Question

I Have data structure like the one mentioned below which us all nested in terms of dictionaries : 
{
'list':{
'5':{
'set_profile_type':'static',
'status_name_now':'OFF',
'name':'BTテザリング',
'status_code_now':'OFF',
'status_name_init':'OFF',
'status_list':{
'0':{},
'1':{}
},
'guide':'some utf text',
'choice_type':'select',
'get_profile':'BTtethering_GetOnOff',
'status_key_name':'onOff'
},
'4':{
'set_profile_type':'static',
'status_name_now':'OFF',
'name':'WiFiテザリング',
'status_code_now':'OFF',
'status_name_init':'OFF',
'status_list':{
'0':{},
'1':{}
},
'guide':'UTF text',
'choice_type':'select',
'get_profile':'PortableHotspot_GetOnOff',
'status_key_name':'onOff'
},
'2':{},
'0':{},
'6':{},
'1':{},
'3':{}
},
'name':'somePropert1'
}

当我试图将这个结构变成csv时，我用json.loads尝试了pandas库，上面的字符串是由于我使用了unserialize的ohpserialization库而形成的;是否有一个很好的和干净的方式来展平这个数据结构我也尝试过json_normalize但是这不能识别上面的字符串为json-

    def deserialize(data):
    import phpserialize
    import ast
    import json
    data = "'"+data+"'"
    data = ast.literal_eval(format(data)).encode('utf-8')
    j1 = phpserialize.unserialize(data)
    j1 = phpserialize.loads(phpserialize.dumps(j1), decode_strings=True)
    j1 = json.dumps(j1)
    return j1
    def CustomParser(data):
    import json
    j1 = json.loads(data)
    j1 = j1
    return j1
    df = pandas.read_csv(input_file,sep="|",converters=
    {'device_udate':deserialize},header=0)
    print (df)
    df.to_csv(output_file,sep=",",encoding="utf-8")
    df1 = pandas.read_csv(output_file,converters=
    {'device_udate':CustomParser},header=0)
    df1[sorted(df1['device_udate'][0].keys())] = 
    df1['device_udate'].apply(pandas.Series)
    print (df1)

展平具有多级字典嵌套的未序列化嵌套json数据

0 个答案: