This package adds functionalities to the Eloquent model and Query builder for MongoDB, using the original Laravel API. This library extends the original Laravel classes, so it uses exactly the same methods.
- Laravel MongoDB
Make sure you have the MongoDB PHP driver installed. You can find installation instructions at http://php.net/manual/en/mongodb.installation.php
Laravel | Package | Maintained |
---|---|---|
9.x | 3.9.x | ✅ |
8.x | 3.8.x | ✅ |
7.x | 3.7.x | ❌ |
6.x | 3.6.x | ❌ |
5.8.x | 3.5.x | ❌ |
5.7.x | 3.4.x | ❌ |
5.6.x | 3.4.x | ❌ |
5.5.x | 3.3.x | ❌ |
5.4.x | 3.2.x | ❌ |
5.3.x | 3.1.x or 3.2.x | ❌ |
5.2.x | 2.3.x or 3.0.x | ❌ |
5.1.x | 2.2.x or 3.0.x | ❌ |
5.0.x | 2.1.x | ❌ |
4.2.x | 2.0.x | ❌ |
Install the package via Composer:
$ composer require jenssegers/mongodb
In case your Laravel version does NOT autoload the packages, add the service provider to config/app.php
:
Jenssegers\Mongodb\MongodbServiceProvider::class,
For usage with Lumen, add the service provider in bootstrap/app.php
. In this file, you will also need to enable Eloquent. You must however ensure that your call to $app->withEloquent();
is below where you have registered the MongodbServiceProvider
:
$app->register(Jenssegers\Mongodb\MongodbServiceProvider::class);
$app->withEloquent();
The service provider will register a MongoDB database extension with the original database manager. There is no need to register additional facades or objects.
When using MongoDB connections, Laravel will automatically provide you with the corresponding MongoDB objects.
For usage outside Laravel, check out the Capsule manager and add:
$capsule->getDatabaseManager()->extend('mongodb', function($config, $name) {
$config['name'] = $name;
return new Jenssegers\Mongodb\Connection($config);
});
To run the test for this package, run:
docker-compose up
To reset the database after each test, add:
use Illuminate\Foundation\Testing\DatabaseMigrations;
Also inside each test classes, add:
use DatabaseMigrations;
Keep in mind that these traits are not yet supported:
use Database Transactions;
use RefreshDatabase;
To configure a new MongoDB connection, add a new connection entry to config/database.php
:
'mongodb' => [
'driver' => 'mongodb',
'dsn' => env('DB_DSN'),
'database' => env('DB_DATABASE', 'homestead'),
],
The dsn
key contains the connection string used to connect to your MongoDB deployment. The format and available options are documented in the MongoDB documentation.
Instead of using a connection string, you can also use the host
and port
configuration options to have the connection string created for you.
'mongodb' => [
'driver' => 'mongodb',
'host' => env('DB_HOST', '127.0.0.1'),
'port' => env('DB_PORT', 27017),
'database' => env('DB_DATABASE', 'homestead'),
'username' => env('DB_USERNAME', 'homestead'),
'password' => env('DB_PASSWORD', 'secret'),
'options' => [
'appname' => 'homestead',
],
],
The options
key in the connection configuration corresponds to the uriOptions
parameter.
This package includes a MongoDB enabled Eloquent class that you can use to define models for corresponding collections.
use Jenssegers\Mongodb\Eloquent\Model;
class Book extends Model
{
//
}
Just like a normal model, the MongoDB model class will know which collection to use based on the model name. For Book
, the collection books
will be used.
To change the collection, pass the $collection
property:
use Jenssegers\Mongodb\Eloquent\Model;
class Book extends Model
{
protected $collection = 'my_books_collection';
}
NOTE: MongoDB documents are automatically stored with a unique ID that is stored in the _id
property. If you wish to use your own ID, substitute the $primaryKey
property and set it to your own primary key attribute name.
use Jenssegers\Mongodb\Eloquent\Model;
class Book extends Model
{
protected $primaryKey = 'id';
}
// MongoDB will also create _id, but the 'id' property will be used for primary key actions like find().
Book::create(['id' => 1, 'title' => 'The Fault in Our Stars']);
Likewise, you may define a connection
property to override the name of the database connection that should be used when utilizing the model.
use Jenssegers\Mongodb\Eloquent\Model;
class Book extends Model
{
protected $connection = 'mongodb';
}
This package includes a MongoDB Authenticatable Eloquent class Jenssegers\Mongodb\Auth\User
that you can use to replace the default Authenticatable class Illuminate\Foundation\Auth\User
for your User
model.
use Jenssegers\Mongodb\Auth\User as Authenticatable;
class User extends Authenticatable
{
}
When soft deleting a model, it is not actually removed from your database. Instead, a deleted_at timestamp is set on the record.
To enable soft deletes for a model, apply the Jenssegers\Mongodb\Eloquent\SoftDeletes
Trait to the model:
use Jenssegers\Mongodb\Eloquent\SoftDeletes;
class User extends Model
{
use SoftDeletes;
}
For more information check Laravel Docs about Soft Deleting.
When choosing between guarding attributes or marking some as fillable, Taylor Otwell prefers the fillable route. This is in light of recent security issues described here.
Keep in mind guarding still works, but you may experience unexpected behavior.
Eloquent allows you to work with Carbon or DateTime objects instead of MongoDate objects. Internally, these dates will be converted to MongoDate objects when saved to the database.
use Jenssegers\Mongodb\Eloquent\Model;
class User extends Model
{
protected $casts = ['birthday' => 'datetime'];
}
This allows you to execute queries like this:
$users = User::where(
'birthday', '>',
new DateTime('-18 years')
)->get();
Retrieving all models
$users = User::all();
Retrieving a record by primary key
$user = User::find('517c43667db388101e00000f');
Where
$posts =
Post::where('author.name', 'John')
->take(10)
->get();
OR Statements
$posts =
Post::where('votes', '>', 0)
->orWhere('is_approved', true)
->get();
AND statements
$users =
User::where('age', '>', 18)
->where('name', '!=', 'John')
->get();
whereIn
$users = User::whereIn('age', [16, 18, 20])->get();
When using whereNotIn
objects will be returned if the field is non-existent. Combine with whereNotNull('age')
to leave out those documents.
whereBetween
$posts = Post::whereBetween('votes', [1, 100])->get();
whereNull
$users = User::whereNull('age')->get();
whereDate
$users = User::whereDate('birthday', '2021-5-12')->get();
The usage is the same as whereMonth
/ whereDay
/ whereYear
/ whereTime
Advanced wheres
$users =
User::where('name', 'John')
->orWhere(function ($query) {
return $query
->where('votes', '>', 100)
->where('title', '<>', 'Admin');
})->get();
orderBy
$users = User::orderBy('age', 'desc')->get();
Offset & Limit (skip & take)
$users =
User::skip(10)
->take(5)
->get();
groupBy
Selected columns that are not grouped will be aggregated with the $last
function.
$users =
Users::groupBy('title')
->get(['title', 'name']);
Distinct
Distinct requires a field for which to return the distinct values.
$users = User::distinct()->get(['name']);
// Equivalent to:
$users = User::distinct('name')->get();
Distinct can be combined with where:
$users =
User::where('active', true)
->distinct('name')
->get();
Like
$spamComments = Comment::where('body', 'like', '%spam%')->get();
Aggregation
Aggregations are only available for MongoDB versions greater than 2.2.x
$total = Product::count();
$price = Product::max('price');
$price = Product::min('price');
$price = Product::avg('price');
$total = Product::sum('price');
Aggregations can be combined with where:
$sold = Orders::where('sold', true)->sum('price');
Aggregations can be also used on sub-documents:
$total = Order::max('suborder.price');
NOTE: This aggregation only works with single sub-documents (like EmbedsOne
) not subdocument arrays (like EmbedsMany
).
Incrementing/Decrementing the value of a column
Perform increments or decrements (default 1) on specified attributes:
Cat::where('name', 'Kitty')->increment('age');
Car::where('name', 'Toyota')->decrement('weight', 50);
The number of updated objects is returned:
$count = User::increment('age');
You may also specify additional columns to update:
Cat::where('age', 3)
->increment('age', 1, ['group' => 'Kitty Club']);
Car::where('weight', 300)
->decrement('weight', 100, ['latest_change' => 'carbon fiber']);
Exists
Matches documents that have the specified field.
User::where('age', 'exists', true)->get();
All
Matches arrays that contain all elements specified in the query.
User::where('roles', 'all', ['moderator', 'author'])->get();
Size
Selects documents if the array field is a specified size.
Post::where('tags', 'size', 3)->get();
Regex
Selects documents where values match a specified regular expression.
use MongoDB\BSON\Regex;
User::where('name', 'regex', new Regex('.*doe', 'i'))->get();
NOTE: you can also use the Laravel regexp operations. These are a bit more flexible and will automatically convert your regular expression string to a MongoDB\BSON\Regex
object.
User::where('name', 'regexp', '/.*doe/i')->get();
The inverse of regexp:
User::where('name', 'not regexp', '/.*doe/i')->get();
Type
Selects documents if a field is of the specified type. For more information check: http://docs.mongodb.org/manual/reference/operator/query/type/#op._S_type
User::where('age', 'type', 2)->get();
Mod
Performs a modulo operation on the value of a field and selects documents with a specified result.
User::where('age', 'mod', [10, 0])->get();
Near
$bars = Bar::where('location', 'near', [
'$geometry' => [
'type' => 'Point',
'coordinates' => [
-0.1367563, // longitude
51.5100913, // latitude
],
],
'$maxDistance' => 50,
])->get();
GeoWithin
$bars = Bar::where('location', 'geoWithin', [
'$geometry' => [
'type' => 'Polygon',
'coordinates' => [
[
[-0.1450383, 51.5069158],
[-0.1367563, 51.5100913],
[-0.1270247, 51.5013233],
[-0.1450383, 51.5069158],
],
],
],
])->get();
GeoIntersects
$bars = Bar::where('location', 'geoIntersects', [
'$geometry' => [
'type' => 'LineString',
'coordinates' => [
[-0.144044, 51.515215],
[-0.129545, 51.507864],
],
],
])->get();
GeoNear
You are able to make a geoNear
query on mongoDB.
You don't need to specify the automatic fields on the model.
The returned instance is a collection. So you're able to make the Collection operations.
Just make sure that your model has a location
field, and a 2ndSphereIndex.
The data in the location
field must be saved as GeoJSON.
The location
points must be saved as WGS84 reference system for geometry calculation. That means, basically, you need to save longitude and latitude
, in that order specifically, and to find near with calculated distance, you need to do the same way
.
Bar::find("63a0cd574d08564f330ceae2")->update(
[
'location' => [
'type' => 'Point',
'coordinates' => [
-0.1367563,
51.5100913
]
]
]
);
$bars = Bar::raw(function ($collection) {
return $collection->aggregate([
[
'$geoNear' => [
"near" => [ "type" => "Point", "coordinates" => [-0.132239, 51.511874] ],
"distanceField" => "dist.calculated",
"minDistance" => 0,
"maxDistance" => 6000,
"includeLocs" => "dist.location",
"spherical" => true,
]
]
]);
});
Inserting, updating and deleting records works just like the original Eloquent. Please check Laravel Docs' Eloquent section.
Here, only the MongoDB-specific operations are specified.
Raw Expressions
These expressions will be injected directly into the query.
User::whereRaw([
'age' => ['$gt' => 30, '$lt' => 40],
])->get();
User::whereRaw([
'$where' => '/.*123.*/.test(this.field)',
])->get();
User::whereRaw([
'$where' => '/.*123.*/.test(this["hyphenated-field"])',
])->get();
You can also perform raw expressions on the internal MongoCollection object. If this is executed on the model class, it will return a collection of models.
If this is executed on the query builder, it will return the original response.
Cursor timeout
To prevent MongoCursorTimeout
exceptions, you can manually set a timeout value that will be applied to the cursor:
DB::collection('users')->timeout(-1)->get();
Upsert
Update or insert a document. Additional options for the update method are passed directly to the native update method.
// Query Builder
DB::collection('users')
->where('name', 'John')
->update($data, ['upsert' => true]);
// Eloquent
$user->update($data, ['upsert' => true]);
Projections
You can apply projections to your queries using the project
method.
DB::collection('items')
->project(['tags' => ['$slice' => 1]])
->get();
DB::collection('items')
->project(['tags' => ['$slice' => [3, 7]]])
->get();
Projections with Pagination
$limit = 25;
$projections = ['id', 'name'];
DB::collection('items')
->paginate($limit, $projections);
Push
Add items to an array.
DB::collection('users')
->where('name', 'John')
->push('items', 'boots');
$user->push('items', 'boots');
DB::collection('users')
->where('name', 'John')
->push('messages', [
'from' => 'Jane Doe',
'message' => 'Hi John',
]);
$user->push('messages', [
'from' => 'Jane Doe',
'message' => 'Hi John',
]);
If you DON'T want duplicate items, set the third parameter to true
:
DB::collection('users')
->where('name', 'John')
->push('items', 'boots', true);
$user->push('items', 'boots', true);
Pull
Remove an item from an array.
DB::collection('users')
->where('name', 'John')
->pull('items', 'boots');
$user->pull('items', 'boots');
DB::collection('users')
->where('name', 'John')
->pull('messages', [
'from' => 'Jane Doe',
'message' => 'Hi John',
]);
$user->pull('messages', [
'from' => 'Jane Doe',
'message' => 'Hi John',
]);
Unset
Remove one or more fields from a document.
DB::collection('users')
->where('name', 'John')
->unset('note');
$user->unset('note');
The only available relationships are:
- hasOne
- hasMany
- belongsTo
- belongsToMany
The MongoDB-specific relationships are:
- embedsOne
- embedsMany
Here is a small example:
use Jenssegers\Mongodb\Eloquent\Model;
class User extends Model
{
public function items()
{
return $this->hasMany(Item::class);
}
}
The inverse relation of hasMany
is belongsTo
:
use Jenssegers\Mongodb\Eloquent\Model;
class Item extends Model
{
public function user()
{
return $this->belongsTo(User::class);
}
}
The belongsToMany relation will not use a pivot "table" but will push id's to a related_ids attribute instead. This makes the second parameter for the belongsToMany method useless.
If you want to define custom keys for your relation, set it to null
:
use Jenssegers\Mongodb\Eloquent\Model;
class User extends Model
{
public function groups()
{
return $this->belongsToMany(
Group::class, null, 'user_ids', 'group_ids'
);
}
}
If you want to embed models, rather than referencing them, you can use the embedsMany
relation. This relation is similar to the hasMany
relation but embeds the models inside the parent object.
REMEMBER: These relations return Eloquent collections, they don't return query builder objects!
use Jenssegers\Mongodb\Eloquent\Model;
class User extends Model
{
public function books()
{
return $this->embedsMany(Book::class);
}
}
You can access the embedded models through the dynamic property:
$user = User::first();
foreach ($user->books as $book) {
//
}
The inverse relation is automagically available. You don't need to define this reverse relation.
$book = Book::first();
$user = $book->user;
Inserting and updating embedded models works similar to the hasMany
relation:
$book = $user->books()->save(
new Book(['title' => 'A Game of Thrones'])
);
// or
$book =
$user->books()
->create(['title' => 'A Game of Thrones']);
You can update embedded models using their save
method (available since release 2.0.0):
$book = $user->books()->first();
$book->title = 'A Game of Thrones';
$book->save();
You can remove an embedded model by using the destroy
method on the relation, or the delete
method on the model (available since release 2.0.0):
$book->delete();
// Similar operation
$user->books()->destroy($book);
If you want to add or remove an embedded model, without touching the database, you can use the associate
and dissociate
methods.
To eventually write the changes to the database, save the parent object:
$user->books()->associate($book);
$user->save();
Like other relations, embedsMany assumes the local key of the relationship based on the model name. You can override the default local key by passing a second argument to the embedsMany method:
use Jenssegers\Mongodb\Eloquent\Model;
class User extends Model
{
public function books()
{
return $this->embedsMany(Book::class, 'local_key');
}
}
Embedded relations will return a Collection of embedded items instead of a query builder. Check out the available operations here: https://laravel.com/docs/master/collections
The embedsOne relation is similar to the embedsMany relation, but only embeds a single model.
use Jenssegers\Mongodb\Eloquent\Model;
class Book extends Model
{
public function author()
{
return $this->embedsOne(Author::class);
}
}
You can access the embedded models through the dynamic property:
$book = Book::first();
$author = $book->author;
Inserting and updating embedded models works similar to the hasOne
relation:
$author = $book->author()->save(
new Author(['name' => 'John Doe'])
);
// Similar
$author =
$book->author()
->create(['name' => 'John Doe']);
You can update the embedded model using the save
method (available since release 2.0.0):
$author = $book->author;
$author->name = 'Jane Doe';
$author->save();
You can replace the embedded model with a new model like this:
$newAuthor = new Author(['name' => 'Jane Doe']);
$book->author()->save($newAuthor);
The database driver plugs right into the original query builder.
When using MongoDB connections, you will be able to build fluent queries to perform database operations.
For your convenience, there is a collection
alias for table
as well as some additional MongoDB specific operators/operations.
$books = DB::collection('books')->get();
$hungerGames =
DB::collection('books')
->where('name', 'Hunger Games')
->first();
If you are familiar with Eloquent Queries, there is the same functionality.
To see the available operations, check the Eloquent section.
Transactions require MongoDB version ^4.0 as well as deployment of replica set or sharded clusters. You can find more information in the MongoDB docs
DB::transaction(function () {
User::create(['name' => 'john', 'age' => 19, 'title' => 'admin', 'email' => 'john@example.com']);
DB::collection('users')->where('name', 'john')->update(['age' => 20]);
DB::collection('users')->where('name', 'john')->delete();
});
// begin a transaction
DB::beginTransaction();
User::create(['name' => 'john', 'age' => 19, 'title' => 'admin', 'email' => 'john@example.com']);
DB::collection('users')->where('name', 'john')->update(['age' => 20]);
DB::collection('users')->where('name', 'john')->delete();
// commit changes
DB::commit();
To abort a transaction, call the rollBack
method at any point during the transaction:
DB::beginTransaction();
User::create(['name' => 'john', 'age' => 19, 'title' => 'admin', 'email' => 'john@example.com']);
// Abort the transaction, discarding any data created as part of it
DB::rollBack();
NOTE: Transactions in MongoDB cannot be nested. DB::beginTransaction() function will start new transactions in a new created or existing session and will raise the RuntimeException when transactions already exist. See more in MongoDB official docs Transactions and Sessions
DB::beginTransaction();
User::create(['name' => 'john', 'age' => 20, 'title' => 'admin']);
// This call to start a nested transaction will raise a RuntimeException
DB::beginTransaction();
DB::collection('users')->where('name', 'john')->update(['age' => 20]);
DB::commit();
DB::rollBack();
The database driver also has (limited) schema builder support. You can easily manipulate collections and set indexes.
Schema::create('users', function ($collection) {
$collection->index('name');
$collection->unique('email');
});
You can also pass all the parameters specified in the MongoDB docs to the $options
parameter:
Schema::create('users', function ($collection) {
$collection->index(
'username',
null,
null,
[
'sparse' => true,
'unique' => true,
'background' => true,
]
);
});
Inherited operations:
- create and drop
- collection
- hasCollection
- index and dropIndex (compound indexes supported as well)
- unique
MongoDB specific operations:
- background
- sparse
- expire
- geospatial
All other (unsupported) operations are implemented as dummy pass-through methods because MongoDB does not use a predefined schema.
Read more about the schema builder on Laravel Docs
Geospatial indexes are handy for querying location-based documents.
They come in two forms: 2d
and 2dsphere
. Use the schema builder to add these to a collection.
Schema::create('bars', function ($collection) {
$collection->geospatial('location', '2d');
});
To add a 2dsphere
index:
Schema::create('bars', function ($collection) {
$collection->geospatial('location', '2dsphere');
});
If you're using a hybrid MongoDB and SQL setup, you can define relationships across them.
The model will automatically return a MongoDB-related or SQL-related relation based on the type of the related model.
If you want this functionality to work both ways, your SQL-models will need to use the Jenssegers\Mongodb\Eloquent\HybridRelations
trait.
This functionality only works for hasOne
, hasMany
and belongsTo
.
The MySQL model should use the HybridRelations
trait:
use Jenssegers\Mongodb\Eloquent\HybridRelations;
class User extends Model
{
use HybridRelations;
protected $connection = 'mysql';
public function messages()
{
return $this->hasMany(Message::class);
}
}
Within your MongoDB model, you should define the relationship:
use Jenssegers\Mongodb\Eloquent\Model;
class Message extends Model
{
protected $connection = 'mongodb';
public function user()
{
return $this->belongsTo(User::class);
}
}
If you want to use Laravel's native Auth functionality, register this included service provider:
Jenssegers\Mongodb\Auth\PasswordResetServiceProvider::class,
This service provider will slightly modify the internal DatabaseReminderRepository to add support for MongoDB based password reminders.
If you don't use password reminders, you don't have to register this service provider and everything else should work just fine.
If you want to use MongoDB as your database backend, change the driver in config/queue.php
:
'connections' => [
'database' => [
'driver' => 'mongodb',
// You can also specify your jobs specific database created on config/database.php
'connection' => 'mongodb-job',
'table' => 'jobs',
'queue' => 'default',
'expire' => 60,
],
],
If you want to use MongoDB to handle failed jobs, change the database in config/queue.php
:
'failed' => [
'driver' => 'mongodb',
// You can also specify your jobs specific database created on config/database.php
'database' => 'mongodb-job',
'table' => 'failed_jobs',
],
Add the service provider in config/app.php
:
Jenssegers\Mongodb\MongodbQueueServiceProvider::class,
With Lumen, add the service provider in bootstrap/app.php
. You must however ensure that you add the following after the MongodbServiceProvider
registration.
$app->make('queue');
$app->register(Jenssegers\Mongodb\MongodbQueueServiceProvider::class);
In this new major release which supports the new MongoDB PHP extension, we also moved the location of the Model class and replaced the MySQL model class with a trait.
Please change all Jenssegers\Mongodb\Model
references to Jenssegers\Mongodb\Eloquent\Model
either at the top of your model files or your registered alias.
use Jenssegers\Mongodb\Eloquent\Model;
class User extends Model
{
//
}
If you are using hybrid relations, your MySQL classes should now extend the original Eloquent model class Illuminate\Database\Eloquent\Model
instead of the removed Jenssegers\Eloquent\Model
.
Instead use the new Jenssegers\Mongodb\Eloquent\HybridRelations
trait. This should make things more clear as there is only one single model class in this package.
use Jenssegers\Mongodb\Eloquent\HybridRelations;
class User extends Model
{
use HybridRelations;
protected $connection = 'mysql';
}
Embedded relations now return an Illuminate\Database\Eloquent\Collection
rather than a custom Collection class. If you were using one of the special methods that were available, convert them to Collection operations.
$books = $user->books()->sortBy('title')->get();
To report a security vulnerability, follow these steps.