fix: vocabulary replacement fails for English words adjacent to CJK characters

[sysalpha01]

Summary

• Fix vocabulary replacement regex that fails when English words appear adjacent to CJK (Japanese/Chinese/Korean) characters
• Change \p{L} to \p{Script=Latin} in word boundary lookahead/lookbehind
• Latin word boundary protection still works (e.g., "apple" in "pineapple" is not replaced)

Problem

The word boundary regex in applyTextReplacements() uses \p{L} (all Unicode letters) in negative lookahead/lookbehind. Since \p{L} matches CJK characters, English words adjacent to Japanese text (e.g., Xavix in Xavixの設定) are never replaced because the CJK character triggers the boundary check.

Example

With vocabulary entry Xavix → ZABBIX:

• Input: Xavixの設定
• Before fix: Xavixの設定 (no replacement - の matches \p{L})
• After fix: ZABBIXの設定 (correctly replaced - の doesn't match \p{Script=Latin})

Changes

apps/desktop/src/utils/text-replacement.ts:

• Replace \p{L} with \p{Script=Latin} in the word boundary regex
• This ensures only Latin script characters are considered for word boundaries
• CJK characters adjacent to English words no longer block replacement

Test Plan

• Verified Xavixの設定 → ZABBIXの設定 works correctly
• Verified pineapple does not have apple replaced (Latin boundary still works)
• Verified pure CJK replacements (e.g., ザビックス → ZABBIX) still work
• Tested with Amical Cloud transcription on Windows

Summary by CodeRabbit

• Bug Fixes
  • Text replacement now properly handles mixed-script text, including CJK characters alongside Latin characters.

[coderabbitai]

📝 Walkthrough

Walkthrough

The change modifies the word-boundary regex in the non-CJK branch of applyTextReplacements to use Latin-script boundaries instead of broader alphabetic/numeric boundaries. This allows CJK characters to be adjacent to matched text while preventing unintended matches within non-Latin words.

Changes

Cohort / File(s)	Summary
Text Replacement Logic apps/desktop/src/utils/text-replacement.ts	Modified word-boundary regex from \p{L}\p{N} to [a-zA-Z0-9] in non-CJK branch, narrowing boundary checks to Latin script and allowing CJK character adjacency. Updated comments to reflect Latin-script focus.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• amicalhq/amical#81: Introduced the original applyTextReplacements logic that this PR directly modifies.

Suggested reviewers

• haritabh-z01

Poem

🐰 A boundary drawn with Latin grace,
No CJK chars out of place,
"Xavixの設定" now sings true,
Word edges sharp, regex anew! ✨

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main fix: it addresses a specific bug where vocabulary replacement fails when English words are adjacent to CJK characters, which aligns with the primary change in the regex modification.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

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

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

No actionable comments were generated in the recent review. 🎉

---

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.}