python 数据序列化 详解Python之数据序列化（json、pickle、shelve）

# 序列化：将Python对象转换成json字符串
dumps(obj, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, cls=None, indent=None, separators=None, default=None, sort_keys=False, **kw)

# 反序列化：将json字符串转换成Python对象
loads(s, encoding=None, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw)

# 序列化：将Python对象转换成json字符串并存储到文件中
dump(obj, fp, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, cls=None, indent=None, indent=None, separators=None, default=None, sort_keys=False, **kw)

# 反序列化：读取指定文件中的json字符串并转换成Python对象
load(fp, cls=None, object_hook=None, parse_float=None, parse_int=None, parse_constant=None, object_pairs_hook=None, **kw)

# 序列化
>>> json.dumps({'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)})
'{"a": "str", "c": true, "b": 11.1, "e": 10, "d": null, "g": [4, 5, 6], "f": [1, 2, 3]}'

# 序列化并对key进行排序
>>> json.dumps({'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}, sort_keys=True)
'{"a": "str", "b": 11.1, "c": true, "d": null, "e": 10, "f": [1, 2, 3], "g": [4, 5, 6]}'

# 序列化并对key进行排序及格式化输出
>>> print(json.dumps({'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}, sort_keys=True, indent=4)) 
{
 "a": "str", 
 "b": 11.1, 
 "c": true, 
 "d": null, 
 "e": 10, 
 "f": [
  1, 
  2, 
  3
 ], 
 "g": [
  4, 
  5, 
  6
 ]
}

>>> json.dumps({'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)})
'{"a": "str", "c": true, "b": 11.1, "e": 10, "d": null, "g": [4, 5, 6], "f": [1, 2, 3]}'

>>> json.dumps({'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}, separators=(',',':'))
'{"a":"str","c":true,"b":11.1,"e":10,"d":null,"g":[4,5,6],"f":[1,2,3]}'

# 反序列化
>>> json.loads('{"a": "str", "c": true, "b": 11.1, "e": 10, "d": null, "g": [4, 5, 6], "f": [1, 2, 3]}')
{'c': True, 'e': 10, 'a': 'str', 'g': [4, 5, 6], 'd': None, 'f': [1, 2, 3], 'b': 11.1}

>>> json.loads('{"a":"str","c":true,"b":11.1,"e":10,"d":null,"g":[4,5,6],"f":[1,2,3]}')
{'c': True, 'e': 10, 'a': 'str', 'g': [4, 5, 6], 'd': None, 'f': [1, 2, 3], 'b': 11.1}

# 序列化到文件中
>>> with open('test.json', 'w') as fp:
...  json.dump({'a':'str中国', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}, fp, indent=4)

# 反序列化文件中的内容
>>> with open('test.json', 'r') as fp:
...  json.load(fp)
{'e': 10, 'g': [4, 5, 6], 'b': 11.1, 'c': True, 'd': None, 'a': 'str中国', 'f': [1, 2, 3]}

class Student(object):
 def __init__(self, name, age, sno):
  self.name = name
  self.age = age
  self.sno = sno
 
 def __repr__(self):
  return 'Student [name: %s, age: %d, sno: %d]' % (self.name, self.age, self.sno)

>>> stu = Student('Tom', 19, 1)
>>> print(stu)
Student [name: Tom, age: 19, sno: 1]
>>>
>>> json.dumps(stu)
...
TypeError: Student [name: Tom, age: 19, sno: 1] is not JSON serializable

def obj2dict(obj):
 d = {}
 d['__class__'] = obj.__class__.__name__
 d['__module__'] = obj.__module__
 d.update(obj.__dict__)
 return d

def dict2obj(d):
 if '__class__' in d:
  class_name = d.pop('__class__')
  module_name = d.pop('__module__')
  module = __import__(module_name)
  class_ = getattr(module, class_name)
  args = dict((key.encode('ascii'), value) for key, value in d.items())
  instance = class_(**args)
 else:
  instance = d
 return instance

>>> for chunk in MyJSONEncoder().iterencode(stu):
...  print(chunk)
...
{
"__class__"
:
"Student"
,
"name"
:
"Tom"
,
"__module__"
:
"__main__"
,
"sno"
:
1
,
"age"
:
19
}

for chunk in JSONEncoder().iterencode(bigobject):
 mysocket.write(chunk)

>>> import json

>>> obj2dict(stu)
{'sno': 1, '__module__': '__main__', 'age': 19, '__class__': 'Student', 'name': 'Tom'}

>>> json.dumps(obj2dict(stu))
'{"sno": 1, "__module__": "__main__", "age": 19, "__class__": "Student", "name": "Tom"}'

>>> json.dumps(stu, default=obj2dict)
'{"sno": 1, "__module__": "__main__", "age": 19, "__class__": "Student", "name": "Tom"}'

>>> json.loads('{"sno": 1, "__module__": "__main__", "age": 19, "__class__": "Student", "name": "Tom"}')
{u'sno': 1, u'__module__': u'__main__', u'age': 19, u'name': u'Tom', u'__class__': u'Student'}

>>> dict2obj(json.loads('{"sno": 1, "__module__": "__main__", "age": 19, "__class__": "Student", "name": "Tom"}'))
Student [name: Tom, age: 19, sno: 1]

>>> json.loads('{"sno": 1, "__module__": "__main__", "age": 19, "__class__": "Student", "name": "Tom"}', object_hook=dict2obj)
Student [name: Tom, age: 19, sno: 1]

import json

class MyJSONEncoder(json.JSONEncoder):
 def default(self, obj):
  d = {}
  d['__class__'] = obj.__class__.__name__
  d['__module__'] = obj.__module__
  d.update(obj.__dict__)
  return d

class MyJSONDecoder(json.JSONDecoder):
 def __init__(self):
  json.JSONDecoder.__init__(self, object_hook=self.dict2obj)
 
 def dict2obj(self, d):
  if '__class__' in d:
   class_name = d.pop('__class__')
   module_name = d.pop('__module__')
   module = __import__(module_name)
   class_ = getattr(module, class_name)
   args = dict((key.encode('ascii'), value) for key, value in d.items())
   instance = class_(**args)
  else:
   instance = d
  return instance

>>> stu = Student('Tom', 19, 1)

# 方式一：直接调用子类MyJSONEncoder的encode()方法进行序列化
>>> MyJSONEncoder().encode(stu)
'{"__class__": "Student", "__module__": "__main__", "name": "Tom", "age": 19, "sno": 1}'
>>> MyJSONEncoder(separators=(',', ':')).encode(stu)
'{"__class__":"Student","__module__":"__main__","name":"Tom","age":19,"sno":1}'

# 方式二：将子类MyJSONEncoder作为cls参数的值传递给json.dumps()函数
>>> json.dumps(stu, cls=MyJSONEncoder)
'{"__class__": "Student", "__module__": "__main__", "name": "Tom", "age": 19, "sno": 1}'
>>> json.dumps(stu, cls=MyJSONEncoder, separators=(',', ':'))
'{"__class__":"Student","__module__":"__main__","name":"Tom","age":19,"sno":1}'

>>> MyJSONDecoder().decode('{"sno": 1, "__module__": "__main__", "age": 19, "__class__": "Student", "name": "Tom"}')
Student [name: Tom, age: 19, sno: 1]

# 将指定的Python对象通过pickle序列化作为bytes对象返回，而不是将其写入文件
dumps(obj, protocol=None, *, fix_imports=True)

# 将通过pickle序列化后得到的字节对象进行反序列化，转换为Python对象并返回
loads(bytes_object, *, fix_imports=True, encoding="ASCII", errors="strict")

# 将指定的Python对象通过pickle序列化后写入打开的文件对象中，等价于`Pickler(file, protocol).dump(obj)`
dump(obj, file, protocol=None, *, fix_imports=True)

# 从打开的文件对象中读取pickled对象表现形式并返回通过pickle反序列化后得到的Python对象
load(file, *, fix_imports=True, encoding="ASCII", errors="strict")

>>> import pickle
>>> 
>>> var_a = {'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}

# 序列化
>>> var_b = pickle.dumps(var_a)
>>> var_b
"(dp0\nS'a'\np1\nS'str'\np2\nsS'c'\np3\nI01\nsS'b'\np4\nF11.1\nsS'e'\np5\nI10\nsS'd'\np6\nNsS'g'\np7\n(I4\nI5\nI6\ntp8\nsS'f'\np9\n(lp10\nI1\naI2\naI3\nas."

# 反序列化
>>> var_c = pickle.loads(var_b)
>>> var_c
{'a': 'str', 'c': True, 'b': 11.1, 'e': 10, 'd': None, 'g': (4, 5, 6), 'f': [1, 2, 3]}

>>> import pickle
>>>
>>> var_a = {'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}

# 序列化
>>> var_b = pickle.dumps(var_a)
>>> var_b
b'\x80\x03}q\x00(X\x01\x00\x00\x00eq\x01K\nX\x01\x00\x00\x00aq\x02X\x03\x00\x00\x00strq\x03X\x01\x00\x00\x00fq\x04]q\x05(K\x01K\x02K\x03eX\x01\x00\x00\x00gq\x06K\x04K\x05K\x06\x87q\x07X\x01\x00\x00\x00bq\x08G@&333333X\x01\x00\x00\x00cq\t\x88X\x01\x00\x00\x00dq\nNu.'

# 反序列化
>>> var_c = pickle.loads(var_b)
>>> var_c
{'e': 10, 'a': 'str', 'f': [1, 2, 3], 'g': (4, 5, 6), 'b': 11.1, 'c': True, 'd': None}

>>> import pickle
>>>
>>> var_a = {'a':'str', 'c': True, 'e': 10, 'b': 11.1, 'd': None, 'f': [1, 2, 3], 'g':(4, 5, 6)}

# 持久化到文件
>>> with open('pickle.txt', 'wb') as f:
...  pickle.dump(var_a, f)
...

# 从文件中读取数据
>>> with open('pickle.txt', 'rb') as f:
...  var_b = pickle.load(f)
...
>>> var_b
{'e': 10, 'a': 'str', 'f': [1, 2, 3], 'g': (4, 5, 6), 'b': 11.1, 'c': True, 'd': None}
>>>

class Student(object):
 def __init__(self, name, age, sno):
  self.name = name
  self.age = age
  self.sno = sno
 
 def __repr__(self):
  return 'Student [name: %s, age: %d, sno: %d]' % (self.name, self.age, self.sno)

>>> stu = Student('Tom', 19, 1)
>>> print(stu)
Student [name: Tom, age: 19, sno: 1]

# 序列化
>>> var_b = pickle.dumps(stu)
>>> var_b
b'\x80\x03c__main__\nStudent\nq\x00)\x81q\x01}q\x02(X\x04\x00\x00\x00nameq\x03X\x03\x00\x00\x00Tomq\x04X\x03\x00\x00\x00ageq\x05K\x13X\x03\x00\x00\x00snoq\x06K\x01ub.'

# 反序列化
>>> var_c = pickle.loads(var_b)
>>> var_c
Student [name: Tom, age: 19, sno: 1]

# 持久化到文件
>>> with open('pickle.txt', 'wb') as f:
...  pickle.dump(stu, f)
...

# 从文件总读取数据
>>> with open('pickle.txt', 'rb') as f:
...  pickle.load(f)
...
Student [name: Tom, age: 19, sno: 1]

open(filename, flag='c', protocol=None, writeback=False)

# 保存数据
with shelve.open('student') as db:
 db['name'] = 'Tom'
 db['age'] = 19
 db['hobby'] = ['篮球', '看电影', '弹吉他']
 db['other_info'] = {'sno': 1, 'addr': 'xxxx'}

# 读取数据
with shelve.open('student') as db:
 for key,value in db.items():
  print(key, ': ', value)

# 自定义class
class Student(object):
 def __init__(self, name, age, sno):
  self.name = name
  self.age = age
  self.sno = sno
 
 def __repr__(self):
  return 'Student [name: %s, age: %d, sno: %d]' % (self.name, self.age, self.sno)

# 保存数据
tom = Student('Tom', 19, 1)
jerry = Student('Jerry', 17, 2)

with shelve.open("stu.db") as db:
 db['Tom'] = tom
 db['Jerry'] = jerry

# 读取数据
with shelve.open("stu.db") as db:
 print(db['Tom'])
 print(db['Jerry'])