一、序列化/反序列化
python中内置了很多序列化/反序列化的方式,最常用的有json、pickle、marshal这三种,示例用法如下:
import json import pickle import marshal author1 = {"name": "菩提树下的杨过", "blog": "http://yjmyzz.cnblogs.com/", "title": "架构师", "pets": ["dog", "cat"]} # json序列化 json_str = json.dumps(author1) print("json=> ", json_str) # json字符串反序列化 author2 = json.loads(json_str) # pickle序列化 pickle_str = pickle.dumps(author1) print("pickle=> ", pickle_str) # pickle字符串反序列化 author3 = pickle.loads(pickle_str) # marshal序列化 marshal_str = marshal.dumps(author1) print("marshal=> ", marshal_str) # marshal反序列化 author4 = marshal.loads(marshal_str) print(" ", id(author1), " ", id(author2), " ", id(author3), " ", id(author4), " ", author1, " ", author2, " ", author3, " ", author4) with open("json.txt", "w") as file1: json.dump(author1, file1) with open("pickle.txt", "wb") as file2: pickle.dump(author1, file2) with open("marshal.txt", "wb") as file3: marshal.dump(author1, file3)
输出:
json=> {"name": "u83e9u63d0u6811u4e0bu7684u6768u8fc7", "blog": "http://yjmyzz.cnblogs.com/", "title": "u67b6u6784u5e08", "pets": ["dog", "cat"]} pickle=> b'x80x03}qx00(Xx04x00x00x00nameqx01Xx15x00x00x00xe8x8fxa9xe6x8fx90xe6xa0x91xe4xb8x8bxe7x9ax84xe6x9dxa8xe8xbfx87qx02Xx04x00x00x00blogqx03Xx1ax00x00x00http://yjmyzz.cnblogs.com/qx04Xx05x00x00x00titleqx05X x00x00x00xe6x9exb6xe6x9ex84xe5xb8x88qx06Xx04x00x00x00petsqx07]qx08(Xx03x00x00x00dogq Xx03x00x00x00catq eu.' marshal=> b'xfbxdax04namexf5x15x00x00x00xe8x8fxa9xe6x8fx90xe6xa0x91xe4xb8x8bxe7x9ax84xe6x9dxa8xe8xbfx87xdax04blogxfax1ahttp://yjmyzz.cnblogs.com/xdax05titlexf5 x00x00x00xe6x9exb6xe6x9ex84xe5xb8x88xdax04pets[x02x00x00x00xdax03dogxdax03cat0' 4307564944 4309277360 4307565016 4309277432 {'name': '菩提树下的杨过', 'blog': 'http://yjmyzz.cnblogs.com/', 'title': '架构师', 'pets': ['dog', 'cat']} {'name': '菩提树下的杨过', 'blog': 'http://yjmyzz.cnblogs.com/', 'title': '架构师', 'pets': ['dog', 'cat']} {'name': '菩提树下的杨过', 'blog': 'http://yjmyzz.cnblogs.com/', 'title': '架构师', 'pets': ['dog', 'cat']} {'name': '菩提树下的杨过', 'blog': 'http://yjmyzz.cnblogs.com/', 'title': '架构师', 'pets': ['dog', 'cat']}
注:api的方法名还是很好记的,dump/dumps意为“倒垃圾”,把对象向xxx里一倒,就算序列化完成了。反之load/loads即从字符串或文件中装载(还原)对象。特别要值得一提的是:pickle、marshal存在安全问题,如果装载的字符串或文件里,包含有精心设计的恶意代码,会让恶意代码执行(关于反序列化的漏洞,大家可以上网查一下,有很多类似的介绍)。另外从序列化后的字符串大小来看,默认情况下,就本示例而言,json序列化后的字符串长度最小,so,综合来看,推荐同学们使用json序列化/反序列化。
二、深拷贝、浅拷贝
import copy list_1 = [1, 2, 3, [4, 5]] list_2 = copy.copy(list_1) # 浅拷贝 list_3 = copy.deepcopy(list_1) # 深拷贝 list_2[3][0] = 99 print(" ", list_1, " ", list_2, " ", list_3) list_3[3][1] = 100 print(" ", list_1, " ", list_2, " ", list_3)
输出:
[1, 2, 3, [99, 5]] [1, 2, 3, [99, 5]] [1, 2, 3, [4, 5]] [1, 2, 3, [99, 5]] [1, 2, 3, [99, 5]] [1, 2, 3, [4, 100]]
当一个对象里的子元素本身也是复杂元素时,浅拷贝不会为这种复杂的子元素生成全新的实例,但深拷贝可以。下面的内存分布示意图有助于大家理解:
list_2是list_1浅拷贝生成的对象,对于第4个元素,都是指向同一个列表[4,5],所以list_2修改了[4,5]中的第1个元素为99后,list_1也受到影响。list_3则是深拷贝的结果,所有元素都是独立的新实例,因此修改list_3里的任何元素,都不会影响list_1、list_2
参考文档:
1. https://docs.python.org/3/library/pickle.html
2. https://docs.python.org/3/library/json.html
3. https://docs.python.org/3/library/marshal.html
4. https://docs.python.org/3/library/copy.html