CSV 文件的字典列表

Question

I have a list of dictionaries:

[   defaultdict(<class 'dict'>,
                {   'account_id': '',
                    'address': {   'address_country': 'ZM',
                                   'city': 'North Matthewland',
                                   'state': 'Nevada',
                                   'street_name': 'Cabrera Extensions',
                                   'street_number': 197,
                                   'zip_code': '81431'},
                    'affiliate_id': 12,
                    'brand': 'TTT',
                    'country': 'ZM',
                    'email': 'rosariojohn@TTT.zed',
                    'first_name': 'Peter',
                    'last_name': 'Green',
                    'leadsource': 559,
                    'password': 'test385',
                    'phone_number': '052839601'},)]

In my situation, I need to put all this data to CSV file, so via csv module, I try to write all this data to CSV file but I receive every time error message:

ValueError: dict contains fields not in fieldnames: 'address' 

So I add to 'fieldnames' address, but the problem is that I receive all data in one column of address.

with open('test_file.csv', 'w') as csvfile:
    filed_names = ['first_name',
                   'last_name',
                   'email',
                   'phone_number',
                   'password',
                   'country',
                   'leadsource',
                   'affiliate_id',
                   'account_id',
                   'brand',
                   'street_number',
                   'street_name',
                   'city',
                   'state',
                   'address_country',
                   'zip_code',
                   ]
    writer = csv.DictWriter(csvfile, fieldnames=filed_names)
    writer.writeheader()
    writer.writerows(list_user_details)

Answer 1

You'll need to flatten out the structure so it's just a list of dictionaries holding strings and numbers, rather than a list of dictionaries holding dictionaries of strings and numbers.

Assuming that you don't want to modify this list just for the writing step, make a new empty list. Copy the dictionaries from one list to the other, but convert the address dictionary into a set of additional fields called address.address_country , address.city , address.state and so on.

Then use csv to write out this new list, and pass it the modified fieldnames.

Also don't forget to test with commas inside the field strings.

Answer 2

Pandas offers a convenient alternative:

# first flatten address
for d in lst:
    d.update(d['address'])
    del d['address']

# read into dataframe
df = pd.DataFrame(d, index=range(len(lst)))

# output to csv
df.to_csv('file.csv', index=False)

# Result
#   account_id address_country  affiliate_id brand               city country  \
# 0                         ZM            12   TTT  North Matthewland      ZM   

#                  email first_name last_name  leadsource password phone_number  \
# 0  rosariojohn@TTT.zed      Peter     Green         559  test385    052839601   

#     state         street_name  street_number zip_code  
# 0  Nevada  Cabrera Extensions            197    81431