如何用Python列表推导式移除不含stats键的会话处理者

问题 #

我有一个存储在变量interactions中的JSON数组，包含呼叫信息、呼叫处理人员，呼叫统计数据存于"stats"键下。部分会话（session）无"stats"键，需移除这些会话对应的处理者（handler），生成新JSON数组。此前用for循环实现但速度慢，请问如何用Python列表推导式完成该操作？

示例输入 #

[
  {
    "callId": "17f2e9a7-2c82-4908-80f4-019ec2b8be57",
    "Start": "2023-04-05T16:38:56",
    "handlers": [
      {
        "handlerId": "dafedee8-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "c7372835-47f3-461c-b506-e604d0633994",
            "stats": [
              {
                "name": "talkcnt",
                "value": 1
              },
              {
                "name": "talktime",
                "value": 903481
              }
            ]
          }
        ]
      },
      {
        "handlerId": "dafeddfds-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "cdsfd-47f3-461c-b506-e604d0633994",
            "stats": [
              {
                "name": "talkcnt",
                "value": 100
              },
              {
                "name": "holdtime",
                "value": 232323
              }
            ]
          }
        ]
      }
    ]
  },
  {
    "callId": "23sdf-2c82-4908-80f4-019ec2b8be57",
    "Start": "2023-04-05T16:38:56",
    "handlers": [
      {
        "handlerId": "sdfsdfee8-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "c7372835-47f3-461c-b506-e6sdfd",
            "stats": [
              {
                "name": "talkcnt",
                "value": 1
              },
              {
                "name": "talktime",
                "value": 903481
              }
            ]
          }
        ]
      },
      {
        "handlerId": "dafeddfds-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "abc94"
          }
        ]
      }
    ]
  }
]

预期输出 #

[
  {
    "callId": "17f2e9a7-2c82-4908-80f4-019ec2b8be57",
    "Start": "2023-04-05T16:38:56",
    "handlers": [
      {
        "handlerId": "dafedee8-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "c7372835-47f3-461c-b506-e604d0633994",
            "stats": [
              {
                "name": "talkcnt",
                "value": 1
              },
              {
                "name": "talktime",
                "value": 903481
              }
            ]
          }
        ]
      },
      {
        "handlerId": "dafeddfds-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "cdsfd-47f3-461c-b506-e604d0633994",
            "stats": [
              {
                "name": "talkcnt",
                "value": 100
              },
              {
                "name": "holdtime",
                "value": 232323
              }
            ]
          }
        ]
      }
    ]
  },
  {
    "callId": "23sdf-2c82-4908-80f4-019ec2b8be57",
    "Start": "2023-04-05T16:38:56",
    "handlers": [
      {
        "handlerId": "sdfsdfee8-2869-414f-970e-657f92016e6b",
        "Name": "test",
        "sessions": [
          {
            "ani": "tel:+120134323",
            "direction": "inbound",
            "dnis": "tel:+3243423",
            "requestedRoutings": [
              "Standard"
            ],
            "sessionId": "c7372835-47f3-461c-b506-e6sdfd",
            "stats": [
              {
                "name": "talkcnt",
                "value": 1
              },
              {
                "name": "talktime",
                "value": 903481
              }
            ]
          }
        ]
      }
    ]
  }
]

解决方案 #

可以通过嵌套列表推导式实现，利用Python内部对推导式的优化提升执行速度，核心逻辑是筛选出所有会话都包含stats键的处理者：

filtered_interactions = [
    {
        **call,
        "handlers": [
            handler for handler in call["handlers"]
            if all("stats" in session for session in handler["sessions"])
        ]
    }
    for call in interactions
]

代码说明 #

• 外层推导式遍历每个呼叫对象，用**call保留原呼叫的所有字段，仅重写handlers字段
• 内层推导式筛选处理者：通过all()函数判断该处理者的所有会话是否都包含stats键，只有满足条件的处理者才会被保留
• 这种写法比普通for循环更简洁，且执行效率更高

内容的提问来源于stack exchange，提问作者chinna7326