[knowledge] Enable No ascii chars in knowledge

[jolestar]

Relates to

https://github.com/elizaOS/eliza/issues/2376

Risks

Low

Background

What does this PR do?

• The issue was fixed by updating the regex pattern to use Unicode properties:
• Changed from: /[^a-zA-Z0-9\s-_./:?=&]/g
• To: /[^\p{L}\p{N}\s-_./:?=&]/gu
• Added 'u' flag for Unicode support
• Used \p{L} to match any kind of letter from any language
• Used \p{N} to match any kind of numeric character
• The fix allows the function to properly handle multilingual content while maintaining the original sanitization goals.

What kind of change is this?

Bug fixes (non-breaking change which fixes an issue)

Documentation changes needed?

My changes do not require a change to the project documentation.

Testing

Where should a reviewer start?

Detailed testing steps

Add a unit test.

[wtfsayo]

please do it via env flag