python笔记：pickle、cPickle模块

pickle包

1) 将内存中的对象转换成为文本流：

import pickle

# define class
class Contact(object):
    def __init__(self, uid, name, age):

	self.uid = uid

	self.age = age

	self.name = name

c = Contact(1001,'aaa',25)

ps = pickle.dumps(c)  #转换成字符串

c_copy = pickle.loads(ps) #转换成对象

使用pickle.dumps()方法可以将对象c转换成了字符串 

当然，我们也可以使用pickle.dump()的方法，将上面两部合二为一:

import pickle

class Contact(object):
    def __init__(self, uid, name, age):
	self.uid = uid
	self.age = age
	self.name = name
c = Contact(1005)with open('./tmp.data', 'w') as f:          
    ps = pickle.dump(summer, f)

对象c存储在文件ymp.data

 

2) 重建对象

首先，我们要从文本中读出文本，存储到字符串 (文本文件的输入输出)。然后使用pickle.loads(str)的方法，将字符串转换成为对象。要记得，此时我们的程序中必须已经有了该对象的类定义。

 

此外，我们也可以使用pickle.load()的方法，将上面步骤合并:

import pickle

class Contact(object):
    def __init__(self, uid, name, age):
	self.uid = uid
	self.age = age
	self.name = name
fn     = './tmp.data'with open(fn, 'r') as f:    summer = pickle.load(f)

 

 

cPickle包

cPickle包的功能和用法与pickle包几乎完全相同 (其存在差别的地方实际上很少用到)，不同在于cPickle是基于c语言编写的，速度是pickle包的1000倍。对于上面的例子，如果想使用cPickle包，我们都可以将import语句改为:

import cPickle as pickle

就不需要再做任何改动了。

下面时存储数据的小实例：

contact.py

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import copy
import json

import base

__all__ = ['Contact']

class Contact(base.Base):
    '''联系人'''
    DEBUG = True

    def __init__(self, **kwargs):
        self.init(**kwargs)

    def init(self, **kwargs):
        self.uid = kwargs.pop('uid', 0)
        self.name = kwargs.pop('name', '') 
        self.age = kwargs.pop('age', 0)
        self.addr = kwargs.pop('addr','')
        self.phone = kwargs.pop('phone', '') 
        self.ext = json.loads(kwargs.pop('ext', '{}'))

    def __str__(self):
        return "id:%s,uid:%s,name:%s,age:%s,%s"%(id(self), self.uid,self.name, self.age,','.join(['%s:%s'%(k,v) for k, v in self.ext.iteritems()]))

    def __getattr__(self, k): 
        if Contact.DEBUG:
            print 'in Contact __getattr__ %s'%k
        return self.ext.get(k)

    def get_key(self):
        return self.uid

    def changed(self):
        if getattr(self, 'manager'):
            self.manager.changed = True

    def set_age(self, age):
        assert type(age)==int, 'age must be interge'
        self.age = age 
        self.changed()

 def set_uid(self, uid):
        assert type(uid)==int, 'uid must be interge'
        self.uid = uid
        self.changed()

    def set_name(self, name):
        assert type(name)==str, 'name must be string'
        self.name = name
        self.changed()

    def load(self, **kwargs):
        return self.init(**kwargs)

    def save(self):
        m = self.__dict__.pop('manager', None)
        data = copy.deepcopy(self.__dict__)
        self.manager = m
        data['ext'] = json.dumps(data['ext'])
        return data

loader.py

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import os
import sys 
import cPickle


class Loader(object):
    '''cPickle loader'''

    def __init__(self, filename=None):
        self.filename = filename

    def load(self, filename=None):
        data = []
        filename = filename or self.filename
        try:
            with open(filename or self.filename, 'r') as fd: 
                data = cPickle.load(fd)
        except:
            print 'load data failed'
        return data

    def save(self, data, filename=None):
        assert (filename or self.filename), 'give a filename to save data'
        try:
            with open(filename or self.filename, 'w') as fd: 
                cPickle.dump(data, fd) 
        except:
            return False
        return True

manager.py

#!/usr/bin/env python
#-*- coding:utf-8 -*-

class Manager(object):
    '''联系人管理类'''
    def __init__(self, memclass, loader, condition=None):
        ''''''
        self.changed = False
        self.loader = loader
        self.memclass = memclass
        self.member = self.loader.load(condition) if condition else {}

    def add_mem(self, memobj):
        okey = memobj.get_key()
        if okey in self.member:
            return False
        memobj.manager = self

<span style="white-space:pre">	</span>self.changed = True
        self.member[okey] = memobj
        return True

    def del_mem(self, memobj=None, mkey=None):
        obj = None
        if memobj:
            obj = self.member.pop(memobj.get_key(), None)
        if mkey:
            obj = self.member.pop(mkey, None)
        if obj:

            <span style="font-family: 'courier new', courier; font-size: 13.3333330154419px;">obj.manager = None</span>

        return obj 

    def find_mem(self, mkey):
        return self.member.get(mkey, None)

    def load(self, condition=None):
        assert not self.changed, "the member is not null, you should save first"
        data = self.loader.load(condition)
        for e in data:
            if isinstance(e, self.memclass):
                memobj = e 
            else:
                memobj = self.memclass(**e)
            memobj.manager = self
            self.member[memobj.get_key()] = memobj

    def save(self):
        if self.changed:
            data = []
            for _, memobj in self.member.iteritems():
                data.append(memobj.save())
            self.loader.save(data)
        self.changed = False

    def display(self):
        for _, m in self.member.iteritems():
            print m

main.py

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import contact
import loader
import manager


def main():
    tmp_loader = loader.Loader('./tmp.data')
    tmp_manager = manager.Manager(contact.Contact, tmp_loader)
    tmp_contact = contact.Contact()
    tmp_contact.set_uid(1001)
    tmp_contact.set_age(25)
    tmp_contact.set_name('nnn')
    tmp_manager.add_mem(tmp_contact)
    tmp_manager.save()

if __name__=='__main__':
    main()