prisma-slug

A slugification middleware for Prisma. It generates slugs for your models by using other model attributes with logic that you can define. It's bundled with the excellent slugify package and comes with reasonable defaults to let you define your Prisma schema without worrying about how you're going to generate URL-safe slugs.

Getting Started

Install the library:

yarn add prisma-slug

Then, include it in the file you use to instantiate your Prisma client:

import { PrismaClient } from '@prisma/client'
import { PrismaSlug } from 'prisma-slug'

const db = new PrismaClient()

db.use(PrismaSlug())

export default db

In your Prisma schema, define a slug field on models that have a name, and when new records are saved your slug field will be automatically populated.

model User {
  id    Int     @id @default(autogenerated())
  name  String  @unique
  slug  String  @unique
}

Usage

The PrismaSlug() function outputs a middleware that will convert model fields called name into a URL-safe slug and persist that to a field called slug on the same model. You can customize how this works by passing options into the function. For example, if you wanted to convert title fields as well as name, you'd configure the middleware like so:

db.use(
  PrismaSlug({
    source(params) {
      return params.args.data.name ?? params.args.data.title
    },
  })
)

You can also configure the function used to generate the slug:

import slug from 'slug'

db.use(
  PrismaSlug({
    slugify: slug,
  })
)

Both the source() and slugify() functions can return promises as well.

For a full list of options, view the documentation at https://tubbo.github.io/prisma-slug/modules.html#PrismaSlugOptions.

Configuring Slugify Options

To configure slugify, define PrismaSlug's slugify() function and pass in the options like so:

import slugify from 'slugify'

db.use(
  PrismaSlug({
    slugify(value) {
      return slugify(value, { trim: false })
    },
  })
)

Unique Slugs

Customize the slugify() function to keep generating slugs until it finds one that's unique:

import { camelCase } from 'camel-case'
import slugify from 'slugify'

db.use(
  PrismaSlug({
    async slugify(source, params) {
      const method = camelCase(params.model)
      const collection = db[method]
      let slug = slugify(source)
      let attempt = 0

      while ((await collection.count({ where: { slug } })) > 0) {
        attempt += 1
        slug = `${slug}-${attempt}`
      }

      return slug
    },
  })
)

Development

This project uses Yarn Plug'n'Play and Zero-Installs, meaning you don't need to run yarn install to begin developing. However, depending on your editor you may need to install an editor SDK.

To run tests:

yarn test

To run lint checks:

yarn lint

To run type checks:

yarn types

You can build the library by running:

yarn build

And tear it down:

yarn clean

Upon committing, your code will be automatically formatted and your commit message checked to make sure it follows the conventional commits standard. All pull requests have tests, lint checks, formatting checks, type checks, and package security checks run against them.