json文件数据解析 -- Python3.6 字典数据的处理

 下面是距离现在两个月左右的一个入门级别的json数据解析，记录一下。

任务目标：将ori1.json和ori2.json这两个文件中的数据格式（见图1），转换成res1.json和res2.json的格式（见图2）

图1

 

图2

 这是字典类型的数据，本质是两种格式的转换就是字典的key和value之间互换。例如，第一组数据：在ori1.json中找到包含value为“vo”的所有的key:"K6","k8"，然后"V0"作为key,"K6","k8"作为value。

下面是ori1.json的数据：

{
  "K2": [
    "V5",
    "V9",
    "V1",
    "V5"
  ],
  "K5": [
    "V8"
  ],
  "K9": [
    "V7",
    "V6",
    "V3"
  ],
  "K1": [
    "V7",
    "V6",
    "V5"
  ],
  "K3": [
    "V7"
  ],
  "K6": [
    "V0"
  ],
  "K4": [
    "V9",
    "V1"
  ],
  "K8": [
    "V3",
    "V0"
  ],
  "K0": [
    "V9",
    "V3",
    "V1"
  ]
}

res1.json的数据：

{
  "V0": [
    "K6",
    "K8"
  ],
  "V1": [
    "K0",
    "K2",
    "K4"
  ],
  "V3": [
    "K0",
    "K8",
    "K9"
  ],
  "V5": [
    "K1",
    "K2",
    "K2"
  ],
  "V6": [
    "K1",
    "K9"
  ],
  "V7": [
    "K1",
    "K3",
    "K9"
  ],
  "V8": [
    "K5"
  ],
  "V9": [
    "K0",
    "K2",
    "K4"
  ]
}

---

实现代码：

#!/usr/bin/python3
import json

def Rad():
    with open('ori1.json',mode='r+') as f:
        data = json.load(f)
    return data

def Wrt(data):
    with open('1.json', mode='w+') as f:
        json.dump(data,f,ensure_ascii=False,indent=2)  #json文件“建行”写入

def Opt1(data): #取出字典中每个key对应的值，并把每个列表中的值取出，去重，排序，得到新的key
    list_2 = []
    list_3 = []
    list_4 = []
    for i in data.values():
        list_2.append(i)
    for i in range(len(list_2)):
        for j in range(len(list_2[i])):
            list_3.append(list_2[i][j])
    list_4 = list(set(list_3))
    list_4.sort(reverse=False)
    return list_4

def Opt2(list_1,data):# 根据新的key 就是旧的value返回 旧的key，作为新的value
    dict1 ={}
    list_5 = []
    for k in range(len(list_1)):
        for i in data.values():
            for j in range(len(i)):
                if i[j] == list_1[k]:
                   list_5.append(list (data.keys()) [list (data.values()).index (i)])# 根据value 返回key
        for a in range(len(list_5)):
            dict1.setdefault(list_1[k],[]).append(str(list_5[a])) # 把value 写入对应的key下面
        list_5.clear()
    print(dict1)
    return dict1

def main():
  data=Rad()
  list_1=Opt1(data)
  dict1=Opt2(list_1,data)
  Wrt(dict1)

if __name__ == '__main__':
        main()

---

运行结果展示：

{
  "V0": [
    "K6",
    "K8"
  ],
  "V1": [
    "K2",
    "K4",
    "K0"
  ],
  "V3": [
    "K9",
    "K8",
    "K0"
  ],
  "V5": [
    "K2",
    "K2",
    "K1"
  ],
  "V6": [
    "K9",
    "K1"
  ],
  "V7": [
    "K9",
    "K1",
    "K3"
  ],
  "V8": [
    "K5"
  ],
  "V9": [
    "K2",
    "K4",
    "K0"
  ]
}

1.json文件中的数据格式和res1.json中的个数完全一致。✌

---

实现是实现了，但是看完大佬的代码，就会觉得自己的特别麻烦，下面是大佬的代码，清爽帅气：

import os
import json
import numpy as np
from collections import OrderedDict


keys = ['K{}'.format(i) for i in range(10)]
vals = ['V{}'.format(i) for i in range(10)]


def create_data(keys, vals, iters = 20):
  """randomly create data formatted as follow:
  {
    K1: [
      V1, 
      V2, 
      Vn
    ], 
    K2: [
      V2, 
      V4, 
      Vn
    ],
    Kn: [
      Vi,
      Vn
    ]
  }
  """
  data = {}
  for i in range(iters):
    k_rd = np.random.choice(len(keys))
    v_rd = np.random.choice(len(vals))
    if keys[k_rd] not in data.keys():
      data[keys[k_rd]] = [vals[v_rd]]
    else:
      data[keys[k_rd]].append(vals[v_rd])
  return data
  
ori_fp = open('ori2.json', 'w')
ori_data = create_data(keys, vals)
json.dump(ori_data, ori_fp, indent = 2)
ori_fp.close()


def convert_data(ori):
  """convert data from K-V to V-K
  traverse all keys and the value, so we can map the value to different keys, for example:
  K1: [
    V1, 
    V2
  ], 
  K3: [
    V2,
    V3
  ]
  we can catch that 
    V1 - K1
    V2 - K1, K3
    V3 - K3
    
  than we sort them and save in an OrderdDict
  """
  res = {}
  for kv in ori.items():
    for vl in kv[1]:
      if vl not in res.keys():
        res[vl] = [kv[0]]
      else:
        res[vl].append(kv[0])
  return res
  
res_fp = open('res2.json', 'w')
sorted_ori = OrderedDict()
for kv in sorted(ori_data.items(), key = lambda x: x[0]):
  sorted_ori[kv[0]] = kv[1]
  
res_data = convert_data(sorted_ori)
sorted_res = OrderedDict()
for kv in sorted(res_data.items(), key = lambda x: x[0]):
  sorted_res[kv[0]] = kv[1]

json.dump(sorted_res, res_fp, indent = 2)
res_fp.close()

感觉学习Python需要对每一个有用的库都非常的熟悉，熟练掌握，不然有些操作就会很头疼。实际上的思路是一样的，但是一步步来的就会很慢，很麻烦。