GraphQL在Django中的使用

简介

特点

• 请求你所要的数据，不多不少
• 获取多个资源，只用一个请求
• 描述所有的可能，类型系统
• 几乎所有语言支持

谁在用

文档

• Graphene-Python
• GraphQL | A query language for your API
  

背景

• 传统restful的接口定义类型多，试图简化接口定义
• django中使用restframework定义restful资源接口时，可能会出现深度查询，造成有时候查询过度
• 例如前端用户需要查询接口用于展示在下拉框时，用户仅需要id与value值时，造成无用字段冗余，影响接口返回性能
• 当一张表字段较多时，例如接口1一共有40个字段，A页面需要5个字段做展示，B页面需要另外10个字段展示，这时我们需要根据用户需求定义返回接口提升性能，且数据不会被暴露

实际问题

问题

• 请求数据量40kB可以根据用户缩减，也就是返回数据量可以做到<40KB
• 后端数据实际耗时783ms,但是数据传输一共耗时5s
  Django中如何使用呢
  安装

pip install graphene-django

django配置

INSTALLED_APPS = [
    "django.contrib.staticfiles", 
    "graphene_django"
    ]
GRAPHENE = {
    "SCHEMA": "test_api.schema.schema" # 下文中需要定义schema.py文件
}

Demo

定义数据库模型

from django.db import models


class Category(models.Model):
    name = models.CharField(max_length=100, help_text="名称")
    id = models.BigAutoField(primary_key=True)


class Ingredient(models.Model):
    id = models.BigAutoField(primary_key=True)
    name = models.CharField(max_length=100, help_text="名称")
    notes = models.TextField(help_text="笔记")
    category = models.ForeignKey(
        Category, related_name="category", on_delete=models.CASCADE
    )

    def __str__(self):
        return self.name

定义serializer

from graphene_django.rest_framework.mutation import SerializerMutation
from rest_framework.serializers import ModelSerializer

from ..models import Category, Ingredient


class CategorySerializer(ModelSerializer):
    class Meta:
        model = Category
        fields = "__all__"


class IngredientSerializer(ModelSerializer):
    class Meta:
        model = Ingredient
        fields = "__all__"

定义接口

import graphene
from graphene import relay
from graphene_django import DjangoObjectType
from graphene_django.filter import DjangoFilterConnectionField
from graphene_django.rest_framework.mutation import SerializerMutation

from ..models import Category, Ingredient

from ..serializer import CategorySerializer, IngredientSerializer


# 为查询添加查询总数
class CountableConnectionBase(relay.Connection):
    class Meta:
        abstract = True

    total_count = graphene.Int()

    def resolve_total_count(self, info, **kwargs):
        return self.iterable.count()


# Ingredient 查看过滤
class IngredientFilter(DjangoObjectType):
    class Meta:
        model = Ingredient
        fields = "__all__"
        filter_fields = {
            "name": ['exact', "contains", "istartswith"],
            "category": ["exact"],
            'category__name': ['exact'],
        }
        interfaces = (relay.Node,)
        connection_class = CountableConnectionBase

    extra_field = graphene.String()

    def resolve_extra_field(self: Ingredient, info):
        return "hello!" + str(self.id)


# CategoryFilter 查询过滤
class CategoryFilter(DjangoObjectType):
    class Meta:
        model = Category
        fields = "__all__"
        filter_fields = {
            "name": ['exact', "contains", "istartswith"],
        }
        interfaces = (relay.Node,)
        connection_class = CountableConnectionBase


# CategoryMutation 修改或新增
class CategoryMutation(SerializerMutation):
    class Meta:
        serializer_class = CategorySerializer


# IngredientMutation 修改或新增
class IngredientMutation(SerializerMutation):
    class Meta:
        serializer_class = IngredientSerializer


# 汇总query接口
class ApiQuery(graphene.ObjectType):
    search_category = DjangoFilterConnectionField(CategoryFilter)
    search_ingredient = DjangoFilterConnectionField(IngredientFilter)


# 汇总操作类接口
class ApiMutation(graphene.ObjectType):
    update_category = CategoryMutation.Field()
    update_ingredient = IngredientMutation.Field()

汇总所有接口

import graphene

from .api import ApiQuery, ApiMutation


class Query(ApiQuery):
    # 新增时提供多继承即可
    pass


class Mutation(ApiMutation):
    # 新增时提供多继承即可
    pass


schema = graphene.Schema(query=Query, mutation=Mutation)

启动

python manage.py runserver 0.0.0.0:8080

• 接口文档
  
  

总结

• 查询时，可以使用django_filter , 快速查询
• 用法基本和drf框架基本类似
• 接口面涉及的深度查询，通过connection实现，如果返回字段中没有改要求，将不会深度查询