Issue Summary

When typing in Korean or Japanese within the Bullet macro in Confluence Cloud, the first character input fails to combine or convert correctly.

For Korean, the initial Jamo separates from the rest of the syllable (e.g., typing '안녕' results in 'ㅇㅏㄴ녕').

For Japanese, the first Romaji input may remain as an alphabet character or fail to merge into the Hiragana conversion (e.g., typing 'はは' results in 'hあ').

This issue is specific to CJK IME inputs and occurs on both Chrome and Firefox.

Our customer has also confirmed that the same IME composition issue occurs in numbered lists (Numbered macro), not only in bullet lists.

Steps to Reproduce

1. Open a Confluence Cloud page in edit mode.
2. Insert a Bullet macro.
3. Switch the input language to Korean or Japanese.
4. Start typing a word (e.g., "안녕" or "はは").

Expected Results

The characters should combine and convert correctly from the start (e.g., "안녕", "はは").

Actual Results

• Korean: The first Jamo is separated (e.g., 'ㅇㅏㄴ녕').

• Japanese: The first character may appear as English or fail to convert properly (e.g., 'hあ').

No relevant client-side exceptions were observed in the browser console. The network errors (409 Conflict, 400 Bad Request) observed in the logs are unrelated to the editor input events.

Workaround

You can avoid this issue by not typing directly inside a bullet/numbered list at first.

1. First, type your Japanese text as normal paragraphs (without any list formatting).
2. After you finish typing, select the text and then apply the bullet or numbered list.

This way, the IME issue with the first character does not occur in most cases.