feat(tokenizers): add experimental Korean tokenizer

[dayongkr]

Why add Korean support?

@orama/tokenizers currently supports experimental CJK tokenizers for Japanese and
Mandarin, but not Korean.
This makes Korean text search inconsistent compared to other CJK languages.

Adding a Korean tokenizer:

• enables proper Korean word segmentation for indexing/search
• keeps language support more consistent across CJK
• provides an official import path for Korean users (@orama/tokenizers/korean)

What changed

New tokenizer

• Added packages/tokenizers/src/korean.ts
  • Implements Korean tokenization using Intl.Segmenter("ko", { granularity: "word" })
  • Follows the same structure and behavior as existing Japanese/Mandarin tokenizers

Exports and package wiring

• Updated packages/tokenizers/src/index.ts
  • Added korean tokenizer export
• Updated packages/tokenizers/package.json
  • Added ./korean export entries (ESM/CJS types + runtime files)
  • Included Korean test in the package test script
  • Added korean keyword

Tests

• Added packages/tokenizers/tests/korean.test.ts
  • Validates Korean tokenization/search behavior with Korean city and university names
  • Keeps style aligned with existing Japanese/Mandarin tokenizer tests