Update: this proposal has been moved to stage 2. Continue discussion here

Body

Summary

Content collections unlocks some huge benefits, especially the DX benefits of type-checked markdown frontmatter. Adding support for relational content, i.e. a blog collection that references an authors collection, would be a huge step to allow more flexible content modeling

Background & Motivation

Taking a blog post as the example, there will likely be author information thats reused across multiple blog posts. To standardize updates when I'm changing something like an author's avatar image, it'd be nice to have an authors or people collection that each blog post can reference by slug.

Astro's content APIs could use that for type-checking to make sure a known author slug is used and prefetch the data so blog.data.author would be the actual data rather than a string that needs to be passed back into getEntryBySlug

Goals

Minimum goal - type-safe references by slug. i.e. The blog collection would validate that blog.data.author is a valid slug, but wouldn't transform or preload that data at all

Bonus points - related content is pre-fetched and the slug is replaced with the related data. i.e. once loaded, blog.data.author would be an object with the author's parsed content instead of the author's slug

Example

Hacky userland experiment

---
title: My blog post
author: tony
---

## Insert awesome post here

---
name: Tony
avatar: uploads/avatar.png
---

import { image, defineCollection, relation, z } from 'astro:content'

export const collections = {
  blog: {
    schema: z.object({
      title: z.string(),
      author: relation("authors").default("tony")
    })
  },
  authors: {
    schema: z.object({
      name: z.string(),
      avatar: image()
    })
  }
}