python对mongodb关于json、unicode、string相关总结

最近接触了mongodb&＃xff0c;处理json和Unicode字符串遇到些问题总结下。

python2环境

mac os系统

# -*- coding: utf-8 -*-import jsonfrom pymongo import MongoClient
from bson import json_util, ObjectIdclient &＃61; MongoClient(&＃39;localhost&＃39;, 27017)
db &＃61; client[&＃39;tumblelog&＃39;]
collection &＃61; db.user
# user 查询出来是字典类型
user &＃61; collection.find_one({"_id": ObjectId(&＃39;5f3cdfac77f96de114e1e989&＃39;)})
# 对user利用python原生的json进行序列化
print json.dumps(user)


报出了如下错误信息&＃xff1a; 

TypeError: ObjectId(&＃39;5f3cdfac77f96de114e1e989&＃39;) is not JSON serializable

不能用python原生的json&＃xff0c;mongo提供了专门的json_util

# -*- coding: utf-8 -*-from pymongo import MongoClient
from bson import json_util, ObjectIdclient &＃61; MongoClient(&＃39;localhost&＃39;, 27017)
db &＃61; client[&＃39;tumblelog&＃39;]
collection &＃61; db.user
# user 查询出来是字典类型
user &＃61; collection.find_one({"_id": ObjectId(&＃39;5f3cdfac77f96de114e1e989&＃39;)})
user_str &＃61; json_util.dumps(user)
print (type(user_str), user_str)

成功序列化&＃xff1a; 

(, &＃39;{"last_name": "Lawley6", "first_name": "Ross6", "_id": {"$oid": "5f3cdfac77f96de114e1e989"}, "email": "ross&＃64;example.com6"}&＃39;)

另一个问题&＃xff1a;

mongodb查询出的数据都是Unicode类型的&＃xff0c;我想要string类型的&＃xff0c;这是python2的原因&＃xff0c;换成python3的环境没问题。

# -*- coding: utf-8 -*-import jsonfrom pymongo import MongoClient
from bson import json_util, ObjectIdclient &＃61; MongoClient(&＃39;localhost&＃39;, 27017)
db &＃61; client[&＃39;tumblelog&＃39;]
collection &＃61; db.user
user &＃61; collection.find_one({"_id": ObjectId(&＃39;5f3cdfac77f96de114e1e989&＃39;)})
print(type(user), user)

输出&＃xff0c;可以看出来都是Unicode类型&＃xff1a;  

(, {u&＃39;last_name&＃39;: u&＃39;Lawley6&＃39;, u&＃39;first_name&＃39;: u&＃39;Ross6&＃39;, u&＃39;_id&＃39;: ObjectId(&＃39;5f3cdfac77f96de114e1e989&＃39;), u&＃39;email&＃39;: u&＃39;ross&＃64;example.com6&＃39;})

 如果用不用python3的话&＃xff0c;可以通过yaml库的safe_load方法解决 

# -*- coding: utf-8 -*-import json
import yamlfrom pymongo import MongoClient
from bson import json_util, ObjectIdclient &＃61; MongoClient(&＃39;localhost&＃39;, 27017)
db &＃61; client[&＃39;tumblelog&＃39;]
collection &＃61; db.user
user &＃61; collection.find_one({"_id": ObjectId(&＃39;5f3cdfac77f96de114e1e989&＃39;)})
user_dict_str &＃61; yaml.safe_load(json_util.dumps(user))
print (type(user_dict_str), user_dict_str)

 通过safe_load方法转化后&＃xff0c;可以看出输出类型都是string了。

 (, {&＃39;first_name&＃39;: &＃39;Ross6&＃39;, &＃39;last_name&＃39;: &＃39;Lawley6&＃39;, &＃39;email&＃39;: &＃39;ross&＃64;example.com6&＃39;, &＃39;_id&＃39;: {&＃39;$oid&＃39;: &＃39;5f3cdfac77f96de114e1e989&＃39;}}) 

再记录一个python2的json对于中文乱码的处理,默认dump中文会乱码&＃xff0c;加一个ensure_ascii参数就好了

import jsonl &＃61; [&＃39;a&＃39;, &＃39;b&＃39;, &＃39;中文&＃39;]
print (l)
l_str &＃61; json.dumps(l)
print (l_str)
l_str &＃61; json.dumps(l, ensure_ascii&＃61;False)
print (l_str)

[&＃39;a&＃39;, &＃39;b&＃39;, &＃39;\xe4\xb8\xad\xe6\x96\x87&＃39;]
["a", "b", "\u4e2d\u6587"]
["a", "b", "中文"]