Vietnamese text input renders with corrupted/doubled diacritical marks

Summary

When typing Vietnamese text in the terminal input on Windows, characters with compound Unicode diacritics (tone mark + vowel modifier) are rendered incorrectly. The base vowel, modifier, and tone mark appear as separate visible characters instead of a single composed glyph.

Expected Behavior

Vietnamese characters with compound diacritics render as single composed glyphs. For example:

• gặp displays as gặp
• lỗi displays as lỗi
• những displays as những

Actual Behavior

Characters with compound diacritics are split/doubled when typed. Examples of what appears on screen:

• gặp → rendered as gaạpaập
• lỗi → rendered as lôiỗi
• những → rendered as nhunữững
• cập nhật → rendered as câạp nhâật

Screenshot showing the bug:

Steps to reproduce the issue

1. Open command-code v0.27.0 in VS Code terminal on Windows 11
2. Switch input language to Vietnamese (Telex or VNI input method)
3. Type any Vietnamese sentence with compound diacritics, e.g.: gặp lỗi trong những bản cập nhật
4. Observe the rendered text in the input field

Command Code Version

v0.27.0

Operating System

Windows

Terminal/IDE

VSCode

Shell

PowerShell

Additional context

No response

Aucun fichier source ni aucun test n’est identifié dans le rapport. Commencez par reproduire le problème dans le terminal de VS Code sous Windows 11 avec PowerShell et une saisie vietnamienne Telex ou VNI, puis suivez le chemin de rendu des entrées du terminal. Le travail est terminé lorsque les signes diacritiques vietnamiens composés s’affichent sous forme de glyphes composés uniques pour les exemples listés.