[yashranaway]

this is the unicode caption segmentation case from the verify expectations. caption chunking avoids ending a caption on a weak one letter word, but it checked the length in bytes, so a one letter word outside ascii like the french "à" counted as two letters and a caption could end on it.

now it counts characters, which is what the reference model already does. the tests run "à", a decomposed "à" and a cjk character through both the chunk end and short tail paths, the xfail is removed, and the full ascii and unicode segmentation oracles pass.

🤖 Generated with Claude Code

Confidence Score: 5/5

This PR appears safe to merge.

Summary

Changes GPUI caption boundaries to count Unicode characters instead of bytes.

• Desktop captions keep one-letter Unicode words with nearby words.

No actionable issues found. Tests were inspected, not run.

_{Reviews (1) · Last reviewed commit: "fix(desktop-gpui): count Unicode charact..." · Reviewed by Greptile}

[vercel]

@yashranaway is attempting to deploy a commit to the Cap Software Inc Team on Vercel.

A member of the Team first needs to authorize it.

[yashranaway]

@greptileai review