Easy creation of slugs for your Eloquent models in Laravel.
NOTE: These instructions are for the latest version of Laravel.
If you are using an older version, please install a version of the package that correlates to your Laravel version.
- Background: What is a slug
- Installation
- Updating your Eloquent Models
- Usage
- The SlugService Class
- Events
- When Is A Model Slugged?
- Configuration
- Extending Sluggable
- SluggableScopeHelpers Trait
- Route Model Binding
- Bugs, Suggestions, Contributions and Support
- Copyright and License
A slug is a simplified version of a string, typically URL-friendly. The act of "slugging" a string usually involves converting it to one case, and removing any non-URL-friendly characters (spaces, accented letters, ampersands, etc.). The resulting string can then be used as an identifier for a particular resource.
For example, if you have a blog with posts, you could refer to each post via the ID:
http://example.com/post/1
http://example.com/post/2
... but that's not particularly friendly (especially for SEO). You probably would prefer to use the post's title in the URL, but that becomes a problem if your post is titled "My Dinner With André & François", because this is pretty ugly too:
http://example.com/post/My+Dinner+With+Andr%C3%A9+%26+Fran%C3%A7ois
The solution is to create a slug for the title and use that instead. You might want
to use Laravel's built-in Str::slug()
method to convert that title into something
friendlier:
http://example.com/post/my-dinner-with-andre-francois
A URL like that will make users happier (it's readable, easier to type, etc.).
For more information, you might want to read this description on Wikipedia.
Slugs tend to be unique as well. So if you write another post with the same title, you'd want to distinguish between them somehow, typically with an incremental counter added to the end of the slug:
http://example.com/post/my-dinner-with-andre-francois
http://example.com/post/my-dinner-with-andre-francois-1
http://example.com/post/my-dinner-with-andre-francois-2
This keeps the URLs unique.
The Eloquent-Sluggable package for Laravel aims to handle all of this for you automatically, with minimal configuration.
Depending on your version of Laravel, you should install a different version of the package.
NOTE: As of version 6.0, the package's version should match the Laravel version.
Laravel Version | Package Version |
---|---|
8.0 | ^8.0 |
7.0 | ^7.0 |
6.0 | ^6.0 |
5.8 | 4.8.* |
5.7 | 4.6.* |
5.6 | 4.5.* |
5.5 | 4.3.* |
5.4 | 4.2.* |
Older versions of Laravel can use older versions of the package, although they are no longer supported or maintained. See CHANGELOG.md and UPGRADING.md for specifics, and be sure that you are reading the correct README.md for your version (Github displays the version in the master branch by default, which might not be what you want).
-
Install the package via Composer:
$ composer require cviebrock/eloquent-sluggable
The package will automatically register its service provider.
-
Optionally, publish the configuration file if you want to change any defaults:
php artisan vendor:publish --provider="Cviebrock\EloquentSluggable\ServiceProvider"
Your models should use the Sluggable trait, which has an abstract method sluggable()
that you need to define. This is where any model-specific configuration is set
(see Configuration below for details):
use Cviebrock\EloquentSluggable\Sluggable;
class Post extends Model
{
use Sluggable;
/**
* Return the sluggable configuration array for this model.
*
* @return array
*/
public function sluggable(): array
{
return [
'slug' => [
'source' => 'title'
]
];
}
}
Of course, your model and database will need a column in which to store the slug.
You can use slug
or any other appropriate name you want; your configuration array
will determine to which field the data will be stored. You will need to add the
column (which should be NULLABLE
) manually via your own migration.
That's it ... your model is now "sluggable"!
Saving a model is easy:
$post = Post::create([
'title' => 'My Awesome Blog Post',
]);
So is retrieving the slug:
echo $post->slug;
NOTE: that if you are replicating your models using Eloquent's
replicate()
method, the package will automatically re-slug the model afterwards to ensure uniqueness.
$post = Post::create([
'title' => 'My Awesome Blog Post',
]);
// $post->slug is "my-awesome-blog-post"
$newPost = $post->replicate();
// $newPost->slug is "my-awesome-blog-post-1"
NOTE: empty strings, non-strings or other "odd" source values will result in different slugs:
Source Value | Resulting Slug |
---|---|
string | string |
empty string | no slug will be set |
null |
no slug will be set |
0 |
"0" |
1 |
"1" |
false |
"0" |
true |
"1" |
(The above values would be subject to any unique or other checks as well.)
All the logic to generate slugs is handled
by the \Cviebrock\EloquentSluggable\Services\SlugService
class.
Generally, you don't need to access this class directly, although there is one static method that can be used to generate a slug for a given string without actually creating or saving an associated model.
use \Cviebrock\EloquentSluggable\Services\SlugService;
$slug = SlugService::createSlug(Post::class, 'slug', 'My First Post');
This would be useful for Ajax-y controllers or the like, where you want to show a user what the unique slug would be for a given test input, before actually creating a model. The first two arguments to the method are the model and slug field being tested, and the third argument is the source string to use for testing the slug.
You can also pass an optional array of configuration values as the fourth argument. These will take precedence over the normal configuration values for the slug field being tested. For example, if your model is configured to use unique slugs, but you want to generate the "base" version of a slug for some reason, you could do:
$slug = SlugService::createSlug(Post::class, 'slug', 'My First Post', ['unique' => false]);
Currently, the model is slugged on Eloquent's saving
event.
This means that the slug is generated before any new data is
written to the database.
For new models, this means that the primary key has not yet been set, so it could not be used as part of the slug source, e.g.:
public function sluggable(): array
{
return [
'slug' => [
'source' => ['title', 'id']
]
];
}
$model->id
is null
before the model is saved. The benefit of hooking into
the saving
event, however, is that we only needed to make one database
query to save all the model's data, including the slug.
Optional, the model can be slugged on Eloquent's saved
event.
This means that all the other model attributes will have already been
persisted to the database and are available for use as slug sources.
So the above configuration would work. The only drawback is that
saving the model to the database requires one extra query: the first one
to save all the non-slug fields, and then a second one to update just
the slug field.
This behaviour is a breaking change, and likely won't affect most users
(unless you are doing some pre-saving validation on a model's slug field).
We feel the benefits outweigh the drawbacks, and so this will likely become
the new default behaviour in a future major release of the package.
Although, to make the transition easier, you can configure this behaviour
via the sluggableEvent
method the trait provides:
public function sluggableEvent(): string
{
/**
* Default behaviour -- generate slug before model is saved.
*/
return SluggableObserver::SAVING;
/**
* Optional behaviour -- generate slug after model is saved.
* This will likely become the new default in the next major release.
*/
return SluggableObserver::SAVED;
}
Keep in mind that you will need to use SluggableObserver::SAVED
if you want
to use your model's primary key as part of the source fields for your slugs.
Sluggable models will fire two Eloquent model events: "slugging" and "slugged".
The "slugging" event is fired just before the slug is generated. If the callback
from this event returns false
, then the slugging is not performed. If anything
else is returned, including null
, then the slugging will be performed.
The "slugged" event is fired just after a slug is generated. It won't be called
in the case where the model doesn't need slugging (as determined by the needsSlugging()
method).
You can hook into either of these events just like any other Eloquent model event:
Post::registerModelEvent('slugging', static function($post) {
if ($post->someCondition()) {
// the model won't be slugged
return false;
}
});
Post::registerModelEvent('slugged', static function($post) {
Log::info('Post slugged: ' . $post->getSlug());
});
Configuration was designed to be as flexible as possible. You can set up defaults for all of your Eloquent models, and then override those settings for individual models.
By default, global configuration is set in the config/sluggable.php
file.
If a configuration isn't set, then the package defaults are used.
Here is an example configuration, with all the default settings shown:
return [
'source' => null,
'method' => null,
'onUpdate' => false,
'separator' => '-',
'unique' => true,
'uniqueSuffix' => null,
'firstUniqueSuffix' => 2,
'includeTrashed' => false,
'reserved' => null,
'maxLength' => null,
'maxLengthKeepWords' => true,
'slugEngineOptions' => [],
];
For individual models, configuration is handled in the sluggable()
method that you
need to implement. That method should return an indexed array where the keys represent
the fields where the slug value is stored and the values are the configuration for that
field. This means you can create multiple slugs for the same model, based on different
source strings and with different configuration options.
public function sluggable(): array
{
return [
'title-slug' => [
'source' => 'title'
],
'author-slug' => [
'source' => ['author.lastname', 'author.firstname'],
'separator' => '_'
],
];
}
This is the field or array of fields from which to build the slug. Each $model->field
is concatenated (with space separation) to build the sluggable string. These can be
model attributes (i.e. fields in the database), relationship attributes, or custom getters.
To reference fields from related models, use dot-notation. For example, the slug for the following book will be generated from its author's name and the book's title:
class Book extends Eloquent
{
use Sluggable;
protected $fillable = ['title'];
public function sluggable(): array
{
return [
'slug' => [
'source' => ['author.name', 'title']
]
];
}
public function author(): \Illuminate\Database\Eloquent\Relations\BelongsTo
{
return $this->belongsTo(Author::class);
}
}
...
class Author extends Eloquent
{
protected $fillable = ['name'];
}
An example using a custom getter:
class Person extends Eloquent
{
use Sluggable;
public function sluggable(): array
{
return [
'slug' => [
'source' => 'fullname'
]
];
}
public function getFullnameAttribute(): string
{
return $this->firstname . ' ' . $this->lastname;
}
}
If source
is empty, false or null, then the value of $model->__toString()
is used
as the source for slug generation.
Defines the method used to turn the sluggable string into a slug. There are three possible options for this configuration:
-
When
method
is null (the default setting), the package uses the default slugging engine -- cocur/slugify -- to create the slug. -
When
method
is a callable, then that function or class method is used. The function/method should expect two parameters: the string to process, and a separator string. For example, to use Laravel'sStr::slug
, you could do:
'method' => ['Illuminate\\Support\\Str', 'slug'],
- You can also define
method
as a closure (again, expecting two parameters):
'method' => static function(string $string, string $separator): string {
return strtolower(preg_replace('/[^a-z]+/i', $separator, $string));
},
Any other values for method
will throw an exception.
For more complex slugging requirements, see Extending Sluggable below.
By default, updating a model will not try and generate a new slug value. It is assumed that once your slug is generated, you won't want it to change (this may be especially true if you are using slugs for URLs and don't want to mess up your SEO mojo).
If you want to regenerate one or more of your model's slug fields, you can set those fields to null or an empty string before the update:
$post->slug = null;
$post->update(['title' => 'My New Title']);
If this is the behaviour you want every time you update a model, then set the onUpdate
option to true.
This defines the separator used when building a slug, and is passed to the method
defined above. The default value is a hyphen.
This is a boolean defining whether slugs should be unique among all models of the given type.
For example, if you have two blog posts and both are called "My Blog Post", then they
will both sluggify to "my-blog-post" if unique
is false. This could be a problem, e.g.
if you use the slug in URLs.
By setting unique
to true, then the second Post model will sluggify to "my-blog-post-1".
If there is a third post with the same title, it will sluggify to "my-blog-post-2"
and so on. Each subsequent model will get an incremental value appended to the end
of the slug, ensuring uniqueness.
If you want to use a different way of identifying uniqueness (other than auto-incrementing
integers), you can set the uniqueSuffix
configuration to a function or callable that
generates the "unique" values for you.
The function should take four parameters:
- the base slug (i.e. the non-unique slug)
- the separator string
- an
\Illuminate\Support\Collection
of all the other slug strings that start with the same slug - the first suffix to use (for the first slug that needs to be made unique) You can then do whatever you want to create a new suffix that hasn't been used by any of the slugs in the collection. For example, if you wanted to use letters instead of numbers as a suffix, this is one way to achieve that:
'uniqueSuffix' => static function(string $slug, string $separator, Collection $list, $firstSuffix): string
{
$size = count($list);
return chr($size + 96);
}
When adding a unique suffix, we start counting at "2", so that the list of generated slugs would look something like:
my-unique-slug
my-unique-slug-2
my-unique-slug-3
- etc.
If you want to start counting at a different number (or pass a different value
into your custom uniqueSuffix
function above), then you can define it here.
NOTE: Prior versions of the package started with a unique suffix of
1
. This was switched to2
in version 8.0.5, as it's a more "intuitive" suffix value to attach to the second slug.
Setting this to true
will also check deleted models when trying to enforce uniqueness.
This only affects Eloquent models that are using the
softDelete feature. Default is false
,
so soft-deleted models don't count when checking for uniqueness.
An array of values that will never be allowed as slugs, e.g. to prevent collisions
with existing routes or controller methods, etc.. This can be an array, or a closure
that returns an array. Defaults to null
: no reserved slug names.
Setting this to a positive integer will ensure that your generated slugs are restricted to a maximum length (e.g. to ensure that they fit within your database fields). By default, this value is null and no limit is enforced.
NOTE: If
unique
is enabled (which it is by default), and you anticipate having several models with the same slug, then you should set this value to a few characters less than the length of your database field. The reason why is that the class will append "-2", "-3", "-4", etc., to subsequent models in order to maintain uniqueness. These incremental extensions aren't included in part of themaxLength
calculation.
If you are truncating your slugs with the maxLength
setting, than you probably
want to ensure that your slugs don't get truncated in the middle of a word. For
example, if your source string is "My First Post", and your maxLength
is 10,
the generated slug would end up being "my-first-p", which isn't ideal.
By default, the maxLengthKeepWords
value is set to true which would trim the
partial words off the end of the slug, resulting in "my-first" instead of "my-first-p".
If you want to keep partial words, then set this configuration to false.
When method
is null (the default setting), the package uses the default slugging
engine -- cocur/slugify -- to create the slug.
If you want to pass a custom set of options to the Slugify constructor when the engine
is instantiated, this is where you would define that.
See the documentation
for Slugify for what those options are. Also, look at
customizeSlugEngine for other ways to customize Slugify
for slugging.
The package supports a really short configuration syntax, if you are truly lazy:
public function sluggable(): array
{
return ['slug'];
}
This will use all the default options from config/sluggable.php
, use the model's
__toString()
method as the source, and store the slug in the slug
field.
Sometimes the configuration options aren't sufficient for complex needs (e.g. maybe the uniqueness test needs to take other attributes into account).
In instances like these, the package offers hooks into the slugging workflow where you can use your own functions, either on a per-model basis, or in your own trait that extends the package's trait.
NOTE: If you are putting these methods into your own trait, you will need to indicate in your models that PHP should use your trait methods instead of the packages (since a class can't use two traits with the same methods), e.g.
/** * Your trait where you collect your common Sluggable extension methods */ class MySluggableTrait { public function customizeSlugEngine(...) {} public function scopeWithUniqueSlugConstraints(...) {} // etc. } /** * Your model */ class MyModel { // Tell PHP to use your methods instead of the packages: use Sluggable, MySluggableTrait { MySluggableTrait::customizeSlugEngine insteadof Sluggable; MySluggableTrait::scopeWithUniqueSlugConstraints insteadof Sluggable; } // ... }
/**
* @param \Cocur\Slugify\Slugify $engine
* @param string $attribute
* @return \Cocur\Slugify\Slugify
*/
public function customizeSlugEngine(Slugify $engine, string $attribute): \Cocur\Slugify\Slugify
{
// ...
return $engine;
}
If you extend this method, the Slugify engine can be customized before slugging occurs. This might be where you change the character mappings that are used, or alter language files, etc..
You can customize the engine on a per-model and per-attribute basis (maybe your model has two slug fields, and one of them needs customization).
Take a look at tests/Models/PostWithCustomEngine.php
for an example.
Also, take a look at the slugEngineOptions configuration for other ways to customize Slugify.
/**
* @param \Illuminate\Database\Eloquent\Builder $query
* @param \Illuminate\Database\Eloquent\Model $model
* @param string $attribute
* @param array $config
* @param string $slug
* @return \Illuminate\Database\Eloquent\Builder
*/
public function scopeWithUniqueSlugConstraints(
Builder $query,
Model $model,
string $attribute,
array $config,
string $slug
): Builder
{
// ...
}
This method is applied to the query that is used to determine if a given slug is unique. The arguments passed to the scope are:
$model
-- the object being slugged$attribute
-- the slug field being generated,$config
-- the configuration array for the given model and attribute$slug
-- the "base" slug (before any unique suffixes are applied)
Feel free to use these values anyway you like in your query scope. As an example, look at
tests/Models/PostWithUniqueSlugConstraints.php
where the slug is generated for a post from it's title, but
the slug is scoped to the author. So Bob can have a post with the same title as Pam's post, but both
will have the same slug.
/**
* Query scope for finding "similar" slugs, used to determine uniqueness.
*
* @param \Illuminate\Database\Eloquent\Builder $query
* @param string $attribute
* @param array $config
* @param string $slug
* @return \Illuminate\Database\Eloquent\Builder
*/
public function scopeFindSimilarSlugs(Builder $query, string $attribute, array $config, string $slug): Builder
{
// ...
}
This is the default scope for finding "similar" slugs for a model. Basically, the package looks for existing
slugs that are the same as the $slug
argument, or that start with $slug
plus the separator string.
The resulting collection is what is passed to the uniqueSuffix
handler.
Generally, this query scope (which is defined in the Sluggable trait) should be left alone. However, you are free to overload it in your models.
Adding the optional SluggableScopeHelpers
trait to your model allows you to work with models
and their slugs. For example:
$post = Post::whereSlug($slugString)->get();
$post = Post::findBySlug($slugString);
$post = Post::findBySlugOrFail($slugString);
Because models can have more than one slug, this requires a bit more configuration. See SCOPE-HELPERS.md for all the details.
See ROUTE-MODEL-BINDING.md for details.
Thanks to everyone who has contributed to this project! Special thanks to JetBrains for their Open Source License Program ... and the excellent PHPStorm IDE, of course!
Please use Github for reporting bugs, and making comments or suggestions.
See CONTRIBUTING.md for how to contribute changes.
eloquent-sluggable was written by Colin Viebrock and is released under the MIT License.
Copyright (c) 2013 Colin Viebrock