解析 JSON 并将其转换为 CSV

Question

在Parsing values from a JSON file using Python? 的帮助下，我能够使用以下方法读取数据：

with open('data/data.json') as data_file:
    data_loaded = json.load(data_file)

但是，我无法进入要将其转换为 CSV 文件的下一步。示例代码：

import pandas as pd
import json 
from io import StringIO
import io
import csv
import sys

sample = {'Table': [{'AA12_Requestor_0': 'John Doe',
   'AA28_Requestor Business Unit_0': 'Manchester United',
   'AA29_Target Remediation Date_0': '2017-07-01',
   'AA29_Exception Lead_0': 'Ji Sung Park',
   'AA29_Level 1 Approver_0': 'Phil Knight',
   'AA29_Date of last WF change_0': '2017-02-12',
   'AA29_Approval Cycles_0': '1',
   'AA29_Level 2 Approver_0': 'Jae In Moon',
   'AA29_Exception Submitted Date_0': '2015-10-12',
   'AA29_Days in Awaiting Assignment_0': '0',
   'AA29_Days in Awaiting Info Sec Review_0': '0',
   'AA29_Days in Awaiting Governance Management Review_0': '2',
   'AA29_Days in Awaiting Level 1 Review - No Level 2 Rev_0': '',
   'AA29_Days from Submitted to Approved_0': '5',
   'AA29_Days from Submitted to Rejected_0': '',
   'AA29_Geography_0': None,
   'AA29_Blood Type': 'Love',
   'AA32_EX: Area_0': None,
   'AA32_Impact_0': None,
   'AA32_Rating_0': None,
   'AA32_Rating (Number)_0': '',
   'AA32_Likelihood_0': None,
   'AA32_Risk Sensitivity_0': None,
   'AA3_Description_0': 'Title: Get your job done | Norm: Be happy.\xa0 Love OS. | Blood type: Z',
   'AA39_Record ID_0': '12345',
   'AA39_Escalation Flag_0': 'No',
   'AA39_Submitted Date_0': '2015-10-15',
   'AA40_ID_0': '123',
   'AF13_Record Workflow State': 'Canceled',
   'FINDING_ID': 12345}]}

在我的 JSON 文件中有更多上述字典的迭代

我试过了：

1. 参考http://blog.appliedinformaticsinc.com/how-to-parse-and-convert-json-to-csv-using-python/

   json_parsed = json.loads(sample)
   

   给我一​​个错误，说 JSON 对象必须是字符串而不是字典。

   TypeError: the JSON object must be str, bytes or bytearray, not 'dict'
   

2. 试图创建csvwriter对象

   csvwriter = csv.writer(sample)
   

   给我一​​个错误提示：

   TypeError: argument 1 must have a "write" method
   

3. 我也读过How can I convert JSON to CSV?，但我无法通过第一步，即json.loads(sample)

4. 根据Python - JSON to CSV table? DictWriter 听起来很有趣。但不知道如何将其应用到我的任务中。

5. pd.read_json(sample) 也期待该字符串。给出一个错误提示

   TypeError: Expected String or Unicode
   

我希望所有键都是标题（列）和行中键的所有值。

Answer 1

sample 不是 JSON 对象。

pd.read_json 会失败。

pd.read_json 接受一个指向 JSON 文档的字符串。您正在传递一个 dict 对象。

尝试传递这个：

sample = {'AA12_Requestor_0': 'John Doe',
   'AA28_Requestor Business Unit_0': 'Manchester United',
   'AA29_Target Remediation Date_0': '2017-07-01',
   'AA29_Exception Lead_0': 'Ji Sung Park',
   'AA29_Level 1 Approver_0': 'Phil Knight',
   'AA29_Date of last WF change_0': '2017-02-12',
   'AA29_Approval Cycles_0': '1',
   'AA29_Level 2 Approver_0': 'Jae In Moon',
   'AA29_Exception Submitted Date_0': '2015-10-12',
   'AA29_Days in Awaiting Assignment_0': '0',
   'AA29_Days in Awaiting Info Sec Review_0': '0',
   'AA29_Days in Awaiting Governance Management Review_0': '2',
   'AA29_Days in Awaiting Level 1 Review - No Level 2 Rev_0': '',
   'AA29_Days from Submitted to Approved_0': '5',
   'AA29_Days from Submitted to Rejected_0': '',
   'AA29_Geography_0': None,
   'AA29_Blood Type': 'Love',
   'AA32_EX: Area_0': None,
   'AA32_Impact_0': None,
   'AA32_Rating_0': None,
   'AA32_Rating (Number)_0': '',
   'AA32_Likelihood_0': None,
   'AA32_Risk Sensitivity_0': None,
   'AA3_Description_0': 'Title: Get your job done | Norm: Be happy.\xa0 Love OS. | Blood type: Z',
   'AA39_Record ID_0': '12345',
   'AA39_Escalation Flag_0': 'No',
   'AA39_Submitted Date_0': '2015-10-15',
   'AA40_ID_0': '123',
   'AF13_Record Workflow State': 'Canceled',
   'FINDING_ID': 12345}

data = pd.DataFrame(sample, index=[0]) # Only one row so index has a length of 1

从那里您可以执行 ```data.to_csv("path/to/out/file.csv")`` 将 JSON 对象传递给 csv。