Skip to content

NouanceLabs/payload-meilisearch

Repository files navigation

Payload Meilisearch Plugin (ALPHA)

Expect breaking changes.

A plugin for Payload CMS to connect Meilisearch and Payload.

Roadmap to stable release:

  • Sync collections on create and update
  • Delete collections
  • Support all field types of Payload (in progress)
  • Support for Payload's draft system (planned)
  • Support for nested fields (planned)
  • Support Meilisearch index options for filtering and sorting (planned)

Installation

  yarn add @nouance/payload-meilisearch
  # OR
  npm i @nouance/payload-meilisearch

Basic Usage

In the plugins array of your Payload config, call the plugin with options:

import { buildConfig } from "payload/config";
import meilisearchPlugin from "@nouance/payload-meilisearch";

const config = buildConfig({
  plugins: [
    meilisearchPlugin({
      host: process.env.MEILISEARCH_HOST,
      apiKey: process.env.MEILISEARCH_API_KEY,
    }),
  ],
});

export default config;

Options

  • host

    Required. Your Meilisearch host URL.

  • apiKey

    Required. Your Meilisearch API key. Must have permissions to read, create and delete indexes and documents.

  • sync

    Required. An array of sync configs. This will automatically configure a sync between Payload collections and Meilisearch indexes. See sync for more details.

  • logs

    Optional. When true, logs sync events to the console as they happen.

Sync

This option will sync collection's data and their fields to your Meilisearch instance. Each collection slug maps to a unique index in Meilisearch.

import { buildConfig } from "payload/config";
import meilisearchPlugin from "@nouance/payload-meilisearch";

const config = buildConfig({
  plugins: [
    meilisearchPlugin({
      host: process.env.MEILISEARCH_HOST,
      apiKey: process.env.MEILISEARCH_API_KEY,
      sync: [
        {
          collection: "posts",
          fields: [
            {
              name: "title",
            },
            {
              name: "publishedDate",
              alias: "publicationDate",
            },
          ],
        },
      ],
    }),
  ],
});

export default config;

Alias

You can optionally set an alias that maps the field to a different name in Meilisearch's index in case you don't want to expose inner field names or want to simplify it.

Alias the ID

In Meilisearch, every document must have an ID, by default we're setting this to be the document's ID, however you can change it to any other field by making your alias id. Note that this has to be unique and a valid format (eg. no spaces), so we recommend leaving it as default or applying a transformer function to make sure your data format is always correct.

Transformers

On each field you can apply a transformer function that takes in the field value as it is from the doc object and then update operation's full doc object so that you can modify the values before the collection is indexed. This is useful for dealing with complex fields like RichText where you might want to serialise it into a string so it's easily searchable.

Example transformer to set the name of a category as the ID of the document:

{
  name: "name",
  alias: "id",
  transformer: (doc) => {
    const name: string = doc.name;
    return name.replaceAll(" ", "-").toLocaleLowerCase();
  },
},

Curently supported field types

  • Code
  • Date
  • Email
  • JSON
  • Number
  • Point
  • Radio group
  • Select
  • Text
  • Textarea

Development

For development purposes, there is a full working example of how this plugin might be used in the demo of this repo. This demo can be developed locally using any Meilisearch Cloud account, you just need a working API key. Then:

git clone [email protected]:NouanceLabs/payload-meilisearch.git \
  cd payload-meilisearch && yarn \
  cd demo && yarn \
  cp .env.example .env \
  vim .env \ # add your Meilisearch creds to this file
  yarn dev