Japanese IME doesn't preserve the text that it overwrites #8569

Open
opened 2026-01-31 01:32:48 +00:00 by claunia · 0 comments
Owner

Originally created by @j4james on GitHub (May 25, 2020).

Environment

Windows build number: Version 10.0.18362.720
Windows Terminal version (if applicable): Commit 2af722b43d

Steps to reproduce

  1. Install the Japanese language pack.
  2. Open a cmd shell in conhost (built from source).
  3. Open the Properties dialog and change the font to MS Gothic.
  4. Switch the input language to Japanese Microsoft IME.
  5. Make sure the IME option is initially Half-Width Alphanumeric.
  6. Type abcdefgh
  7. Move the cursor back to the start of the line.
  8. Switch the IME option to Hiragana.
  9. Type sushi then press Esc to close the IME.
  10. Press Esc again to cancel the added text.

Expected behavior

In the process of typing sushi, the latin characters abcdef will be temporarily overwritten, but after step 9, I'd expect the ef restored, and only the abcd overwritten with すし. After step 10, I'd expect to see all the latin characters restored.

This is what those two steps looks like in the cmd shell in my current version of Windows.

image

Actual behavior

After step 9, the ef characters have been replaced with blanks, and after step 10 they're still blank.

image

What is weird is that I tried reverting the code back to the first public commit and it still seems to have that problem. I would have thought the cmd shell that shipped with Windows was much more recent than that. So I don't know if maybe it's just something about the build process that I'm getting wrong.

Originally created by @j4james on GitHub (May 25, 2020). # Environment Windows build number: Version 10.0.18362.720 Windows Terminal version (if applicable): Commit 2af722b43d3213858df156cc42c47e12341b836a # Steps to reproduce 1. Install the Japanese language pack. 2. Open a cmd shell in conhost (built from source). 3. Open the _Properties_ dialog and change the font to _MS Gothic_. 4. Switch the input language to _Japanese Microsoft IME_. 5. Make sure the IME option is initially _Half-Width Alphanumeric_. 6. Type `abcdefgh` 7. Move the cursor back to the start of the line. 8. Switch the IME option to _Hiragana_. 9. Type `sushi` then press <kbd>Esc</kbd> to close the IME. 10. Press <kbd>Esc</kbd> again to cancel the added text. # Expected behavior In the process of typing `sushi`, the latin characters `abcdef` will be temporarily overwritten, but after step 9, I'd expect the `ef` restored, and only the `abcd` overwritten with `すし`. After step 10, I'd expect to see all the latin characters restored. This is what those two steps looks like in the cmd shell in my current version of Windows. ![image](https://user-images.githubusercontent.com/4181424/82819714-bf4e0380-9e98-11ea-8cdc-25d57390f2b8.png) # Actual behavior After step 9, the `ef` characters have been replaced with blanks, and after step 10 they're still blank. ![image](https://user-images.githubusercontent.com/4181424/82819740-ca089880-9e98-11ea-854b-e77db8b9138d.png) What is weird is that I tried reverting the code back to the first public commit and it still seems to have that problem. I would have thought the cmd shell that shipped with Windows was much more recent than that. So I don't know if maybe it's just something about the build process that I'm getting wrong.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: starred/terminal#8569