@adonisjs/lucid-slugify

Use Lucid models to create URL-safe unique slugs and persist them in the database.

Introduction

Generating slugs is easy, but keeping them unique is hard. This package abstracts the hard parts and gives you a simple API to create and persist unique slugs to the database.

Lucid slugify exports the @slugify decorator, which you can use on the model fields to mark them as slugs and define the source fields from which the slug should be generated. Under the hood, the decorator registers beforeCreate and beforeUpdate hooks to compute the slug and persist it to the database.

In the following example, we mark the slug field as the slug and compute its value using the title field. We also use the dbIncrement strategy to keep slugs unique.

import { PostSchema } from '#database/schema'
import { slugify } from '@adonisjs/lucid-slugify'

export default class Post extends PostSchema {
  @slugify({
    strategy: 'dbIncrement',
    fields: ['title'],
  })
  declare slug: string
}

Installation and usage

You can install the @adonisjs/lucid-slugify package from the npm packages registry. Ensure your application uses @adonisjs/core@7 and @adonisjs/lucid@22.

npm i @adonisjs/lucid-slugify

yarn add @adonisjs/lucid-slugify

pnpm add @adonisjs/lucid-slugify

Once done, you can mark a field as a slug using the @slugify decorator. Make sure to specify the source field(s) from which the slug should be generated.

import { PostSchema } from '#database/schema'

// 👇 Import decorator
import { slugify } from '@adonisjs/lucid-slugify'

export default class Post extends PostSchema {
  // 👇 Use it on a column
  @slugify({
    fields: ['title'],
  })
  declare slug: string
}

The decorator computes the slug only when the column does not already have a value. If you assign a value to the slug column manually, your value is left untouched.

Uniqueness of slug

In the previous example, if two posts are created with the same title, they will have the same slug value.

This won't be a problem if you are the only author of your blog since you can always rename titles or might never write two articles with the same title.

However, if it's a community blog or forum, the chances of creating two or more posts with the same title are quite high.

You can use one of the following strategies to prevent duplicate slugs even when the titles are the same.

dbIncrement

The dbIncrement strategy performs a select query to find similar slugs and appends a counter when a duplicate slug is found. Given you have a database table with the following slugs:

• Creating a post with slug=hello-world will result in hello-world-6.
• Similarly, creating a post with slug=introduction-to-social-auth will result in introduction-to-social-auth-5.

+----+-----------------------------+-------------------------------+
| id | title                       | slug                          |
+----+-----------------------------+-------------------------------+
| 1  | Hello world                 | hello-world                   |
| 2  | Hello world                 | hello-world-5                 |
| 3  | Hello world                 | hello10world                  |
| 4  | Hello world                 | hello-10-world                |
| 5  | Introduction to social auth | introduction-to-social-auth   |
| 6  | Introduction to social auth | introduction-to-social-auth-4 |
| 7  | Hello world                 | hello-world-2                 |
| 8  | Hello world fanny           | hello-world-fanny             |
| 9  | Hello world                 | post-hello-world              |
| 10 | Hello world                 | post-11am-hello-world11       |
| 11 | Hello world                 | post-11am-hello-world         |
| 12 | Introduction to social auth | introdUction-to-Social-auTH-1 |
+----+-----------------------------+-------------------------------+

This strategy uses the same database connection as the model being persisted and works with PostgreSQL, MySQL (including 5.7), SQLite, and SQL Server. Using it with any other dialect throws the E_UNSUPPORTED_DB_DIALECT exception.

shortId

The shortId strategy appends a 10-character short id to the slug to make it unique. This strategy does not perform any additional database queries.

export default class Post extends PostSchema {
  @slugify({
    strategy: 'shortId',
    fields: ['title'],
  })
  declare slug: string
}

+----+-------------+------------------------+
| id | title       | slug                   |
+----+-------------+------------------------+
| 1  | Hello world | hello-world-yRPZZIWGgC |
+----+-------------+------------------------+

No strategy

When you do not specify a strategy, the slug is generated and persisted as-is, without any uniqueness guarantees. This is the default behavior and is suitable when the source fields are already unique (for example, a SKU or an email address).

Updating slugs

By default, slugs are not updated when you update a model instance, and this is how it should be when slugs are used to look up a record, as changing a slug will result in a broken URL.

However, if slugs are not primarily used to look up records, you may want to update them.

You can enable updates by using the allowUpdates flag.

export default class Post extends PostSchema {
  @slugify({
    strategy: 'dbIncrement',
    fields: ['title'],
    allowUpdates: true, // 👈
  })
  declare slug: string
}

When updates are enabled, the slug is recomputed only when one of the source fields has been modified. If you mutate the slug column manually, your value takes precedence and the slug is not recomputed.

You may also pass a function to allowUpdates to decide at runtime whether the slug should be updated for a given model instance.

@slugify({
  fields: ['title'],
  allowUpdates: (post) => post.status === 'draft',
})
declare slug: string

Null values and slug generation

The slugify decorator does not generate slugs when the value of one or more source fields is undefined or null.

Available options

Following is the list of available options accepted by the @slugify decorator.

Generating slugs from multiple fields

When you specify more than one field, their values are concatenated using the separator (in the order they are defined) and the result is slugified.

@slugify({
  fields: ['seriesName', 'title'],
  separator: '-',
})
declare slug: string

Customizing the transformer

The transformer is handy when a source field holds a rich data type that should not be stringified using the default rules.

@slugify({
  fields: ['title', 'publishedAt'],
  transformer: (post, field, value) => {
    if (field === 'publishedAt' && value instanceof DateTime) {
      return value.toFormat('yyyy-MM-dd')
    }

    return String(value)
  },
})
declare slug: string

Using custom strategies

Custom strategies can be used if you want to handle the uniqueness of slugs yourself. A strategy must implement the SlugifyStrategyContract interface.

import { LucidRow } from '@adonisjs/lucid/types/model'
import { SlugifyStrategyContract } from '@adonisjs/lucid-slugify/types'

export class MyCustomStrategy implements SlugifyStrategyContract {
  maxLengthBuffer: number = 0

  async makeSlugUnique(modelInstance: LucidRow, field: string, value: string) {
    return value
  }
}

The makeSlugUnique method receives the following arguments.

• modelInstance: Reference to the model instance that will be persisted in the database.
• field: The name of the field for which the unique slug will be created.
• value: The base value to convert to a unique value.

Once you have created the strategy, you can use it with the @slugify decorator by registering it as a factory function. The factory receives the model class, the property name, and the resolved decorator config, allowing you to build the strategy with the same configuration.

export default class Post extends PostSchema {
  @slugify({
    strategy: (model, propertyName, config) => {
      return new MyCustomStrategy()
    },
    fields: ['title'],
  })
  declare slug: string
}

The built-in strategy classes are also exported, so you can reuse or extend them from a factory function.

import { DbIncrementStrategy, ShortIdStrategy } from '@adonisjs/lucid-slugify/strategies'

Self creating slugs

The default implementation used by Lucid slugify for creating slugs works great with English words. However, if you are using Non-Latin alphabets, replace the implementation for creating slugs with a custom one.

You can override the static slugify method on the Slugifier class. The following code has to be executed only once (for example, inside a service provider).

import { Slugifier } from '@adonisjs/lucid-slugify'

/**
 * Make sure to install the "transliteration" package
 */
import { slugify } from 'transliteration'

Slugifier.slugify = function (value, options) {
  return slugify(value, {
    separator: options.separator,
    lowercase: options.lower,
  })
}

API reference

Entry points

slugify(config)

A property decorator that marks a Lucid model column as a slug. It registers beforeCreate and beforeUpdate hooks on the model to compute and persist the slug.

import { slugify } from '@adonisjs/lucid-slugify'

@slugify({ fields: ['title'], strategy: 'dbIncrement' })
declare slug: string

See Available options for the complete list of accepted config properties.

Slugifier

The class that powers slug generation. You usually do not interact with it directly — the @slugify decorator creates and manages a Slugifier instance per slug field for you.

The one member relevant to end-users is the static slugify method, which converts a string into a URL-safe slug. Override it to plug in a custom slugification routine (see Self creating slugs).

Slugifier.slugify(value: string, options: { separator: string; lower: boolean }): string

SlugifyStrategyContract

The interface that every strategy (built-in or custom) implements.

interface SlugifyStrategyContract {
  /**
   * The number of characters the strategy may add to the slug to make it
   * unique. The slugifier reserves this many characters when applying the
   * `maxLength` option, so the final value stays within the column size.
   */
  maxLengthBuffer: number

  /**
   * Convert the computed slug value into a unique value that is safe to
   * persist to the database.
   */
  makeSlugUnique(modelInstance: LucidRow, field: string, value: string): Promise<string> | string
}

errors

The package exports its exceptions under the errors namespace.

import { errors } from '@adonisjs/lucid-slugify'

errors.E_UNSUPPORTED_DB_DIALECT

Contributing

One of the primary goals of AdonisJS is to have a vibrant community of users and contributors who believe in the principles of the framework.

We encourage you to read the contribution guide before contributing to the framework.

Code of Conduct

To ensure that the AdonisJS community is welcoming to all, please review and abide by the Code of Conduct.

License

AdonisJS Lucid slugify is open-sourced software licensed under the MIT license.