[英]django-rest-framework: How Do I Serialize a Field That Already Contains JSON?
I am pretty new to the django-rest-framework, so could use some help. 我是django-rest-framework的新手,所以可以使用一些帮助。
I have an object with a TextField that is a string containing JSON. 我有一个TextField的对象,它是一个包含JSON的字符串。
I'm using django-rest-framework to serialize the whole object as JSON. 我正在使用django-rest-framework将整个对象序列化为JSON。 However, this one string that is already JSON gets serialized as an encoded string containing the JSON rather than the JSON itself. 但是,这个已经是JSON的字符串被序列化为包含JSON而不是JSON本身的编码字符串。
How can I tell the serializer to send this field as-is rather than trying to transform this string to JSON? 如何告诉序列化器按原样发送此字段而不是尝试将此字符串转换为JSON? Is there some sort of "ignore" decorator or override I can use? 我可以使用某种“忽略”装饰器或覆盖吗? Or can I pre-parse this JSON before serializing? 或者我可以在序列化之前预先解析这个JSON吗?
This is the difference between having: 这是有以下区别:
{"data": data}
and 和
{"data": "data"}
The latter being more of a nuisance to use on the client side... 后者在客户端使用起来更麻烦......
I solved this another way: 我解决了另一种方式:
1: use a JSON-Field for the JSON content ( django-jsonfield
or django-json-field
should be fine). 1:对JSON内容使用JSON-Field( django-jsonfield
或django-json-field
应该没问题)。 These then will to loads/dumps as needed 然后根据需要加载/转储
2: in my serializer, use the transform-method to prevent the data added as string to the response 2:在我的序列化程序中,使用transform-method来防止数据作为字符串添加到响应中
class MyModelSerializer(serializers.ModelSerializer):
def transform_myjsonfield(self, obj, value):
return obj.myjsonfield
class Meta:
model = MyModel
If you need write-access, you only have to add a method validate_myjsonfield
which converts back. 如果您需要写访问权限,则只需添加一个返回的方法validate_myjsonfield
。
(of course, this could be also done with a custom DRF serializer field. (当然,这也可以通过自定义DRF序列化器字段完成。
You can simply decode the json into python object: 您可以简单地将json解码为python对象:
json_obj = json.loads(model.json_text)
Once you serialize your object, replace this field with the decoded object: 序列化对象后,将此字段替换为已解码的对象:
data = serializer.data
data["field"] = json_obj
return Response(data)
Here is what works for me on, djangorestframework==3.9.1
: 这对我djangorestframework==3.9.1
, djangorestframework==3.9.1
:
import json
from rest_framework import serializers
from .models import WeatherLocation
class WeatherLocationSerializer(serializers.ModelSerializer):
LocationId = serializers.CharField(source='location_id')
City = serializers.CharField(source='city')
Region = serializers.CharField(source='region')
Name = serializers.CharField(source='name')
Country = serializers.CharField(source='country')
WeatherForecastLongTermTimePeriods = serializers.JSONField(required=False, allow_null=True,
source='weather_forecast_long_term_time_periods')
WeatherForecastShortTermTimePeriods = serializers.JSONField(required=False, allow_null=True,
source='weather_forecast_short_term_time_periods')
def to_representation(self, instance):
ret = super(WeatherLocationSerializer, self).to_representation(instance)
ret['WeatherForecastLongTermTimePeriods'] = json.loads(ret['WeatherForecastLongTermTimePeriods'])
ret['WeatherForecastShortTermTimePeriods'] = json.loads(ret['WeatherForecastShortTermTimePeriods'])
return ret
class Meta:
model = WeatherLocation
fields = ['LocationId', 'City', 'Region', 'Name', 'Country',
'WeatherForecastLongTermTimePeriods', 'WeatherForecastShortTermTimePeriods', ]
I thought there would be an easier way to do this, but by changing the behaviour of to_representation
, I can convert my text fields to JSON. 我认为有一种更简单的方法可以做到这一点,但通过改变to_representation
的行为,我可以将我的文本字段转换为JSON。 For reference, here is my models.py
: 作为参考,这是我的models.py
:
from django.db import models
class WeatherLocation(models.Model):
"""
Weather app schema, from southparc
"""
location_id = models.CharField(primary_key=True, null=False, blank=False, default=None, max_length=254,
editable=True)
region = models.CharField(max_length=2, null=False, blank=False)
city = models.CharField(null=False, blank=False, max_length=254)
province = models.CharField(null=True, blank=True, max_length=254)
name = models.CharField(null=True, blank=True, max_length=254)
country = models.CharField(null=True, blank=True, max_length=254)
# JSON fields
weather_forecast_long_term_time_periods = models.TextField(default="", blank=True)
weather_forecast_short_term_time_periods = models.TextField(default="", blank=True)
# Dates
created_date = models.DateTimeField(auto_now_add=True)
modified_date = models.DateTimeField(auto_now=True)
Hope that helps. 希望有所帮助。 If you use a JSONField
that is supported by Postgres, I'm sure you won't need to do this. 如果你使用Postgres支持的JSONField
,我相信你不需要这样做。 I'm using a TextField to save my JSON here. 我正在使用TextField来保存我的JSON。 I thought that specifying the field type as serializers.JSONField
on the serializer would be enough but its not the case. 我认为在字符串serializers.JSONField
上将字段类型指定为serializers.JSONField
就足够了,但事实并非如此。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.