How to Use the Serializer

Warning: You are browsing the documentation for Symfony 7.0, which is no longer maintained.

Read the updated version of this page for Symfony 7.3 (the current stable version).

Symfony provides a serializer to serialize/deserialize to and from objects and different formats (e.g. JSON or XML). Before using it, read the Serializer component docs to get familiar with its philosophy and the normalizers and encoders terminology.

Installation

In applications using Symfony Flex, run this command to install the serializer Symfony pack before using it:

        1
        $ composer require symfony/serializer-pack
    

Using the Serializer Service

Once enabled, the serializer service can be injected in any service where you need it or it can be used in a controller:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
        // src/Controller/DefaultController.php
namespace App\Controller;

use Symfony\Bundle\FrameworkBundle\Controller\AbstractController;
use Symfony\Component\HttpFoundation\Response;
use Symfony\Component\Serializer\SerializerInterface;

class DefaultController extends AbstractController
{
    public function index(SerializerInterface $serializer): Response
    {
        // keep reading for usage examples
    }
}
    

Or you can use the serialize Twig filter in a template:

        1
        {{ object|serialize(format = 'json') }}
    

See the twig reference for more information.

Adding Normalizers and Encoders

Once enabled, the serializer service will be available in the container. It comes with a set of useful encoders and normalizers.

Encoders supporting the following formats are enabled:

JSON: JsonEncoder
XML: XmlEncoder
CSV: CsvEncoder
YAML: YamlEncoder

As well as the following normalizers:

Other built-in normalizers and custom normalizers and/or encoders can also be loaded by tagging them as serializer.normalizer and serializer.encoder. It's also possible to set the priority of the tag in order to decide the matching order.

Danger

Always make sure to load the DateTimeNormalizer when serializing the DateTime or DateTimeImmutable classes to avoid excessive memory usage and exposing internal details.

Serializer Context

The serializer can define a context to control the (de)serialization of resources. This context is passed to all normalizers. For example:

DateTimeNormalizer uses datetime_format key as date time format;
AbstractObjectNormalizer uses preserve_empty_objects to represent empty objects as {} instead of [] in JSON.
Serializer uses empty_array_as_object to represent empty arrays as {} instead of [] in JSON.

You can pass the context as follows:

        1
2
3
4
5
6
7
        $serializer->serialize($something, 'json', [
    DateTimeNormalizer::FORMAT_KEY => 'Y-m-d H:i:s',
]);

$serializer->deserialize($someJson, Something::class, 'json', [
    DateTimeNormalizer::FORMAT_KEY => 'Y-m-d H:i:s',
]);
    

You can also configure the default context through the framework configuration:

        1
2
3
4
5
6
7
        # config/packages/framework.yaml
framework:
    # ...
    serializer:
        default_context:
            enable_max_depth: true
            yaml_indentation: 2
    

        1
2
3
4
5
6
7
        <!-- config/packages/framework.xml -->
<framework:config>
    <!-- ... -->
    <framework:serializer>
        <default-context enable-max-depth="true" yaml-indentation="2"/>
    </framework:serializer>
</framework:config>
    

        1
2
3
4
5
6
7
8
9
10
11
12
13
        // config/packages/framework.php
use Symfony\Component\Serializer\Encoder\YamlEncoder;
use Symfony\Component\Serializer\Normalizer\AbstractObjectNormalizer;
use Symfony\Config\FrameworkConfig;

return static function (FrameworkConfig $framework): void {
    $framework->serializer()
        ->defaultContext([
            AbstractObjectNormalizer::ENABLE_MAX_DEPTH => true,
            YamlEncoder::YAML_INDENTATION => 2,
        ])
    ;
};
    

You can also specify the context on a per-property basis:

        1
2
3
4
5
6
7
8
9
10
11
12
        namespace App\Model;

use Symfony\Component\Serializer\Annotation\Context;
use Symfony\Component\Serializer\Normalizer\DateTimeNormalizer;

class Person
{
    #[Context([DateTimeNormalizer::FORMAT_KEY => 'Y-m-d'])]
    public \DateTimeInterface $createdAt;

    // ...
}
    

        1
2
3
4
5
        App\Model\Person:
    attributes:
        createdAt:
            contexts:
                - { context: { datetime_format: 'Y-m-d' } }
    

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
        <?xml version="1.0" encoding="UTF-8" ?>
<serializer xmlns="http://symfony.com/schema/dic/serializer-mapping"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://symfony.com/schema/dic/serializer-mapping
        https://symfony.com/schema/dic/serializer-mapping/serializer-mapping-1.0.xsd"
>
    <class name="App\Model\Person">
        <attribute name="createdAt">
            <context>
                <entry name="datetime_format">Y-m-d</entry>
            </context>
        </attribute>
    </class>
</serializer>
    

Use the options to specify context specific to normalization or denormalization:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
        namespace App\Model;

use Symfony\Component\Serializer\Annotation\Context;
use Symfony\Component\Serializer\Normalizer\DateTimeNormalizer;

class Person
{
    #[Context(
        normalizationContext: [DateTimeNormalizer::FORMAT_KEY => 'Y-m-d'],
        denormalizationContext: [DateTimeNormalizer::FORMAT_KEY => '!Y-m-d'], // To prevent to have the time from the moment of denormalization
    )]
    public \DateTimeInterface $createdAt;

    // ...
}
    

You can also restrict the usage of a context to some groups:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
        namespace App\Model;

use Symfony\Component\Serializer\Annotation\Context;
use Symfony\Component\Serializer\Annotation\Groups;
use Symfony\Component\Serializer\Normalizer\DateTimeNormalizer;

class Person
{
    #[Groups(['extended'])]
    #[Context([DateTimeNormalizer::FORMAT_KEY => \DateTime::RFC3339])]
    #[Context(
        context: [DateTimeNormalizer::FORMAT_KEY => \DateTime::RFC3339_EXTENDED],
        groups: ['extended'],
    )]
    public \DateTimeInterface $createdAt;

    // ...
}
    

The attribute can be repeated as much as needed on a single property. Context without group is always applied first. Then context for the matching groups are merged in the provided order.

If you repeat the same context in multiple properties, consider using the #[Context] attribute on your class to apply that context configuration to all the properties of the class:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
        namespace App\Model;

use Symfony\Component\Serializer\Annotation\Context;
use Symfony\Component\Serializer\Normalizer\DateTimeNormalizer;

#[Context([DateTimeNormalizer::FORMAT_KEY => \DateTime::RFC3339])]
#[Context(
    context: [DateTimeNormalizer::FORMAT_KEY => \DateTime::RFC3339_EXTENDED],
    groups: ['extended'],
)]
class Person
{
    // ...
}
    

Using Context Builders

To define the (de)serialization context, you can use "context builders", which are objects that help you to create that context by providing autocompletion, validation, and documentation:

        1
2
3
4
        use Symfony\Component\Serializer\Context\Normalizer\DateTimeNormalizerContextBuilder;

$contextBuilder = (new DateTimeNormalizerContextBuilder())->withFormat('Y-m-d H:i:s');
$serializer->serialize($something, 'json', $contextBuilder->toArray());
    

Each normalizer/encoder has its related context builder. To create a more complex (de)serialization context, you can chain them using the withContext() method:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
        use Symfony\Component\Serializer\Context\Encoder\CsvEncoderContextBuilder;
use Symfony\Component\Serializer\Context\Normalizer\ObjectNormalizerContextBuilder;

$initialContext = [
    'custom_key' => 'custom_value',
];

$contextBuilder = (new ObjectNormalizerContextBuilder())
    ->withContext($initialContext)
    ->withGroups(['group1', 'group2']);

$contextBuilder = (new CsvEncoderContextBuilder())
    ->withContext($contextBuilder)
    ->withDelimiter(';');

$serializer->serialize($something, 'csv', $contextBuilder->toArray());
    

You can also create your context builders to have autocompletion, validation, and documentation for your custom context values.

Using Serialization Groups Attributes

You can add #[Groups] attributes to your class properties:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
        // src/Entity/Product.php
namespace App\Entity;

use Doctrine\ORM\Mapping as ORM;
use Symfony\Component\Serializer\Annotation\Groups;

#[ORM\Entity]
class Product
{
    #[ORM\Id]
    #[ORM\GeneratedValue]
    #[ORM\Column(type: 'integer')]
    #[Groups(['show_product', 'list_product'])]
    private int $id;

    #[ORM\Column(type: 'string', length: 255)]
    #[Groups(['show_product', 'list_product'])]
    private string $name;

    #[ORM\Column(type: 'text')]
    #[Groups(['show_product'])]
    private string $description;
}
    

You can also use the #[Groups] attribute on class level:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
        #[ORM\Entity]
#[Groups(['show_product'])]
class Product
{
    #[ORM\Id]
    #[ORM\GeneratedValue]
    #[ORM\Column(type: 'integer')]
    #[Groups(['list_product'])]
    private int $id;

    #[ORM\Column(type: 'string', length: 255)]
    #[Groups(['list_product'])]
    private string $name;

    #[ORM\Column(type: 'text')]
    private string $description;
}
    

In this example, the id and the name properties belong to the show_product and list_product groups. The description property only belongs to the show_product group.

Now that your groups are defined, you can choose which groups to use when serializing:

        1
2
3
4
5
6
7
        use Symfony\Component\Serializer\Context\Normalizer\ObjectNormalizerContextBuilder;

$context = (new ObjectNormalizerContextBuilder())
    ->withGroups('show_product')
    ->toArray();

$json = $serializer->serialize($product, 'json', $context);
    

Tip

The value of the groups key can be a single string, or an array of strings.

In addition to the #[Groups] attribute, the Serializer component also supports YAML or XML files. These files are automatically loaded when being stored in one of the following locations:

All *.yaml and *.xml files in the config/serializer/ directory.
The serialization.yaml or serialization.xml file in the Resources/config/ directory of a bundle;
All *.yaml and *.xml files in the Resources/config/serialization/ directory of a bundle.

Using Nested Attributes

To map nested properties, use the SerializedPath configuration to define their paths using a valid PropertyAccess syntax:

        1
2
3
4
5
6
7
8
9
10
11
        namespace App\Model;

use Symfony\Component\Serializer\Annotation\SerializedPath;

class Person
{
    #[SerializedPath('[profile][information][birthday]')]
    private string $birthday;

    // ...
}
    

        1
2
3
4
        App\Model\Person:
    attributes:
        dob:
            serialized_path: '[profile][information][birthday]'
    

        1
2
3
4
5
6
7
8
9
10
        <?xml version="1.0" encoding="UTF-8" ?>
<serializer xmlns="http://symfony.com/schema/dic/serializer-mapping"
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
    xsi:schemaLocation="http://symfony.com/schema/dic/serializer-mapping
        https://symfony.com/schema/dic/serializer-mapping/serializer-mapping-1.0.xsd"
>
    <class name="App\Model\Person">
        <attribute name="dob" serialized-path="[profile][information][birthday]"/>
    </class>
</serializer>
    

Using the configuration from above, denormalizing with a metadata-aware normalizer will write the birthday field from $data onto the Person object:

        1
2
3
4
5
6
7
8
9
        $data = [
    'profile' => [
        'information' => [
            'birthday' => '01-01-1970',
        ],
    ],
];
$person = $normalizer->denormalize($data, Person::class, 'any');
$person->getBirthday(); // 01-01-1970
    

When using attributes, the SerializedPath can either be set on the property or the associated _getter_ method. The SerializedPath cannot be used in combination with a SerializedName for the same property.

Configuring the Metadata Cache

The metadata for the serializer is automatically cached to enhance application performance. By default, the serializer uses the cache.system cache pool which is configured using the cache.system option.

Enabling a Name Converter

The use of a name converter service can be defined in the configuration using the name_converter option.

The built-in CamelCase to snake_case name converter can be enabled by using the serializer.name_converter.camel_case_to_snake_case value:

        1
2
3
4
5
        # config/packages/framework.yaml
framework:
    # ...
    serializer:
        name_converter: 'serializer.name_converter.camel_case_to_snake_case'
    

        1
2
3
4
5
        <!-- config/packages/framework.xml -->
<framework:config>
    <!-- ... -->
    <framework:serializer name-converter="serializer.name_converter.camel_case_to_snake_case"/>
</framework:config>
    

        1
2
3
4
5
6
        // config/packages/framework.php
use Symfony\Config\FrameworkConfig;

return static function (FrameworkConfig $framework): void {
    $framework->serializer()->nameConverter('serializer.name_converter.camel_case_to_snake_case');
};
    

Debugging the Serializer

Use the debug:serializer command to dump the serializer metadata of a given class:

        1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
        $ php bin/console debug:serializer 'App\Entity\Book'

    App\Entity\Book
    ---------------

    +----------+------------------------------------------------------------+
    | Property | Options                                                    |
    +----------+------------------------------------------------------------+
    | name     | [                                                          |
    |          |   "groups" => [                                            |
    |          |       "book:read",                                         |
    |          |       "book:write",                                        |
    |          |   ]                                                        |
    |          |   "maxDepth" => 1,                                         |
    |          |   "serializedName" => "book_name"                          |
    |          |   "ignore" => false                                        |
    |          |   "normalizationContexts" => [],                           |
    |          |   "denormalizationContexts" => []                          |
    |          | ]                                                          |
    | isbn     | [                                                          |
    |          |   "groups" => [                                            |
    |          |       "book:read",                                         |
    |          |   ]                                                        |
    |          |   "maxDepth" => null,                                      |
    |          |   "serializedName" => null                                 |
    |          |   "ignore" => false                                        |
    |          |   "normalizationContexts" => [],                           |
    |          |   "denormalizationContexts" => []                          |
    |          | ]                                                          |
    +----------+------------------------------------------------------------+
    

Going Further with the Serializer

API Platform provides an API system supporting the following formats:

JSON-LD along with the Hydra Core Vocabulary
OpenAPI v2 (formerly Swagger) and v3
GraphQL
JSON:API
HAL
JSON
XML
YAML
CSV

It is built on top of the Symfony Framework and its Serializer component. It provides custom normalizers and a custom encoder, custom metadata and a caching system.

If you want to leverage the full power of the Symfony Serializer component, take a look at how this bundle works.

This work, including the code samples, is licensed under a Creative Commons BY-SA 3.0 license.

TOC

Version