我正在用python创建一个应用程序，我使用postgresql数据库，用tortoise-orm进行查询。在数据库中有几个表，但是对于问题我有："project"，"assignment"，它们之间是一对多的关系，我需要根据某个参数过滤"assignments"（这个问题无关紧要），然后我想把它们分组，在json中返回，这就是我要实现的函数：

from datetime import date
from itertools import groupby

from app.models.assignment import Assignment

async def calendar(self, start_filter: date, final_filter:date):
        assignments = await self.model.filter(
            start_date__lte=final_filter,
            final_date__gte=start_filter
        ).prefetch_related("collaborator__job", "project").all()
        grouped_assignments = {}
        for project_id, project_assignments in groupby(assignments, lambda x: x.project_id):
            grouped_assignments[project_id] = list(project_assignments)
        return grouped_assignments

为了直接解决这个问题，在作业中应用了过滤器，我得到了这样的结果：

[
  {
    "name": "tarea de backend",
    "start_date": "2023-02-10",
    "final_date": "2023-02-20",
    "id": 1,
    "collaborator_id": 1,
    "project_id": 1
  },
  {
    "name": "tarea de backend",
    "start_date": "2023-02-10",
    "final_date": "2023-02-25",
    "id": 2,
    "collaborator_id": 2,
    "project_id": 1
  },
  {
    "name": "tarea de data science",
    "start_date": "2023-02-10",
    "final_date": "2023-02-20",
    "id": 3,
    "collaborator_id": 3,
    "project_id": 1
  },
  {
    "name": "tarea de data science",
    "start_date": "2023-02-10",
    "final_date": "2023-02-25",
    "id": 4,
    "collaborator_id": 4,
    "project_id": 1
  },
  {
    "name": "tarea de frontend",
    "start_date": "2023-02-20",
    "final_date": "2023-02-25",
    "id": 5,
    "collaborator_id": 5,
    "project_id": 1
  },
  {
    "name": "tarea de frontend",
    "start_date": "2023-02-20",
    "final_date": "2023-02-25",
    "id": 6,
    "collaborator_id": 5,
    "project_id": 2
  },
  {
    "name": "tarea de frontend",
    "start_date": "2023-02-20",
    "final_date": "2023-02-27",
    "id": 7,
    "collaborator_id": 6,
    "project_id": 1
  },
  {
    "name": "tarea de frontend",
    "start_date": "2023-02-20",
    "final_date": "2023-02-27",
    "id": 8,
    "collaborator_id": 6,
    "project_id": 2
  },
  {
    "name": "Tarea de backend",
    "start_date": "2023-03-11",
    "final_date": "2023-03-17",
    "id": 9,
    "collaborator_id": 1,
    "project_id": 2
  }
]

现在只有两个项目（Project_id = 1和Project_id = 2），我想要的是所有的任务都根据每个project_id分组，所以在json中我期望如下所示：

{
  "1": [
    {
      "start_date": "2023-02-20",
      "collaborator_id": 6,
      "final_date": "2023-02-27",
      "id": 7,
      "name": "tarea de frontend",
      "project_id": 1
    }
  ],
  "2": [
    {
      "start_date": "2023-02-20",
      "collaborator_id": 6,
      "final_date": "2023-02-27",
      "id": 8,
      "name": "tarea de frontend",
      "project_id": 2
    },
    {
      "start_date": "2023-03-11",
      "collaborator_id": 1,
      "final_date": "2023-03-17",
      "id": 9,
      "name": "Tarea de backend",
      "project_id": 2
    }

]}
实际上上面的json就是我得到的，由于某些原因我没有得到每个项目的所有任务（我测试的日期过滤器返回了上面我放置的9个赋值语句的列表）我看代码的方式是，在for循环中，我迭代赋值语句，并根据它们的project_id对它们进行分组，group_by应该只返回两个分组，一个用于project_id = 1，另一个用于project_id = 2，但是它似乎对相同的project_id进行了不同的分组。我不知道我做错了什么，或者我误解了groupby的工作原理。