Database Batch Mutations

The Grafbase Database provides mutations to create, update, and delete entries for any configured model. However, if you wanted to execute multiple mutations, you needed to conduct separate mutations.

We're excited to announce that Grafbase Database now supports batch mutations!

Batch mutations are engineered to function in a manner akin to their single mutation equivalents.

The examples below use the following schema:

import { g } from '@grafbase/sdk'

g.model('Post', {
  title: g.string(),
  slug: g.string().unique(),
})

This will automatically generate additional GraphQL mutations for all database models:

mutation {
  postUpdateMany(input: [PostUpdate!]!): PostUpdateManyPayload
  postCreateMany(input: [PostCreate!]!): PostCreateManyPayload
  postDeleteMany(input: [PostDelete!]!): PostDeleteManyPayload
}

To create multiple entries at once, passed an array objects to input:

mutation {
  postCreateMany(
    input: [
      { input: { title: "Batch", slug: "batch" } }
      { input: { title: "Operations", slug: "operations" } }
    ]
  ) {
    postCollection {
      id
      title
      slug
    }
  }
}

To update multiple entries at once, passed an array objects to input that contains by and new input values:

mutation {
  postUpdateMany(
    input: [
      { by: { id: "..." }, input: { title: "Batch!" } }
      { by: { slug: "operations" }, input: { title: "Operations!" } }
    ]
  ) {
    postCollection {
      id
      title
      slug
    }
  }
}

To delete multiple entries at once, passed an array objects to input that contains the by values:

mutation {
  postDeleteMany(input: [{ by: { id: "..." } }, { by: { slug: "batch" } }]) {
    deletedIds
  }
}

• Fixes an issue introduced in 0.26.1 that would cause an infinite reinstallation of dependencies if package.json was located in the grafbase/ directory.
• Change the working directory in which resolver builds take place allowing them to import files relative to the project directory.
• Fixes allOf schema support when using the @openapi connector.
• JSON scalars will now accept any literal Value in documents.
• Added the missing description field to __Schema in introspection.
• Fixes the __type field in introspection.
• The @graphql connector now pulls in an objects interfaces correctly.
• Type conditions on inline & fragment spreads now match interfaces correctly.
• Type conditions on unions will now match if the condition is the union type.
• For update mutations, like postUpdate, nullable fields can now be set to null. Previously, those updates would be ignored.
• Improved support for v2 of the OpenAPI specification.

We'd love to hear your feedback and ideas, so join us on Discord.