fix(scripts): pin stderr to UTF-8 in the config scripts

[aranellaeth]

Problem

The three scripts under src/scripts write diagnostics that quote user-controlled text:

• resolve_config.py:67 and resolve_customization.py:84 — f"error: {error}\n", where the message carries the offending config path and tomllib's parse error.
• memlog.py:148 — quotes path; memlog.py:154 — quotes the --field key=value pair straight from argv.

Only stdout is pinned (write_json_stdout). stderr is left at the platform default, so on a Windows console (cp1252) the filename the user needs in order to fix the problem comes back as escapes.

Measured, same run with and without the pin:

with:     error: failed to parse ...\proje-şık\_bmad\custom\bmad-fake.toml: ...
without:  error: failed to parse ...\proje-şık\_bmad\custom\bmad-fake.toml: ...

Change

Generalizes the pin_utf8 helper already merged for brain.py in #2578 to the three scripts, and calls it on stderr at main() entry. write_json_stdout now goes through the same helper instead of its own inline reconfigure.

errors= is passed through deliberately: reconfigure(encoding=...) alone resets the handler to "strict", which would silently downgrade stderr's POSIX backslashreplace default and turn a diagnostic about an undecodable path into a traceback.

Test

Adds one subprocess test that drives resolve_customization.py with PYTHONIOENCODING=cp1252 against a project root whose path contains Turkish characters, and asserts the path appears readable in stderr with no traceback.

Verified it discriminates: with the pin_utf8(sys.stderr) line stripped the test fails on exactly that assertion; restored, it passes.

src/scripts/tests — 42 passed (41 before, +1 new).

Diff is +102 / -7 across three scripts and one test file.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: 505b8dcf-29f5-4c85-ac40-0fd7e44f4d38

📥 Commits

Reviewing files that changed from the base of the PR and between c16dc72 and 89baad4.

📒 Files selected for processing (4)

• src/scripts/memlog.py
• src/scripts/resolve_config.py
• src/scripts/resolve_customization.py
• src/scripts/tests/test_resolve_customization.py

Included review availability: Your plan provides up to 10 included reviews per hour; 9 remain after this review.

---

📝 Walkthrough

Walkthrough

The scripts now pin console streams to UTF-8 while preserving existing error handlers. A regression test verifies that malformed configuration diagnostics with non-ASCII paths remain readable under cp1252 console settings.

Changes

Console encoding handling

Layer / File(s)	Summary
Stream pinning and script integration src/scripts/memlog.py, src/scripts/resolve_config.py, src/scripts/resolve_customization.py	Each script adds or uses a helper that sets UTF-8 encoding and preserves the stream error handler. stdout and stderr are pinned before output or diagnostics.
Non-ASCII diagnostic regression test src/scripts/tests/test_resolve_customization.py	The test runs with cp1252 console encoding and verifies that a malformed configuration reports a non-ASCII path without UnicodeEncodeError or a traceback.

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

Merge Risk: ⚪ Minimal · up to 89baa

The scripts now emit readable UTF-8 diagnostics while preserving existing error handling, improving troubleshooting for paths containing non-ASCII characters. No actionable merge-blocking risk remains beyond normal checks and review.

Suggested reviewers: bmadcode

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 30.77% which is insufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 13 functions across 4 files.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely identifies the main change: pinning stderr to UTF-8 in the configuration scripts.
Description check	✅ Passed	The description accurately explains the encoding problem, implementation, error-handler behavior, affected scripts, and test coverage.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

• Fix all pre-merge checks with AI

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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.}

[greptile-apps]

Greptile Summary

Pins the configuration scripts’ console streams to UTF-8 while preserving each stream’s existing error handler.

• Reuses a common pin_utf8 helper for JSON output and stderr diagnostics.
• Applies UTF-8 stream configuration at each script’s entry point.
• Adds subprocess coverage for a non-ASCII path under a simulated cp1252 console.

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failure remains.

Important Files Changed

Filename	Overview
src/scripts/memlog.py	Adds UTF-8 pinning for stdout and stderr before dispatching memlog subcommands.
src/scripts/resolve_config.py	Refactors stdout encoding through pin_utf8 and pins diagnostic output to UTF-8.
src/scripts/resolve_customization.py	Applies the shared UTF-8 stream behavior to resolved customization output and errors.
src/scripts/tests/test_resolve_customization.py	Verifies that parse errors preserve readable non-ASCII paths when the inherited I/O encoding is cp1252.

_{Reviews (2): Last reviewed commit: "docs(scripts): restore and extend docstr..." | Re-trigger Greptile}

[aranellaeth]

Pushed ff78f08 to address the docstring coverage check, and it turned out to be covering a real slip rather than a style gap.

The first commit folded write_json_stdout's body into the new pin_utf8 helper and dropped its existing docstring in resolve_config.py along with it:

-def write_json_stdout(output) -> None:
-    """Pin stdout to UTF-8 — a Windows cp1252 default cannot encode emoji icons."""

That was unintended. It is restored, and its sibling in resolve_customization.py now carries the same. The follow-up also documents main() in the three scripts and the two test entities this PR touches.

Docstring coverage over the touched set goes from 4/13 to 11/13 (84.6%). The follow-up is +8/-0, pure additions. src/scripts/tests still passes 42.

The two test methods that remain undocumented predate this PR and I did not touch their bodies, so I left them alone rather than widening the diff.