django開發-mongodb的配置與使用

今天整理了一下在django項目中如何使用mongodb, 環境如下：
ubuntu18.04, django2.0.5, drf3.9, mongoengine0.16

第一步：在settings.py中配置mongodb和mysql,配置如下(可以同時使用mysql和mongodb):

DATABASES = {
    'default': {
        'ENGINE': 'django.db.backends.mysql',   # 數據庫引擎
        'NAME': 'django_test2',                  # 你要存儲數據的庫名，事先要創建之
        'USER': 'root',                         # 數據庫用戶名
        'PASSWORD': 'wyzane',                     # 密碼
        'HOST': 'localhost',                    # 主機
        'PORT': '3306',                         # 數據庫使用的端口
    },
    'mongotest': {
        'ENGINE': None,
    }
}

import mongoengine
# 連接mongodb中數據庫名稱爲mongotest5的數據庫
conn = mongoengine.connect("mongotest")

第二步：向mongodb中插入數據
1.插入json類型數據

models.py:
    import mongoengine
    class StudentModel(mongoengine.Document):
        name = mongoengine.StringField(max_length=32)
        age = mongoengine.IntField()
        password = mongoengine.StringField(max_length=32)

views.py:
    from rest_framework.views import APIView
    class FirstMongoView(APIView):
        def post(self, request):
            name = request.data["name"]
            age = request.data["age"]
            password = request.data["password"]
            StudentModel.objects.create(name=name, age=age, password=password)
            return Response(dict(msg="OK", code=10000))

插入數據格式爲:

{
    "name": "nihao",
    "age": 18,
    "password": "123456"
}

2.插入含有list的json數據

models.py:
    import mongoengine
    class Student2Model(mongoengine.Document):
        name = mongoengine.StringField(max_length=32)
        # 用於存儲list類型的數據
        score = mongoengine.ListField()

views.py:
    from rest_framework.views import APIView
    class FirstMongo2View(APIView):
        def post(self, request):
            name = request.data["name"]
            score = request.data["score"]
            Student2Model.objects.create(name=name, score=score)
            return Response(dict(msg="OK", code=10000))

插入數據格式爲：

{
     "name": "test",
     "score": [12, 13]
}

3.插入含有dict和list的複雜json數據

models.py:
    import mongoengine
    class Student3Model(mongoengine.Document):
        name = mongoengine.StringField(max_length=32)
        # DictField用於存儲字典類型的數據
        score = mongoengine.DictField()
views.py:
    from rest_framework.views import APIView
    class FirstMongo3View(APIView):
        def post(self, request):
            name = request.data["name"]
            score = request.data["score"]
            Student3Model.objects.create(name=name, score=score)
            return Response(dict(msg="OK", code=10000))

插入數據格式爲:

{
    "name": "test",
    "score": {"xiaoming": 12, "xiaoli": 13}
}
或者:
{
    "name": "test",
    "score": {"xiaoming": 12, "xiaoli": {"xiaozhao": 14}}
}
或者:
{
"name": "test",
"score": {"xiaoming": 12, "xiaoli": {"xiaozhao": {"xiaoliu": 12, "xiaojian": 18}}}
}
或者:
{
"name": "test",
"score": {"xiaoming": 12, "xiaoli": {"xiaozhao": {"xiaoliu": 12, "xiaojian": [12,13,14]}}}
}

第三步：查詢mongodb中的數據
1.查詢並序列化複雜json數據

serializers.py:
    class StudentSerializer(serializers.Serializer):
        name = serializers.CharField()
        score = serializers.DictField()  # 序列化複雜的json數據
        # DictField與EmbeddedDocumentField類似，但是比EmbeddedDocumentField更靈活
views.py:
    class FirstMongo4View(APIView):
        def get(self, request):
            student_info = Student3Model.objects.all()
            # 增加過濾條件
            # student_info = Student3Model.objects.filter(name="test1")
            ser = StudentSerializer(instance=student_info, many=True)
            return Response(dict(msg="OK", code="10000", data=ser.data))

2.序列化mongodb中含有嵌套關係的兩個document

models.py:
    class AuthorModel(mongoengine.EmbeddedDocument):
        author_name = mongoengine.StringField(max_length=32)
        age = mongoengine.IntField()


    class BookModel(mongoengine.Document):
        book_name = mongoengine.StringField(max_length=64)
        publish = mongoengine.DateTimeField(default=datetime.datetime.utcnow())
        words = mongoengine.IntField()
        author = mongoengine.EmbeddedDocumentField(AuthorModel)

serializers.py: 序列化時注意與rest_framework的序列化中DictField()的區別
    from rest_framework_mongoengine import serializers as s1
    class AuthorSerializer(s1.DocumentSerializer):  
        # DocumentSerializer繼承自drf中的ModelSerializer，用於代替ModelSerializer序列化mongodb中的document.
        # 具體可以到官網上查看
        class Meta:
            model = AuthorModel
            fields = ('author_name', 'age')


    class BookSerializer(s1.DocumentSerializer):
        author = AuthorSerializer()

        class Meta:
            model = BookModel
            fields = ('book_name', 'publish', 'words', 'author')

    AuthorSerializer還可以這樣寫:
    class AuthorSerializer(s1.EmbeddedDocumentSerializer):
        # EmbeddedDocumentSerializer繼承了DocumentSerializer
        class Meta:
            model = AuthorModel
            fields = ('author_name', 'age')

views.py:
    class BookView(APIView):
        def get(self, request):
            """
            查詢數據
            :param request:
            :return:
            """
            books = BookModel.objects.all()
            ser = BookSerializer(instance=books, many=True)
            return Response(dict(msg="OK", code="10000", data=ser.data))

序列化mongodb中相關聯的兩個表時，如果序列化器繼承自rest_framework中的Serializer和ModelSerializer，會拋出如下異常：

Django serialization to JSON error: 'MetaDict' object has no attribute 'concrete_model'

此時，序列化器需要繼承自rest_framework_mongoengine的類，具體可以查看官網：
http://umutbozkurt.github.io/...

今天就聊到這裏，如有問題，歡迎交流。