今天和大家分享一个python入库mongodb的脚本。。。
涉及到python和mongodb，那么安装相应的模块四必不可少的，最简单的安装方法，或者非pip不可了。
# pip install pymongo==3.0.4顺便也记录下源码安装的方式
# wget https://pypi.python.org/packages/source/p/pymongo/pymongo-2.8.tar.gz#md5=23100361c9af1904eb2d7722f2658114 --no-check-certificate
# tar xf pymongo-2.8.tar.gz
# cd pymongo-2.8
# python setup.py install摘自一则日志
35783 s100 android 47 5 192.168.1.100 2015-09-05 08:03:19 strengthenHeroByHeroes {"consume_gold":{"ogold":2893821,"cgold":1700,"gold":2892121,"tag":"strengthenHeroByHeroes"},"taskInfo":[{"id":2310033,"progress":2,"status":0}],"delHeroList":{"id":102014,"id":102014,"id":102014,"id":102010,"id":102010},"id":100026,"olevel":46,"oexp":1700,"cexp":1700,"level":46,"exp":3400} 865982021462182 XiaoMi入库mongodb的python脚本
[root@localhost opt]# cat analytical.py
#!/usr/bin/env python
#coding:utf8
import os,sys,json
from datetime import *
from pymongo import MongoClient
def ConMongo(host,port,cur_db,username,password):
client = MongoClient(host,port)
db = client[cur_db]
db.authenticate(username,password)
table = db.gamelogs
return table
def parseLog(logfile,table):
dic = {}
dl = []
with open(file_log) as fd:
for line in fd:
try:
tokens = line.strip().split('\t')
uid = tokens[0]
server = tokens[1]
system = tokens[2]
level = int(tokens[3])
vip_level = tokens[4]
ip = tokens[5]
time = datetime.strptime(tokens[6], "%Y-%m-%d %H:%M:%S") #将时间字符串转换成时间格式
action = tokens[7]
result = json.loads(tokens[8]) #特殊字符串转换成json格式
uuid = tokens[9]
if len(tokens) == 12:
channel = tokens[11]
else:
channel = ''
dic = {'uid':uid,'server':server,'system':system,'level':level,'vip_level':vip_level,'ip':ip,'time':time,'action':action,'result':result,'uuid':uuid,'channel':channel}
dl.append(dic)
if len(dl) == 10000:
table.insert_many(dl)
dl = []
except Exception,e:
print e, line
if len(dl) > 0:
table.insert_many(dl)
if __name__ == '__main__':
table = ConMongo('localhost',27017,'talefundb','talefun','123456')
try:
logfile = sys.argv[1]
parseLog(logfile,table)
except IndexError,e:
print e注意事项：
(1)insert_many参数是mongodb 3.0.4中新加的，允许你将一个大列表直接insert到mongodb数据库中
(2)脚本中做了限制，如果字典中有2000个值，就向mongodb插入一次数据，这样在效率上得到了保证
(3)不建议直接复制脚本测试，很多粘贴出来后，很多制表符等会出现问题。我会吧脚本放在云盘上大家可以下载，测试用。
点击可下载：http://pan.baidu.com/s/1qWtbgjq
