这工作得相对较好。 它将json压缩成csv文件。 嵌套元素被管理:)

这是python 3的

import json

o = json.loads('your json string') # Be careful, o must be a list, each of its objects will make a line of the csv.

def flatten(o, k='/'):
    global l, c_line
    if isinstance(o, dict):
        for key, value in o.items():
            flatten(value, k + '/' + key)
    elif isinstance(o, list):
        for ov in o:
            flatten(ov, '')
    elif isinstance(o, str):
        o = o.replace('\r',' ').replace('\n',' ').replace(';', ',')
        if not k in l:
            l[k]={}
        l[k][c_line]=o

def render_csv(l):
    ftime = True

    for i in range(100): #len(l[list(l.keys())[0]])
        for k in l:
            if ftime :
                print('%s;' % k, end='')
                continue
            v = l[k]
            try:
                print('%s;' % v[i], end='')
            except:
                print(';', end='')
        print()
        ftime = False
        i = 0

def json_to_csv(object_list):
    global l, c_line
    l = {}
    c_line = 0
    for ov in object_list : # Assumes json is a list of objects
        flatten(ov)
        c_line += 1
    render_csv(l)

json_to_csv(o)

享受。

我已经尝试了很多建议的解决方案(也熊猫没有正确地规范化我的JSON)，但真正好的是正确解析JSON数据来自Max Berman。

我写了一个改进，以避免每一行都有新列 在解析期间将其放置到现有列。 如果只有一个数据存在，则将值存储为字符串，如果该列有更多值，则将值存储为列表。

它有一个输入。Json文件作为输入，并输出一个output.csv。

import json
import pandas as pd

def flatten_json(json):
    def process_value(keys, value, flattened):
        if isinstance(value, dict):
            for key in value.keys():
                process_value(keys + [key], value[key], flattened)
        elif isinstance(value, list):
            for idx, v in enumerate(value):
                process_value(keys, v, flattened)
                # process_value(keys + [str(idx)], v, flattened)
        else:
            key1 = '__'.join(keys)
            if not flattened.get(key1) is None:
                if isinstance(flattened[key1], list):
                    flattened[key1] = flattened[key1] + [value]
                else:
                    flattened[key1] = [flattened[key1]] + [value]
            else:
                flattened[key1] = value

    flattened = {}
    for key in json.keys():
        k = key
        # print("Key: " + k)
        process_value([key], json[key], flattened)
    return flattened

try:
    f = open("input.json", "r")
except:
    pass
y = json.loads(f.read())
flat = flatten_json(y)
text = json.dumps(flat)
df = pd.read_json(text)
df.to_csv('output.csv', index=False, encoding='utf-8')

我对丹提出的解决方案感到困惑，但这对我来说很管用:

import json
import csv 

f = open('test.json')
data = json.load(f)
f.close()

f=csv.writer(open('test.csv','wb+'))

for item in data:
  f.writerow([item['pk'], item['model']] + item['fields'].values())

“测试的地方。Json”包含以下内容:

[ 
{"pk": 22, "model": "auth.permission", "fields": 
  {"codename": "add_logentry", "name": "Can add log entry", "content_type": 8 } }, 
{"pk": 23, "model": "auth.permission", "fields": 
  {"codename": "change_logentry", "name": "Can change log entry", "content_type": 8 } }, {"pk": 24, "model": "auth.permission", "fields": 
  {"codename": "delete_logentry", "name": "Can delete log entry", "content_type": 8 } }
]

令人惊讶的是，我发现到目前为止贴在这里的答案都没有正确处理所有可能的场景(例如，嵌套字典，嵌套列表，无值等)。

这个解决方案应该适用于所有场景:

def flatten_json(json):
    def process_value(keys, value, flattened):
        if isinstance(value, dict):
            for key in value.keys():
                process_value(keys + [key], value[key], flattened)
        elif isinstance(value, list):
            for idx, v in enumerate(value):
                process_value(keys + [str(idx)], v, flattened)
        else:
            flattened['__'.join(keys)] = value

    flattened = {}
    for key in json.keys():
        process_value([key], json[key], flattened)
    return flattened

修改了Alec McGail的答案，以支持包含列表的JSON

    def flattenjson(self, mp, delim="|"):
            ret = []
            if isinstance(mp, dict):
                    for k in mp.keys():
                            csvs = self.flattenjson(mp[k], delim)
                            for csv in csvs:
                                    ret.append(k + delim + csv)
            elif isinstance(mp, list):
                    for k in mp:
                            csvs = self.flattenjson(k, delim)
                            for csv in csvs:
                                    ret.append(csv)
            else:
                    ret.append(mp)

            return ret

谢谢!

使用csv.DictWriter()很容易，详细的实现可以像这样:

def read_json(filename):
    return json.loads(open(filename).read())
def write_csv(data,filename):
    with open(filename, 'w+') as outf:
        writer = csv.DictWriter(outf, data[0].keys())
        writer.writeheader()
        for row in data:
            writer.writerow(row)
# implement
write_csv(read_json('test.json'), 'output.csv')

注意，这假设所有JSON对象都具有相同的字段。

这是一份可能对你有帮助的参考资料。