Issues with indexing Japanese text inside 【 】 brackets in filenames

Hi everyone,

I noticed an interesting behavior in the project search (Your Videos page) when using Japanese full-width brackets 【 】 in filenames, and I'd love to hear if other users have encountered the same thing — and how you work around it.

## What I observed

I have a set of filenames that start with a bracketed category tag like:
  【TagWord】MainTitleHere

I ran these queries against the folder containing those files (TagWord = a 4-character Japanese word inside the brackets):

| Query | Hits |
|---|---|
| First 2 characters of the bracketed word | 0 |
| Full bracketed word (no brackets) | 0 |
| Bracketed word + closing 】 only | 0 |
| Full 【TagWord】 (both brackets included) | All matching files ✅ |
| A Japanese word that appears OUTSIDE any brackets in other filenames | Hits normally ✅ |

So partial-match search works perfectly for Japanese text OUTSIDE the brackets, but content INSIDE 【 】 seems to be excluded from the index. Only the full bracketed phrase works — likely a fallback to raw filename substring match when special characters are in the query.

## Questions

1. Is this intentional? (e.g., 【 】 treated as a "decoration tag" and stripped during indexing)
2. Has anyone else using Japanese filenames noticed this?
3. Any recommended workaround other than removing the brackets from filenames?

For context, many Japanese creators use 【 】 as a visual category prefix (very common convention in Japan), so this affects a lot of files in practice.

Thanks!