JSON 未正确保存

Question

所以我从 API 中提取数据，并且只想从 JSON 响应中保存特定的字典和列表。问题是，当我在循环中转储数据时，它会在文件中创建看起来非常奇怪的数据，而这些数据实际上并不是 JSON。

r=requests.get(url,headers=header)
result=r.json()
with open ('myfile.json','a+') as file:
    for log in result['logs']:
        hello=json.dump(log['log']['driver']['username'], file)
        hello=json.dump(log['log']['driver']['first_name'],file)
        hello=json.dump(log['log']['driver']['last_name'],file)
        for event in log['log']['events']:
            hello=json.dump(event['event']['id'],file)
            hello=json.dump(event['event']['start_time'],file)
            hello=json.dump(event['event']['type'],file)
            hello=json.dump(event['event']['location'],file)

这里的最终目标是将此数据转换为 CSV。我将它保存到 JSON 文件的唯一原因是我可以加载它并将其保存到 CSV 中。我的目标 API 端点是 Logs：

https://developer.keeptruckin.com/reference#get-logs

Answer 1

我认为@GBrandt 在创建有效的 JSON 输出方面有正确的想法，但正如我在评论中所说，我认为 JSON 到 JSON 的转换步骤并不是真正必要的——因为你可以创建您已经拥有的 JSON 中的 CSV 文件：

（修改为根据您的后续问题将start_time 拆分为两个单独的字段。）

result = r.json()

with open('myfile.csv', 'w', newline='') as csvfile:
    writer = csv.writer(csvfile, quoting=csv.QUOTE_ALL)
    for log in result['logs']:
        username = log['log']['driver']['username']
        first_name = log['log']['driver']['first_name']
        last_name = log['log']['driver']['last_name']

        for event in log['log']['events']:
            id = event['event']['id']
            start_time = event['event']['start_time']
            date, time = start_time.split('T')  # Split time into two fields.
            _type = event['event']['type']  # Avoid using name of built-in.
            location = event['event']['location']
            if not location:
                location = "N/A"
            writer.writerow(
                (username, first_name, last_name, id, date, time, _type, location))

Answer 2

看起来您只是以非结构化方式将单个 JSON 字符串转储到文件中。

json.dump 不会神奇地创建类似 JSON 字典的对象并将其保存到文件中。见：

json.dump(log['log']['driver']['username'], file)

它实际上所做的只是将驱动程序的用户名字符串化并将其直接转储到文件中，因此文件将只有一个字符串，而不是 JSON 对象（我猜这就是你想要的）。它是 JSON，只是不是很有用。

你要找的是这个：

r=requests.get(url,headers=header)
result=r.json()
with open ('myfile.json','w+') as file:
    logs = []
    for log in result['logs']:
        logs.append({
            'username': log['log']['driver']['username'],
            'first_name': log['log']['driver']['first_name'],
            'last_name': log['log']['driver']['last_name'],
            # ...
            'events': [
                ({
                    'id': event['event']['id'],
                    'start_time': event['event']['start_time'],
                    # ...
                }) for event in log['log']['events']
            ]
        })
    json.dump(logs, file)

另外，我建议不要对 JSON 文件使用附加模式，.json 预计会保存一个 JSON 对象（就我而言）。

Answer 3

下面的代码怎么样（示例 json 从文件加载，而不是通过 HTTP 调用以获取数据）。

从 https://developer.keeptruckin.com/reference#get-logs 获取的 JSON 示例

import json

with open('input.json', 'r') as f_in:
    data = json.load(f_in)

data_to_collect = []
logs = data['logs']
with open('output.json', 'w') as f_out:
    for log in logs:
        _log = log['log']
        data_to_collect.append({key: _log['driver'].get(key) for key in ['username', 'first_name', 'last_name']})
        data_to_collect[-1]['events'] = []
        for event in _log['events']:
            data_to_collect[-1]['events'].append(
                {key: event['event'].get(key) for key in ['id', 'start_time', 'type', 'location']})
    json.dump(data_to_collect, f_out)

输出文件

[
  {
    "username": "demo_driver",
    "first_name": "Demo",
    "last_name": "Driver",
    "events": [
      {
        "start_time": "2016-10-16T07:00:00Z",
        "type": "driving",
        "id": 221,
        "location": "Mobile, AL"
      },
      {
        "start_time": "2016-10-16T09:00:00Z",
        "type": "sleeper",
        "id": 474,
        "location": null
      },
      {
        "start_time": "2016-10-16T11:00:00Z",
        "type": "driving",
        "id": 475,
        "location": null
      }
    ]
  }
]