python json module tips

the python json module is typically used for data sequence, such as

def get_user_info(user_id):
    res = {"user_id": 190013234,"nick": "havana"}
    json_str = json.dumps(res)
    return json_str

But when you want to serialize the data which contains Chinese characters, will this be

import json
res = {"user_id": 190027134, "nick": "愚蠢的萝卜"}
print(json.dumps(res))

# {"nick": "\u611a\u8822\u7684\u841d\u535c", "user_id": 190027134}

Solution: Add a parameter in json.dumps () insideensure_ascii=False

import json
res = {"user_id": 190027134, "nick": "愚蠢的萝卜"}
print(json.dumps(res))

# {"nick": "愚蠢的萝卜", "user_id": 190027134}

There is a problem, if json structure is very complex, the resulting output will be a mess, is not clear, it is formatted so that it can add a parameter

import json
res = {"user_id": 190027134, "nick": "愚蠢的萝卜"}
print(json.dumps(res))

# {
    "nick": "愚蠢的萝卜", 
    "user_id": 190027134
}

If the sequence of the data inside the python datetime.datetime type, then the error will be serialized

import datetime

res = {"user_id": 190027134, "nick": "愚蠢的萝卜", "reg_ts": datetime.datetime.now()}
print(json.dumps(res, ensure_ascii=False, indent=4))

#TypeError: datetime.datetime(2019, 11, 3, 19, 8, 45, 987000) is not JSON serializable

The solution is to use a custom jsonencoder instead of native jsonencoder

class DatetimeEncoder(json.JSONEncoder):

    def default(self, obj):
        if isinstance(obj, datetime.datetime):
            return obj.strftime('%Y-%m-%d %H:%M:%S')
        elif isinstance(obj, datetime.date):
            return obj.strftime('%Y-%m-%d')
        else:
            return json.JSONEncoder.default(self, obj)


res = {"user_id": 190027134, "nick": "愚蠢的萝卜", "reg_ts": datetime.datetime.now(), "last_login_date": datetime.date.today()}
print(json.dumps(res, ensure_ascii=False, indent=4, cls=DatetimeEncoder))

# {
    "nick": "愚蠢的萝卜", 
    "user_id": 190027134, 
    "reg_ts": "2019-11-03 20:09:55", 
    "last_login_date": "2019-11-03"
}