3.4 .序列化模块

1. 序列化模块(非常非常重要 )

   • 序列化:将一个数据结构 (list,dict....)转化成一个特殊的序列(特殊的字符串)的过程.
   • 反序列化：将这个特殊的序列反转回原来的数据结构。

   l1 = [1,2,3]
   ret = str(l1)
   print(ret,type(ret))
   print(eval(ret))#不用
   文件存取时,遇到的矛盾.
   dic = {'username': '太白', 'password': 123}
   dic = {1: {'username': '太白', 'password': 123,'status': False},
          2:{'username': 'alex', 'password': 123,'status': False}
          } 
   with open('register.json',encoding = 'utf-8',mode='w+')as f:
   	f1.write(str(dic))
   	f.seek(0)
   	print(ret,type(ret))
   数据结构 --- >str()存储在文件 读取出来时 反转不回去(报错)
   

2. 网络传输

   l1 = [i for i in range(10)]
   凡是数据通过网络传出去最终的格式必须bytes
   s1 = str(l1)
   b1 = s1.encode('utf-8')
   print(b1)  #b'[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]'  b1可以发送出去
   s2 =b1.decode('utf-8')
   print(s2,type(s2))#[0, 1, 2, 3, 4, 5, 6, 7, 8, 9] <class 'str'>
   s2 转化不成列表了.
   

   我们现在要解决的问题: 如果有一种特殊的字符串,这个字符串可以与任何的数据结构互相转换.

3. 序列化模块: 将一中数据结构转化成特殊的序列(特殊的字符串,bytes)并且还可以反转回去.

   json模块: 是所有语言公认的一种序列.最最常用的
   所以支持的python数据结构有限: int str bool dict list(tuple),None,float
    None  ---> Null ----> NUll
    dict  --->
    pickle模块: 只能python语言中使用的,序列化模块:
    支持python所有的数据类型以及对象.
    shevle模块(不讲): 课下了解(只能是文件存取).
   

4. json序列化

   两种四个方法.

   dumps,loads

   import json
   dumps,loads 主要用于网络传输,可以用于文件的存取.
   dic = {'username': '太白', 'password': 123,'status': False}
   ret = json.dumps(dic)
   print(ret,type(ret))#{"username": "\u592a\u767d", "password": 123, "status": false} <class 'str'>
   ret_dict = json.loads(ret)
   print(ret_dict)#{'username': '太白', 'password': 123, 'status': False}<class 'dict'>
   

   dumps,loads 写入文件

   特殊的参数
   dic = {'username': '太白', 'password': 123,'status': False}
   ret = json.dumps(dic,ensure_ascii=False,sort_keys=True)
   print(ret,type(ret))#{"password": 123, "status": false, "username": "太白"} <class 'str'>
   以json写入文件,dumps转化成特殊的字符串
   s_dict = json.dumps(dic)
   with open('jsonlx.json',encoding='utf-8',mode='w') as f1:
   	f1.write(s_dict)
   with open('jsonlx.json',encoding='utf-8') as f2:
       content = f2.read()
       print(json.loads(content))
   

   dump load:单个数据的存取文件.

   dic = {'username': '太白', 'password': 123,'status': False}
   with open('jsonlx1.json',encoding='utf-8',mode='w') as f1:
       json.dump(dic,f1)
   with open('jsonlx1.json',encoding='utf-8') as f1:
       dic1 = json.load(f1)
   print(dic1,type(dic1))
   

5. 多个数据如何存储到一个文件中?

   # 错误演示:
   # import json
   # dic1 = {'username': '太白', 'password': 123,'status': False}
   # dic2 = {'username': 'alex', 'password': 123,'status': False}
   # dic3 = {'username': 'ly', 'password': 123,'status': False}
   # with open('jsonmore.json',encoding='utf-8',mode='a') as f1:
   #     # f1.write(json.dumps(dic1))
   #     # f1.write(json.dumps(dic2))
   #     # f1.write(json.dumps(dic3))
   #     f1.write(f'{json.dumps(dic1)}{json.dumps(dic2)}{json.dumps(dic3)}')
   # with open('jsonmore.json',encoding='utf-8') as f1:
   #     ret = json.loads(f1.read())
   #     print(ret)
   

   正确做法:
   dic1 = {'username': '太白', 'password': 123,'status': False}
   dic2 = {'username': 'alex', 'password': 123,'status': False}
   dic3 = {'username': 'ly', 'password': 123,'status': False}
   with open('jsonmore.json',encoding='utf-8',mode='a') as f1:
     f1.write(f'{json.dumps(dic1)}\n{json.dumps(dic2)}\n{json.dumps(dic3)}')
   with open('jsonmore.json',encoding='utf-8') as f1:
       for line in f1:
           ret = json.loads(line)
           print(ret,type(ret))
   

6. pickle:pickle模块是将Python所有的数据结构以及对象等转化成bytes类型，然后还可以反序列化还原回去。

   dumps,loads 只能是网络传输

   import pickle
   l1 = ['wusir', '太白', '小黑', 666]
   ret = pickle.dumps(l1)
   print(ret)#b'\x80\x03]q\x00(X\x05\x00\x00\x00wusirq\x01X\x06\x00\x00\x00\xe5\xa4\xaa\xe7\x99\xbdq\x02X\x06\x00\x00\x00\xe5\xb0\x8f\xe9\xbb\x91q\x03M\x9a\x02e.'
   l2 = pickle.loads(ret)
   print(l2,type(l2))#['wusir', '太白', '小黑', 666] <class 'list'>
   

   dump load 数据结构存取文件.

   写入单个文件

   import pickle
   l1 = ['wusir', '太白', '小黑', 666]
   with open('pickle练习.pickle',mode='wb') as f1:
       pickle.dump(l1,f1)
   with open('pickle练习.pickle', mode='rb') as f1:
       ret = pickle.load(f1)
       print(ret,type(ret))
   

   多个数据写入文件

   l1 = ['wusir', '太白', '小黑1', 666]
   l2 = ['wusir', '太白', '小黑2', 666]
   l3 = ['wusir', '太白', '小黑3', 666]
   with open('pickle练习1.pickle',mode='wb') as f1:
       pickle.dump(l1,f1)
       pickle.dump(l2,f1)
       pickle.dump(l3,f1)
   with open('pickle练习1.pickle', mode='rb') as f1:
       ret1 = pickle.load(f1)
       ret2 = pickle.load(f1)
       ret3 = pickle.load(f1)
       print(ret1,ret2,ret3)
       #['wusir', '太白', '小黑1', 666] ['wusir', '太白', '小黑2', 666] ['wusir', '太白', '小黑3', 666]