在Python中解析嵌套的Json以刪除列中的特殊字符

Question

這是我的Json文件

{
    "highest_table": {
        "items": [{
                "key": "Human 1",
                "columns": {
                    "Na$me": "Tom",
                    "Description(ms/2)": "Table Number One on the Top",
                    "A&ge": "24",
                    "Ge_nder": "M"
                }
            },
            {
                "key": "Human 2",
                "columns": {
                    "Na$me": "John",
                    "Description(ms/2)": "Table Number One on the Top",
                    "A&ge": "23",
                    "Ge_nder": "M"
                }
                }
        ]
    }
}

目標是刪除列名中的所有特殊字符（或者，如果更容易的話，刪除.json文件中的所有特殊字符），並返回一個.json文件。 我最初的想法是將其轉換為熊貓，刪除列標題中的特殊字符，然后將其轉換回.json文件。

到目前為止，這是我嘗試過的。 它們都只打印一行。

import json
from pandas.io.json import json_normalize    

data_file = r"C:\characters.json"

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

df = json_normalize(data)  

-

data_file = r"C:\characters.json"

df = pd.read_json(data_file)  

如何提取列，刪除特殊字符並將其放回.json文件中？

Answer 1

一點問題與fixkey -您必須為fixkey提供完整的實現，但這應該可以解決您的問題。

import json

def fixkey(key):
    # toy implementation
    #print("fixing {}".format(key))
    return key.replace("&", "").replace("$", "")

def normalize(data):
    #print("normalizing {}".format(data))
    if isinstance(data, dict):
        data = {fixkey(key): normalize(value) for key, value in data.items()}
    elif isinstance(data, list):
        data = [normalize(item) for item in data]
    return data

jsdata = """
{
    "highest_table": {
        "items": [{
                "key": "Human 1",
                "columns": {
                    "Na$me": "Tom",
                    "Description(ms/2)": "Table Number One on the Top",
                    "A&ge": "24",
                    "Ge_nder": "M"
                }
            },
            {
                "key": "Human 2",
                "columns": {
                    "Na$me": "John",
                    "Description(ms/2)": "Table Number One on the Top",
                    "A&ge": "23",
                    "Ge_nder": "M"
                }
                }
        ]
    }
}
"""


data = json.loads(jsdata)

data = normalize(data)

result = json.dumps(data, indent=2)
print(result)

Answer 2

坦白地說，這很丑陋，但我還沒有找到更通用的方法。 這非常特定於您的特定JSON（問題確實需要在API中解決）。

import json


response = """{
    "highest_table": {
        "items": [{
                "key": "Human 1",
                "columns": {
                    "Na$me": "Tom",
                    "Description(ms/2)": "Table Number One on the Top",
                    "A&ge": "24",
                    "Ge_nder": "M"
                }
            },
            {
                "key": "Human 2",
                "columns": {
                    "Na$me": "John",
                    "Description(ms/2)": "Table Number One on the Top",
                    "A&ge": "23",
                    "Ge_nder": "M"
                }
                }
        ]
    }
}"""

def fix_json(resp):

    output = {'highest_table': {'items': []}}
    for item in resp['highest_table']['items']:
        inner_dict = item['columns']
        fixed_values = {'Name': inner_dict['Na$me'],
                        'Description(ms/2)': inner_dict['Description(ms/2)'],
                        'Age': inner_dict['A&ge'],
                        'Gender': inner_dict['Ge_nder']
                        }
        new_inner = {'key': item['key'], 'columns': fixed_values}
        output['highest_table']['items'].append(new_inner)
    return output



response = json.loads(response)
fixed = fix_json(response)