[codex] Add Cantonese language option

[pppan2003]

Summary

• add Cantonese (yue) as an explicit language option for capture transcription/STT settings
• add Cantonese (yue) to Qwen TTS and Qwen CustomVoice language lists
• map yue to cantonese when loading backend engines
• update English, Traditional Chinese, and Simplified Chinese UI labels so Chinese and Cantonese are distinct

Notes

• This intentionally keeps Cantonese TTS limited to Qwen/Qwen CustomVoice, where the backend can pass a Cantonese language name through the existing language-code mapping.
• Other TTS engines are unchanged because Cantonese behavior was not verified for them.
• Related: Feature Request: Chinese Language Support — Traditional Script (zh-TW) + Refine Output Language #754 covers broader Chinese language/script refinement, but this PR is specifically for Cantonese language selection.

Validation

• bun run typecheck
• Pydantic request validation for GenerationRequest, SpeakRequest, and VoiceProfileCreate with language="yue"
• Verified backend language mapping resolves yue to cantonese
• End-to-end Qwen TTS runtime smoke test with an existing cloned profile using language="yue", engine="qwen", model_size="1.7B"; generation completed successfully with a 4.0s audio result and no backend error

Summary by CodeRabbit

• New Features
  • Added Cantonese as a distinct language option for both transcription and voice generation, with updated UI labels in English, Simplified Chinese, and Traditional Chinese.
  • Separated “Mandarin / Chinese” from “Cantonese” in the language picker labels.
• Refactor
  • Centralized backend language validation rules and expanded them to accept the new Cantonese language tag across relevant requests.
• Documentation
  • Updated the Unreleased changelog entry to reflect the new Cantonese language support and label changes.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 72449b6a-3d67-4d1a-8faa-64af32d87322

📥 Commits

Reviewing files that changed from the base of the PR and between bc65f85 and 255de43.

📒 Files selected for processing (8)

• CHANGELOG.md
• app/src/components/ServerTab/CapturesPage.tsx
• app/src/i18n/locales/en/translation.json
• app/src/i18n/locales/zh-CN/translation.json
• app/src/i18n/locales/zh-TW/translation.json
• app/src/lib/constants/languages.ts
• backend/backends/__init__.py
• backend/models.py

✅ Files skipped from review due to trivial changes (5)

• app/src/i18n/locales/zh-TW/translation.json
• CHANGELOG.md
• app/src/components/ServerTab/CapturesPage.tsx
• app/src/i18n/locales/en/translation.json
• app/src/i18n/locales/zh-CN/translation.json

🚧 Files skipped from review as they are similar to previous changes (2)

• app/src/lib/constants/languages.ts
• backend/backends/init.py

---

📝 Walkthrough

Walkthrough

Cantonese (yue) is added as a supported language end-to-end: shared regex constants in backend/models.py replace inline patterns and include yue for voice validation; backends/__init__.py registers yue in the language map and all four Qwen model configs; the frontend constants, three locale files, the captures dropdown, and changelog are updated accordingly.

Changes

Cantonese Language Support

Layer / File(s)	Summary
Backend validation patterns and model configs backend/models.py, backend/backends/__init__.py	Extracts VOICE_LANGUAGE_PATTERN and STT_LANGUAGE_PATTERN regex constants; adds yue to the voice pattern; replaces five inline regex literals across request models with these constants. Registers yue → "cantonese" in the language-name map and adds yue to the supported languages list of all four Qwen TTS and CustomVoice model configs.
Frontend constants, i18n strings, and UI dropdown app/src/lib/constants/languages.ts, app/src/i18n/locales/en/translation.json, app/src/i18n/locales/zh-CN/translation.json, app/src/i18n/locales/zh-TW/translation.json, app/src/components/ServerTab/CapturesPage.tsx	Adds yue: 'Cantonese' to ALL_LANGUAGES and appends yue to ENGINE_LANGUAGES.qwen and ENGINE_LANGUAGES.qwen_custom_voice. Updates the zh label to "Mandarin / Chinese" (and locale equivalents) and adds yue translations in all three locale files. Adds the yue SelectItem to the captures transcription language dropdown.
Changelog documentation CHANGELOG.md	Documents the introduction of Cantonese as a distinct language option with code yue and clarifies the UI separation between Mandarin/Chinese and Cantonese.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Poem

🐇 Hop hop, a new tongue joins the choir,
Cantonese whispers through circuits and wire,
yue in the dropdown, the backend, the map,
From Guangzhou to bytes in a blink and a tap.
The rabbit nods fondly — 廣東話 is here! 🎉

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 66.67% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately summarizes the main change: adding Cantonese language support to the application, which is reflected across all modified files including UI, translations, backend constants, and validation patterns.
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.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ 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 and usage tips.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@backend/models.py`:
- Line 15: The STT_LANGUAGE_PATTERN regex constant excludes Hindi, creating a
mismatch where frontend UI allows selecting Hindi but backend validation rejects
it. Add `hi` to the regex pattern in the STT_LANGUAGE_PATTERN constant alongside
the existing language codes (en, zh, yue, ja, ko, de, fr, ru, pt, es, it) so
that frontend language selections are consistently accepted by backend
validation.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 1630b0b7-9692-4f30-9147-823a6776e30a

📥 Commits

Reviewing files that changed from the base of the PR and between b35b909 and bc65f85.

📒 Files selected for processing (7)

• app/src/components/ServerTab/CapturesPage.tsx
• app/src/i18n/locales/en/translation.json
• app/src/i18n/locales/zh-CN/translation.json
• app/src/i18n/locales/zh-TW/translation.json
• app/src/lib/constants/languages.ts
• backend/backends/__init__.py
• backend/models.py

[coderabbitai]

⚠️ Potential issue | 🟠 Major | ⚡ Quick win

STT_LANGUAGE_PATTERN still excludes Hindi, which breaks frontend/backend language contract.

At Line 15, STT validation omits hi, but the captures UI still allows selecting hi; those requests can fail validation. Please include hi in the shared STT regex so UI selections are accepted consistently.

Suggested patch

-STT_LANGUAGE_PATTERN = "^(en|zh|yue|ja|ko|de|fr|ru|pt|es|it)$"
+STT_LANGUAGE_PATTERN = "^(en|zh|yue|ja|ko|de|fr|ru|pt|es|it|hi)$"

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change

	STT_LANGUAGE_PATTERN = "^(en|zh|yue|ja|ko|de|fr|ru|pt|es|it)$"
	STT_LANGUAGE_PATTERN = "^(en|zh|yue|ja|ko|de|fr|ru|pt|es|it|hi)$"

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@backend/models.py` at line 15, The STT_LANGUAGE_PATTERN regex constant
excludes Hindi, creating a mismatch where frontend UI allows selecting Hindi but
backend validation rejects it. Add `hi` to the regex pattern in the
STT_LANGUAGE_PATTERN constant alongside the existing language codes (en, zh,
yue, ja, ko, de, fr, ru, pt, es, it) so that frontend language selections are
consistently accepted by backend validation.