Skip to content

fix: restore zh-Hant locale labels - #1323

Closed
dso2ng wants to merge 1 commit into
nesquena:masterfrom
dso2ng:fix/zh-hant-locale-cyrillic
Closed

dso2ng wants to merge 1 commit into
nesquena:masterfrom
dso2ng:fix/zh-hant-locale-cyrillic

Conversation

@dso2ng

@dso2ng dso2ng commented Apr 30, 2026

Copy link
Copy Markdown
Contributor

Thinking Path

  • Hermes WebUI's zh-Hant locale should present Traditional Chinese labels consistently.
  • A few recently added tree/MCP labels in the zh-Hant block were copied from the Russian locale.
  • Those labels appear in JSON/YAML tree toggles and Settings → System → MCP Servers.
  • This PR restores the affected labels to Traditional Chinese and adds a regression check so Cyrillic text does not slip back into zh-Hant.

What Changed

  • Replaced Russian zh-Hant labels for tree/raw view and MCP server settings with Traditional Chinese text.
  • Normalized the parse-failed note to readable Chinese text.
  • Added a targeted zh-Hant regression test that checks the expected labels and rejects Cyrillic characters in the zh-Hant block.

Why It Matters

  • zh-TW users currently see mixed Russian/Chinese UI text in affected areas.
  • The regression test keeps future locale copy/paste fallout easy to catch.

Verification

  • python -m pytest tests/test_chinese_locale.py -q
  • Result: 5 passed

Risks / Follow-ups

  • Low risk; this only changes locale strings plus a static regression test.
  • No behavior or layout changes.

Model Used

  • AI-assisted with OpenAI GPT-5.5 via Hermes Agent, using local git/pytest/GitHub CLI tooling.

@nesquena-hermes

Copy link
Copy Markdown
Collaborator

Thanks. zh-Hant labels accidentally inheriting from the ru block is exactly the kind of locale copy/paste fallout that's hard to spot in review — Cyrillic in a Traditional Chinese context only surfaces when a zh-TW user opens the affected panel.

The two main fixes look right:

  • tree/raw view toggles back to Traditional Chinese
  • Settings → System → MCP Servers labels back to Traditional Chinese
  • parse-failed note normalized

The regression test in test_chinese_locale.py that rejects Cyrillic codepoints in the zh-Hant block is the real win here — keeps this from drifting back even if a future PR copies from ru again. Worth considering extending that pattern to other locale blocks (reject mismatched scripts per locale) in a follow-up.

+51/-25, locale-only. Looks good.

@nesquena-hermes

Copy link
Copy Markdown
Collaborator

Shipped in v0.50.245 via batch release PR #1334 (merge commit 52e1567). Thanks @dso2ng! 🙏

Live now at https://github.com/nesquena/hermes-webui/releases/tag/v0.50.245.

Released alongside 9 other contributor fixes — see the v0.50.245 entry in CHANGELOG.md.

bsgdigital pushed a commit to bsgdigital/hermes-webui that referenced this pull request Apr 30, 2026
From PR nesquena#1323.

Co-authored-by: Dennis Soong <dso2ng@gmail.com>
GeoffBao pushed a commit to GeoffBao/hermes-webui that referenced this pull request May 1, 2026
From PR nesquena#1323.

Co-authored-by: Dennis Soong <dso2ng@gmail.com>
SysAdminDoc pushed a commit to SysAdminDoc/hermes-webui that referenced this pull request Jun 26, 2026
From PR nesquena#1323.

Co-authored-by: Dennis Soong <dso2ng@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants