Laravel Factories and Seeders: All You Need to Know

Seeding data in Laravel is quite simple, but also has a lot of caveats, less-known functions and use-cases. Both Seeders and Factories have so much "hidden" power that we've compiled this HUGE tutorial, with screenshots from real scenarios.

You can treat it as "Laravel docs on steroids": same functions explained in a more practical visual way, with a better flow to read step-by-step.

Among other topics, in this tutorial, we will answer questions, like:

• How do you create model factories with complex relationships?
• What's the best way to test with large datasets?
• How can you make one factory field depend on another?
• Which factory methods are powerful but rarely used?
• How do you make test data look realistic?

So, let's dive in!

Table of Contents

1. Seeding Approaches

• Using the DatabaseSeeder Class
• Using Separate Seeder Files
• Running Individual Seeders

2. Understanding Laravel Factories

• Creating a Factory
• Using Factories in Seeders
• Other Factory Use Cases

3. Factory Methods and Features

• Factory States
• Persistent vs Non-Persistent Factories
• Overriding Factory Values
• Sequence Factories
• Lifecycle Hooks: afterMaking and afterCreating

4. Working with Relationships

• BelongsTo Relationships
• HasMany Relationships
• ManyToMany Relationships
• Polymorphic Relationships
• Reusing Models with recycle()

5. Tips and Best Practices

• Accessing Other Attributes
• Seeding Unique Values
• Lesser-Known Factory Methods
• Matching Real-World Data
• Testing with Large Datasets

1. Seeding Approaches

In Laravel, we have a couple of ways to seed our Database:

• Using the DatabaseSeeder class
• Using separate files for each Seeder and calling them from the DatabaseSeeder class

Let's explore each approach.

Using the DatabaseSeeder Class

Laravel provides a DatabaseSeeder class by default:

database/seeders/DatabaseSeeder.php

// ...
 
class DatabaseSeeder extends Seeder
{
    /**
     * Seed the application's database.
     */
    public function run(): void
    {
        // User::factory(10)->create();
 
        User::factory()->create([
            'name' => 'Test User',
            'email' => '[email protected]',
        ]);
    }
}

Which already contains a seed for Test User. But what if we want to add more things? Let's take our Currency example:

database/seeders/DatabaseSeeder.php

class DatabaseSeeder extends Seeder
{
    /**
     * Seed the application's database.
     */
    public function run(): void
    {
        User::factory(10)->create();
 
        User::factory()->create([
            'name' => 'Test User',
            'email' => '[email protected]',
        ]);
 
        $currencies = [
            [
                'name' => 'US Dollar',
                'code' => 'USD',
            ],
            [
                'name' => 'Euro',
                'code' => 'EUR',
            ],
            [
                'name' => 'British Pound',
                'code' => 'GBP',
            ],
        ];
 
        foreach ($currencies as $currency) {
            Currency::create($currency);
        }
    }
}

Let's try to run the Seeder:

php artisan db:seed
# Or (if you want to refresh the database)
php artisan migrate:fresh --seed

This screen shows that our Seeder ran successfully. But that's all the information we get. We don't know how long it took to run, which can lead us to think that the Seeder is stuck. This is because the Seeder is running synchronously, and we have no way to know how long it will take.

Of course, this is not the only problem. When the system grows - and it will - the DatabaseSeeder class will become a mess. It will be hard to maintain and understand. We should consider using separate files for each Seeder.

Using Separate Seeder Files

Let's take a look at separate files for each Seeder. In this case, we will move the Currency Seeder and User Seeder to separate files:

php artisan make:seeder CurrencySeeder
php artisan make:seeder UserSeeder

Let's fill the CurrencySeeder with the Currency data:

database/seeders/CurrencySeeder.php

class CurrencySeeder extends Seeder
{
    /**
     * Run the database seeds.
     */
    public function run(): void
    {
        $currencies = [
            [
                'name' => 'US Dollar',
                'code' => 'USD',
            ],
            [
                'name' => 'Euro',
                'code' => 'EUR',
            ],
            [
                'name' => 'British Pound',
                'code' => 'GBP',
            ],
        ];
 
        foreach ($currencies as $currency) {
            Currency::create($currency);
        }
    }
}

And the UserSeeder with the User data:

database/seeders/UserSeeder.php

class UserSeeder extends Seeder
{
    /**
     * Run the database seeds.
     */
    public function run(): void
    {
        User::factory(10)->create();
    }
}

Finally, let's call these Seeders from the DatabaseSeeder class:

database/seeders/DatabaseSeeder.php

class DatabaseSeeder extends Seeder
{
    /**
     * Seed the application's database.
     */
    public function run(): void
    {
        User::factory()->create([
            'name' => 'Test User',
            'email' => '[email protected]',
        ]);
 
        $this->call([
            CurrencySeeder::class,
            UserSeeder::class
        ]);
    }
}

Now, let's run the Seeder:

php artisan db:seed
# Or (if you want to refresh the database)
php artisan migrate:fresh --seed

As you can see, we have a detailed report of which Seeders ran and how long they took. This is a much better approach than using the DatabaseSeeder class.

Running Individual Seeders

Another advantage of using separate files is that we can run them individually. This can be useful when we want to run only one Seeder:

php artisan db:seed --class=UserSeeder

This will run only the UserSeeder and not the CurrencySeeder:

This is a great way to test and debug Seeders or in testing environments where we need to run only a specific Seeder.

2. Understanding Laravel Factories

Let's talk about Factories in Laravel. Factories are great if we want to generate fake data for our Database:

Factory Example

public function definition(): array
{
    return [
        'name' => fake()->name(),
        'email' => fake()->unique()->safeEmail(),
        'email_verified_at' => now(),
        'password' => static::$password ??= Hash::make('password'),
        'remember_token' => Str::random(10),
    ];
}

This Factory will generate a User with a random name and a unique email. But how does it work?

Creating a Factory

Let's look at how Factories are Created. For this, we need...