README

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

Installation

Make sure you have the MongoDB PHP driver installed. You can find installation instructions at http://php.net/manual/en/mongodb.installation.php

Laravel version Compatibility

Install the package via Composer:

$ composer require jenssegers/mongodb

Laravel

In case your Laravel version does NOT autoload the packages, add the service provider to config/app.php:

Jenssegers\Mongodb\MongodbServiceProvider::class,

Lumen

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.

Non-Laravel projects

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);
});

Testing

To run the test for this package, run:

docker-compose up

Database Testing

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;

Configuration

You can use MongoDB either as the main database, either as a side database. To do so, add a new mongodb connection to config/database.php:

'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' => [
        // here you can pass more settings to the Mongo Driver Manager
        // see https://www.php.net/manual/en/mongodb-driver-manager.construct.php under "Uri Options" for a list of complete parameters that you can use

        'database' => env('DB_AUTHENTICATION_DATABASE', 'admin'), // required with Mongo 3+
    ],
],

For multiple servers or replica set configurations, set the host to an array and specify each server host:

'mongodb' => [
    'driver' => 'mongodb',
    'host' => ['server1', 'server2', ...],
    ...
    'options' => [
        'replicaSet' => 'rs0',
    ],
],

If you wish to use a connection string instead of full key-value params, you can set it so. Check the documentation on MongoDB's URI format: https://docs.mongodb.com/manual/reference/connection-string/

'mongodb' => [
    'driver' => 'mongodb',
    'dsn' => env('DB_DSN'),
    'database' => env('DB_DATABASE', 'homestead'),
],

Eloquent

Extending the base model

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';
}

// Mongo 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';
}

Extending the Authenticable base model

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
{

}

Soft Deletes

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;

    protected $dates = ['deleted_at'];
}

For more information check Laravel Docs about Soft Deleting.

Guarding attributes

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.

Dates

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 $dates = ['birthday'];
}

This allows you to execute queries like this:

$users = User::where(
    'birthday', '>',
    new DateTime('-18 years')
)->get();

Basic Usage

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();

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']);

MongoDB-specific operators

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();

MongoDB-specific Geo operations

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();

Inserts, updates and deletes

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.

MongoDB specific operations

Raw Expressions

These expressions will be injected directly into the query.

User::whereRaw([
    'age' => ['$gt' => 30, '$lt' => 40],
])->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');

Relationships

Basic Usage

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);
    }
}

belongsToMany and pivots

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'
        );
    }
}

EmbedsMany Relationship

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

EmbedsOne Relationship

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);

Query Builder

Basic Usage

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.

Available operations

To see the available operations, check the Eloquent section.

Schema

The database driver also has (limited) schema builder support. You can easily manipulate collections and set indexes.

Basic Usage

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

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');
});

Extending

Cross-Database Relationships

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);
    }
}

Authentication

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.

Queues

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',
],

Laravel specific

Add the service provider in config/app.php:

Jenssegers\Mongodb\MongodbQueueServiceProvider::class,

Lumen specific

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);

Upgrading

Upgrading from version 2 to 3

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();

Security contact information

To report a security vulnerability, follow these steps.

wangzd / mongodb

Maintainers

Details

README

Installation

Laravel version Compatibility

Laravel

Lumen

Non-Laravel projects

Testing

Database Testing

Configuration

Eloquent

Extending the base model

Extending the Authenticable base model

Soft Deletes

Guarding attributes

Dates

Basic Usage

MongoDB-specific operators

MongoDB-specific Geo operations

Inserts, updates and deletes

MongoDB specific operations

Relationships

Basic Usage

belongsToMany and pivots

EmbedsMany Relationship

EmbedsOne Relationship

Query Builder

Basic Usage

Available operations

Schema

Basic Usage

Geospatial indexes

Extending

Cross-Database Relationships

Authentication

Queues

Laravel specific

Lumen specific

Upgrading

Upgrading from version 2 to 3

Security contact information