**Note**: This is the documentation for the **version 3.0** of REST framework. Documentation for [version 2.4](http://tomchristie.github.io/rest-framework-2-docs/) is also available.
Serializer fields handle converting between primitive values and internal datatypes. They also deal with validating input values, as well as retrieving and setting the values from their parent objects.
**Note:** The serializer fields are declared in `fields.py`, but by convention you should import them using `from rest_framework import serializers` and refer to fields as `serializers.<FieldName>`.
Each serializer field class constructor takes at least these arguments. Some Field classes take additional, field-specific arguments, but the following should always be accepted:
Set this to `True` to ensure that the field is used when serializing a representation, but is not used when creating or updating an instance during deserialization.
Defaults to `False`
### `write_only`
Set this to `True` to ensure that the field may be used when updating or creating an instance, but is not included when serializing the representation.
Normally an error will be raised if `None` is passed to a serializer field. Set this keyword argument to `True` if `None` should be considered a valid value.
If set, this gives the default value that will be used for the field if no input value is supplied. If not set the default behavior is to not populate the attribute at all.
Note that setting a `default` value implies that the field is not required. Including both the `default` and `required` keyword arguments is invalid and will raise an error.
The name of the attribute that will be used to populate the field. May be a method that only takes a `self` argument, such as `URLField('get_absolute_url')`, or may use dotted notation to traverse attributes, such as `EmailField(source='user.email')`.
The value `source='*'` has a special meaning, and is used to indicate that the entire object should be passed through to the field. This can be useful for creating nested representations, or for fields which require access to the complete object in order to determine the output representation.
A list of validator functions which should be applied to the incoming field input, and which either raise a validation error or simply return. Validator functions should typically raise `serializers.ValidationError`, but Django's built-in `ValidationError` is also supported for compatibility with validators defined in the Django codebase or third party Django packages.
A value that should be used for pre-populating the value of HTML form fields.
### `style`
A dictionary of key-value pairs that can be used to control how renderers should render the field. The API for this should still be considered experimental, and will be formalized with the 3.1 release.
Two options are currently used in HTML form generation, `'input_type'` and `'base_template'`.
# Use <inputtype="password"> for the input.
password = serializers.CharField(
style={'input_type': 'password'}
)
# Use a radio input instead of a select input.
color_channel = serializers.ChoiceField(
choices=['red', 'green', 'blue']
style = {'base_template': 'radio.html'}
}
**Note**: The `style` argument replaces the old-style version 2.x `widget` keyword argument. Because REST framework 3 now uses templated HTML form generation, the `widget` option that was used to support Django built-in widgets can no longer be supported. Version 3.1 is planned to include public API support for customizing HTML form generation.
-`max_length` - Validates that the input contains no more than this number of characters.
-`min_length` - Validates that the input contains no fewer than this number of characters.
-`allow_blank` - If set to `True` then the empty string should be considered a valid value. If set to `False` then the empty string is considered invalid and will raise a validation error. Defaults to `False`.
The `allow_null` option is also available for string fields, although its usage is discouraged in favor of `allow_blank`. It is valid to set both `allow_blank=True` and `allow_null=True`, but doing so means that there will be two differing types of empty value permissible for string representations, which can lead to data inconsistencies and subtle application bugs.
Corresponds to `django.db.models.fields.IntegerField`, `django.db.models.fields.SmallIntegerField`, `django.db.models.fields.PositiveIntegerField` and `django.db.models.fields.PositiveSmallIntegerField`.
-`decimal_places` The number of decimal places to store with the number.
-`coerce_to_string` Set to `True` if string values should be returned for the representation, or `False` if `Decimal` objects should be returned. Defaults to the same value as the `COERCE_DECIMAL_TO_STRING` settings key, which will be `True` unless overridden. If `Decimal` objects are returned by the serializer, then the final output format will be determined by the renderer.
-`max_value` Validate that the number provided is no greater than this value.
-`min_value` Validate that the number provided is no less than this value.
#### Example usage
To validate numbers up to 999 with a resolution of 2 decimal places, you would use:
This field also takes an optional argument, `coerce_to_string`. If set to `True` the representation will be output as a string. If set to `False` the representation will be left as a `Decimal` instance and the final representation will be determined by the renderer.
If unset, this will default to the same value as the `COERCE_DECIMAL_TO_STRING` setting, which is `True` unless set otherwise.
*`format` - A string representing the output format. If not specified, this defaults to the same value as the `DATETIME_FORMAT` settings key, which will be `'iso-8601'` unless set. Setting to a format string indicates that `to_representation` return values should be coerced to string output. Format strings are described below. Setting this value to `None` indicates that Python `datetime` objects should be returned by `to_representation`. In this case the datetime encoding will be determined by the renderer.
*`input_formats` - A list of strings representing the input formats which may be used to parse the date. If not specified, the `DATETIME_INPUT_FORMATS` setting will be used, which defaults to `['iso-8601']`.
Format strings may either be [Python strftime formats][strftime] which explicitly specify the format, or the special string `'iso-8601'`, which indicates that [ISO 8601][iso8601] style datetimes should be used. (eg `'2013-01-29T12:34:56.000000Z'`)
When a value of `None` is used for the format `datetime` objects will be returned by `to_representation` and the final output representation will determined by the renderer class.
In the case of JSON this means the default datetime representation uses the [ECMA 262 date time string specification][ecma262]. This is a subset of ISO 8601 which uses millisecond precision, and includes the 'Z' suffix for the UTC timezone, for example: `2013-01-29T12:34:56.123Z`.
When using `ModelSerializer` or `HyperlinkedModelSerializer`, note that any model fields with `auto_now=True` or `auto_now_add=True` will use serializer fields that are `read_only=True` by default.
If you want to override this behavior, you'll need to declare the `DateTimeField` explicitly on the serializer. For example:
class CommentSerializer(serializers.ModelSerializer):
*`format` - A string representing the output format. If not specified, this defaults to the same value as the `DATE_FORMAT` settings key, which will be `'iso-8601'` unless set. Setting to a format string indicates that `to_representation` return values should be coerced to string output. Format strings are described below. Setting this value to `None` indicates that Python `date` objects should be returned by `to_representation`. In this case the date encoding will be determined by the renderer.
*`input_formats` - A list of strings representing the input formats which may be used to parse the date. If not specified, the `DATE_INPUT_FORMATS` setting will be used, which defaults to `['iso-8601']`.
Format strings may either be [Python strftime formats][strftime] which explicitly specify the format, or the special string `'iso-8601'`, which indicates that [ISO 8601][iso8601] style dates should be used. (eg `'2013-01-29'`)
*`format` - A string representing the output format. If not specified, this defaults to the same value as the `TIME_FORMAT` settings key, which will be `'iso-8601'` unless set. Setting to a format string indicates that `to_representation` return values should be coerced to string output. Format strings are described below. Setting this value to `None` indicates that Python `time` objects should be returned by `to_representation`. In this case the time encoding will be determined by the renderer.
*`input_formats` - A list of strings representing the input formats which may be used to parse the date. If not specified, the `TIME_INPUT_FORMATS` setting will be used, which defaults to `['iso-8601']`.
Format strings may either be [Python strftime formats][strftime] which explicitly specify the format, or the special string `'iso-8601'`, which indicates that [ISO 8601][iso8601] style times should be used. (eg `'12:34:56.000000'`)
-`allow_blank` - If set to `True` then the empty string should be considered a valid value. If set to `False` then the empty string is considered invalid and will raise a validation error. Defaults to `False`.
Both the `allow_blank` and `allow_null` are valid options on `ChoiceField`, although it is highly recommended that you only use one and not both. `allow_blank` should be preferred for textual choices, and `allow_null` should be preferred for numeric or other non-textual choices.
A field that can accept a set of zero, one or many values, chosen from a limited set of choices. Takes a single mandatory argument. `to_internal_representation` returns a `set` containing the selected values.
-`allow_blank` - If set to `True` then the empty string should be considered a valid value. If set to `False` then the empty string is considered invalid and will raise a validation error. Defaults to `False`.
As with `ChoiceField`, both the `allow_blank` and `allow_null` options are valid, although it is highly recommended that you only use one and not both. `allow_blank` should be preferred for textual choices, and `allow_null` should be preferred for numeric or other non-textual choices.
The `FileField` and `ImageField` classes are only suitable for use with `MultiPartParser` or `FileUploadParser`. Most parsers, such as e.g. JSON don't support file uploads.
Django's regular [FILE_UPLOAD_HANDLERS] are used for handling uploaded files.
-`use_url` - If set to `True` then URL string values will be used for the output representation. If set to `False` then filename string values will be used for the output representation. Defaults to the value of the `UPLOADED_FILES_USE_URL` settings key, which is `True` unless set otherwise.
-`max_length` - Designates the maximum length for the file name.
-`allow_empty_file` - Designates if empty files are allowed.
-`use_url` - If set to `True` then URL string values will be used for the output representation. If set to `False` then filename string values will be used for the output representation. Defaults to the value of the `UPLOADED_FILES_USE_URL` settings key, which is `True` unless set otherwise.
A field class that simply returns the value of the field without modification.
This field is used by default with `ModelSerializer` when including field names that relate to an attribute rather than a model field.
**Signature**: `ReadOnlyField()`
For example, is `has_expired` was a property on the `Account` model, then the following serializer would automatically generate it as a `ReadOnlyField`:
class AccountSerializer(serializers.ModelSerializer):
The `HiddenField` class is usually only needed if you have some validation that needs to run based on some pre-provided field values, but you do not want to expose all of those fields to the end user.
For further examples on `HiddenField` see the [validators](validators.md) documentation.
A generic field that can be tied to any arbitrary model field. The `ModelField` class delegates the task of serialization/deserialization to its associated model field. This field can be used to create serializer fields for custom model fields, without having to create a new custom serializer field.
The `ModelField` class is generally intended for internal use, but can be used by your API if needed. In order to properly instantiate a `ModelField`, it must be passed a field that is attached to an instantiated model. For example: `ModelField(model_field=MyModel()._meta.get_field('custom_field'))`
This is a read-only field. It gets its value by calling a method on the serializer class it is attached to. It can be used to add any sort of data to the serialized representation of your object.
The serializer method referred to by the `method_name` argument should accept a single argument (in addition to `self`), which is the object being serialized. It should return whatever you want to be included in the serialized representation of the object. For example:
If you want to create a custom field, you'll need to subclass `Field` and then override either one or both of the `.to_representation()` and `.to_internal_value()` methods. These two methods are used to convert between the initial datatype, and a primitive, serializable datatype. Primitive datatypes will typically be any of a number, string, boolean, `date`/`time`/`datetime` or `None`. They may also be any list or dictionary like object that only contains other primitive objects. Other types might be supported, depending on the renderer that you are using.
The `to_internal_value()` method is called to restore a primitive datatype into its internal python representation. This method should raise a `serializer.ValidationError` if the data is invalid.
Note that the `WritableField` class that was present in version 2.x no longer exists. You should subclass `Field` and override `to_internal_value()` if the field supports data input.
By default field values are treated as mapping to an attribute on the object. If you need to customize how the field value is accessed and set you need to override `.get_attribute()` and/or `.get_value()`.
The [drf-compound-fields][drf-compound-fields] package provides "compound" serializer fields, such as lists of simple values, which can be described by other fields rather than serializers with the `many=True` option. Also provided are fields for typed dictionaries and values that can be either a specific type or a list of items of that type.
The [drf-extra-fields][drf-extra-fields] package provides extra serializer fields for REST framework, including `Base64ImageField` and `PointField` classes.
The [django-rest-framework-gis][django-rest-framework-gis] package provides geographic addons for django rest framework like a `GeometryField` field and a GeoJSON serializer.
The [django-rest-framework-hstore][django-rest-framework-hstore] package provides an `HStoreField` to support [django-hstore][django-hstore] `DictionaryField` model field.
