Add hybrid search mode and unify vector, hybrid, and full-text search under the same API

[micheleriva]

This PR unifies full-text, vector, and hybrid searches under the same search method.

import { create, insert, search } from '@orama/orama'

const db = await create({
  schema: {
    text: 'string',
    embedding: 'vector[1536]'
  }
})

await insert(db, { ... })
await insert(db, { ... })
await insert(db, { ... })

const fulltextResults = await search(db, {
  term: 'Super Mario games'
})

const hybridResults = await search(db, {
  term: 'Super Mario games',
  mode: 'hybrid'
})

const vectorResults = await search(db, {
  term: 'Super Mario games',
  mode: 'vector'
})

The text-to-embedding part is managed by Orama Cloud under the Secure Proxy API, which lets you connect to OpenAI or any other service from your front-end directly.

This PR also introduces:

1. Facets for vector search
2. GroupBy operations for vector search
3. "where" filters operations for vector search

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
orama-docs	✅ Ready (Inspect)	Visit Preview	✅ 4 resolved	Jan 14, 2024 1:51pm