-sonnet-5) via Claude Code — NOT human-reviewed. Verify before acting.

Independent review

Verified against current source (origin/dev diff as posted):

• Confirmed via gh api that this is the only re.sub/re.subn call with a dynamic replacement arg in scripts/update_pr_comment.py (grepped full file: only one re.sub site, at the fixed line).
• The lambda _match: new_section fix is the textbook-correct way to avoid re.sub interpreting backslashes/backreferences in the replacement string — it does not change matching behavior (pattern/re.DOTALL untouched), only how the replacement is substituted. No over-matching risk introduced.
• Regression test (TestInsertOrReplaceSectionBackslashSafety) exercises both the replace-path and append-path with \x, \1, \g<name> sequences — this is a real regression test, not just a happy-path add.
• CI green on all 5 Python versions (3.10–3.14); CodeRabbit review completed with no findings (only the auto-summary comment, folded into this PR's own body above).
• Scope check: no other automation path touched — single-file logic change + tests.

Issue list: none.

Verdict: MERGE
Safe for the human to merge? Yes — correctness fix, real regression test (fails before/passes after per PR's own red/green evidence), CI green, no other paths touched.

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro Plus

Run ID: a2b07bd4-159d-4adc-9482-9db67acd6253

📥 Commits

Reviewing files that changed from the base of the PR and between dc48bdf and 376b303.

📒 Files selected for processing (2)

• scripts/update_pr_comment.py
• test/test_update_pr_comment.py

---

📝 Walkthrough

Walkthrough

insert_or_replace_section now preserves literal backslash sequences when it replaces or appends content. Regression tests cover escape and backreference-like sequences.

Changes

Section replacement safety

Layer / File(s)	Summary
Literal replacement and regression coverage scripts/update_pr_comment.py, test/test_update_pr_comment.py	insert_or_replace_section uses a lambda replacement function. Tests verify that \x, \1, and \g<name> remain unchanged during replacement and append operations.

Estimated code review effort: 2 (Simple) | ~10 minutes

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/re-sub-backslash-escape-crash

---

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.}