Repository navigation
Unicode-aware dictation insertion #1892
Description
Feature / improvement for Solenta. Priority: High. Area: Speech.
Confirmed helper gap: prefix+transcript+suffix yields Helloworld, 코드를검토 and Hello worldnext. Handle both boundaries, language-dependent spacing, selection replacement, and provisional/final transcript consistency.
Acceptance criteria:
- Implement the described behavior through existing shared APIs and UI; preserve explicit settings and account/permission boundaries.
- Cover the stated edge cases with focused regression checks and run the relevant typechecks/tests.
- Verify affected visible UI behavior and document platform/dependency limits.
Part of #1869.
Activity
currentbits commentedon Nov 7, 2026
Implemented in PR #1916 (commit 68d2680): dictation adds spacing at word boundaries with Unicode/script handling, replaces selected text, and restores the original selection for empty/cancelled transcription. Incremental fragments remain one provisional range. Added helper cases for punctuation, newlines, combining marks, supplementary-plane letters, Korean/Hindi and unspaced scripts, plus Composer selection coverage. Helper and rendered integration checks pass. Leaving plan:doing until review and merge.
currentbits commentedon Nov 7, 2026
Note
🤖 gpt-6.1-sol responding on behalf of currentbits
Merged in PR #1916 (squash 606efe2). All 14 checks passed and review feedback was resolved.