N
Nuxt•4mo ago
Alex

Nuxt content sqlite modify #nuxtcontent

Hi I have been playing with Nuxt content and I wonder How could a I modify the sqlite dump. I want to save vector embeddings into the generated sqlite.
2 Replies
kapa.ai
kapa.ai•4mo ago
Hi @Alex. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave: ---------------------- Nuxt Content v3 uses an SQLite database (by default at .data/content/contents.sqlite) to store and query your content efficiently. However, the knowledge sources do not provide any official or documented way to directly modify the generated SQLite dump or to add custom fields (such as vector embeddings) directly into the SQLite database outside of the Nuxt Content schema and build process. If you want to store additional data (like vector embeddings) alongside your content, the recommended approach is to extend your collection schema in content.config.ts to include a new field for your embeddings. For example: 
import { defineContentConfig, defineCollection, z } from '@nuxt/content'

export default defineContentConfig({
  collections: {
    docs: defineCollection({
      type: 'page',
      source: '**/*.md',
      schema: z.object({
        embedding: z.array(z.number()).optional() // Add your vector field here
      })
    })
  }
})
import { defineContentConfig, defineCollection, z } from '@nuxt/content'

export default defineContentConfig({
  collections: {
    docs: defineCollection({
      type: 'page',
      source: '**/*.md',
      schema: z.object({
        embedding: z.array(z.number()).optional() // Add your vector field here
      })
    })
  }
})
You would then need to ensure that your content files include this field, or use Nuxt Content hooks to programmatically add or update the embedding field during the build process. For example, you could use the content:file:afterParse hook to compute and attach embeddings: 
export default defineNuxtConfig({
  hooks: {
    'content:file:afterParse'(ctx) {
      // ctx.content.embedding = ... // Compute and assign your embedding here
    }
  }
})
export default defineNuxtConfig({
  hooks: {
    'content:file:afterParse'(ctx) {
      // ctx.content.embedding = ... // Compute and assign your embedding here
    }
  }
})
This way, the embeddings are included in the content that gets stored in the SQLite database, and you can query them using Nuxt Content's APIs. Directly modifying the SQLite dump outside of Nuxt Content's lifecycle is not documented or supported, and may lead to inconsistencies or data loss, as the database is regenerated from your content sources and schema during the build process. If you need to inspect or debug the database, you can use tools like the SQLite VS Code extension, but for persistent custom data, always use the schema and hooks approach provided by Nuxt Content Debugging tools. In summary:
- Extend your schema to include the embedding field. - Use hooks to populate it. - Do not modify the SQLite dump directly, as it is managed by Nuxt Content and may be overwritten during builds. If you need more advanced or custom database logic, you may need to manage a separate database outside of Nuxt Content for your embeddings. The knowledge sources do not provide further details on this scenario. ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ----------------------
kapa.ai
kapa.ai•4mo ago
<@1244273611754438708> Kapa.ai is still learning and improving, please let me know how I did by reacting below.

Did you find this page helpful?