Django -将多个行序列化为一个列表

Question

我有三个表顶点，边，图

1. 顶点可以是许多图的一部分。
2. 边缘表表示图中任意两个顶点之间的边。

模型

class Graph(models.Model):
    id = models.TextField(
        primary_key=True, null=False)
    name = models.TextField()

class Vertex(models.Model):
    id = models.TextField(
        primary_key=True, null=False)
    name = models.TextField()
    extras = models.JsonField()
 
class Edge(models.Model):
    graph = models.ForeignKey(
        "Graph", related_name="edges")
    node = models.ForeignKey(
        "Vertex", related_name="g_vertex")
    dependency = models.ForeignKey(
        "Vertex", related_name="e_vertex")

串行化

class VertexNameSerializer(serializers.ModelSerializer):
    class Meta:
        model = Vertex
        fields = ['name']

class VertexSerializer(serializers.ModelSerializer):
    class Meta:
        model = Vertex
        fields = ['name', 'extras']


class EdgeSerializer(serializers.ModelSerializer):
    node = VertexSerializer()
    dependency = VertexNameSerializer()

    class Meta:
        model = Edge
        fields = ['node', 'dependency']


class GraphSerializer(serializers.ModelSerializer):
    edges = EdgeSerializer(many=True)

    class Meta:
        model = Graph
        fields = ['id', 'name', 'edges']
        depth = 3

GraphViewSet

class GraphViewSet(viewsets.ModelViewSet):
    queryset = Graph.objects.all()
    serializer_class = GraphSerializer

关于上面的图

​

表：

+-----+--------------+
| id  |     Name     |
+-----+--------------+
| g-1 | Sample Graph |
+-----+--------------+

Vertex
+-----+------+-------+
| id  | Name | Extra |
+-----+------+-------+
| v-1 | A    | {}    |
| v-2 | B    | {}    |
| v-3 | C    | {}    |
+-----+------+-------+

Edge
+-------+------+------------+
| Graph | Node | Dependency |
+-------+------+------------+
| g-1   | v-1  | null       |
| g-1   | v-2  | null       |
| g-1   | v-3  | v-1        |
| g-1   | v-3  | v-2        |
+-------+------+------------+

因为我有视图集，如果我对/graph端点执行GET调用，这就是我得到的输出

{
  "id": "g-1",
  "name": "Sample Graph",
  "edges": [
    {
      "node": {
        "name": "A",
        "extras": {}
      },
      "dependency": null
    },
    {
      "node": {
        "name": "B",
        "extras": {}
      },
      "dependency": null
    },
    {
      "node": {
        "name": "C",
        "extras": {}
      },
      "dependency": {
        "name": "A"
      }
    },
    {
      "node": {
        "name": "C",
        "extras": {}
      },
      "dependency": {
        "name": "B"
      }
    }
  ]
}

但我想要输出

{
  "id": "g-1",
  "name": "Sample Graph",
  "edges": [
    {
      "node": {
        "name": "A",
        "extras": {}
      },
      "dependency": null
    },
    {
      "node": {
        "name": "B",
        "extras": {}
      },
      "dependency": null
    },
    {
      "node": {
        "name": "C",
        "extras": {}
      },
      "dependency": ["A", "B"]
    }
  ]
}

如何更改序列化程序以获得所需的输出？

Answer 1

像这样的东西应该能起作用

class GraphSerializer(serializers.ModelSerializer):
    edges = serializers.SerializerMethodField()

    class Meta:
        model = Graph
        fields = ['id', 'name', 'edges']

    def get_edges(self, obj):
        dependencies_raw = obj.edges.values_list('node__name', 'dependency__name')
        dependencies_mapper = {}
        for node_name, dependency_name in dependencies_raw:
            dependencies_mapper.setdefault(node_name, [])
            dependencies_mapper[node_name].append(dependency_name)

        nodes = EdgeSerializer(obj.edges.distinct('name').all(), many=True).data

        data = [
            {
                **node,
                'dependencies': dependencies_mapper.get(node['node']['name'], None)
            } for node in nodes
        ]
        return data

我现在不能测试。基本上，我会发出两个请求，而不是默认的1。

首先获得node name及其依赖项，第二项获取不同节点的依赖项。那我们就把它粘在一起。

并从dependency字段中删除EdgeSerializer字段。