enh: improve fulltext search

[benjaminfrueh]

Fixes #2559

📝 Summary

Collectives fulltext search improvements with a focus on sentence search, stem distance limiting and result ranking.

Changes

• AND search for multi-token queries, only return documents containing all search terms
• improved multi-token ranking using a match type scoring across all matched tokens
• stem distance limit, stems too far from the original token are not indexed anymore, reducing noise
• improve snippet preview, shows the exact phrase if found, otherwise the section containing the most matched tokens
• prefix match always runs alongside exact match for the last token, to prevent short tokens from blocking results

🏁 Checklist

• Code is properly formatted (npm run lint / npm run stylelint / composer run cs:check)
• Sign-off message is added to all commits
• Tests (unit, integration and/or end-to-end) passing and the changes are covered with tests
• Documentation (README or documentation) has been updated or is not required

🤖 AI (if applicable)

• The content of this PR was partly or fully generated using AI tools
• The AI-generated content was reviewed, comprehended and tested by a human

[max-nextcloud]

All in all looks good. Just some possible improvements. Not sure if for now or for later.

[max-nextcloud]

Tested locally. Works great! Thanks a lot!