$remote_addr - $remote_user [$time_local] "$request" $status $body_bytes_sent "$http_referer" "$http_user_agent" "$http_x_forwarded_for"' #使用的是nginx默認日志格式

182.19.31.129 - - [2013-08-13T00:00:01-07:00] "GET /css/anniversary.css HTTP/1.1" 304 0 "http://www.chlinux.net/" "Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/28.0.1500.95 Safari/537.36" "-"

#!/usr/bin/env python
#coding:utf8
import os
import fileinput
import re
import sys
import MySQLdb
#日志的位置
logfile=open("access_20130812.log")
#使用的nginx默認日志格式$remote_addr - $remote_user [$time_local] "$request" $status $body_bytes_sent "$http_referer" "$http_user_agent" "$http_x_forwarded_for"'
#日志分析正則表達式
#203.208.60.230
ipP = r"?P<ip>[\d.]*"
#以[開始,除[]以外的任意字符 防止匹配上下個[]項目(也可以使用非貪婪匹配*?) 不在中括號里的.可以匹配換行外的任意字符 *這樣地重復是"貪婪的“ 表達式引擎會試著重復盡可能多的次數(shù)。#以]結(jié)束
#[21/Jan/2011:15:04:41 +0800]
timeP = r"""?P<time>\[[^\[\]]*\]"""
#以"開始, #除雙引號以外的任意字符 防止匹配上下個""項目(也可以使用非貪婪匹配*?),#以"結(jié)束
#"GET /EntpShop.do?method=view&shop_id=391796 HTTP/1.1"
#"GET /EntpShop.do?method=view&shop_id=391796 HTTP/1.1"
requestP = r"""?P<request>\"[^\"]*\""""
statusP = r"?P<status>\d+"
bodyBytesSentP = r"?P<bodyByteSent>\d+"
#以"開始, 除雙引號以外的任意字符 防止匹配上下個""項目(也可以使用非貪婪匹配*?),#以"結(jié)束
#"http://test.myweb.com/myAction.do?method=view&mod_id=&id=1346"
referP = r"""?P<refer>\"[^\"]*\""""
#以"開始, 除雙引號以外的任意字符 防止匹配上下個""項目(也可以使用非貪婪匹配*?),以"結(jié)束
#"Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)"'
userAgentP = r"""?P<userAgent>\"[^\"]*\""""
#以(開始, 除雙引號以外的任意字符 防止匹配上下個()項目(也可以使用非貪婪匹配*?),以"結(jié)束
#(compatible; Googlebot/2.1; +http://www.google.com/bot.html)"'
userSystems = re.compile(r'\([^\(\)]*\)')
#以"開始，除雙引號以外的任意字符防止匹配上下個""項目(也可以使用非貪婪匹配*?),以"結(jié)束
userlius = re.compile(r'[^\)]*\"')
#原理：主要通過空格和-來區(qū)分各不同項目，各項目內(nèi)部寫各自的匹配表達式
nginxLogPattern = re.compile(r"(%s)\ -\ -\ (%s)\ (%s)\ (%s)\ (%s)\ (%s)\ (%s)" %(ipP, timeP, requestP, statusP, bodyBytesSentP, referP, userAgentP), re.VERBOSE)
#數(shù)據(jù)庫連接信息
conn=MySQLdb.connect(host='192.168.1.22',user='test',passwd='pass',port=3306,db='python')
cur=conn.cursor()
sql = "INSERT INTO python.test VALUES(%s,%s,%s,%s,%s,%s,%s,%s,%s)"
while True:
    line = logfile.readline()
    if not line:break
    matchs = nginxLogPattern.match(line)
    if matchs != None:
        allGroup = matchs.groups()
        ip = allGroup[0]
        time = allGroup[1]
        request = allGroup[2]
        status = allGroup[3]
        bodyBytesSent = allGroup[4]
        refer = allGroup[5]
        userAgent = allGroup[6]
        Time = time.replace('T',' ')[1:-7]
        if len(userAgent) > 20:
            userinfo = userAgent.split(' ')
            userkel =  userinfo[0]
            try:
                usersystem = userSystems.findall(userAgent)
                usersystem = usersystem[0]
                print usersystem
                userliu = userlius.findall(userAgent)
                value = [ip,Time,request,status,bodyBytesSent,refer,userkel,usersystem,userliu[1]]
                conn.commit()
                print value
            except IndexError:
                userinfo = userAgent
                value = [ip,Time,request,status,bodyBytesSent,refer,userinfo,"",""]
        else:
            useraa = userAgent
            value = [ip,Time,request,status,bodyBytesSent,refer,useraa,"",""]
    try:
        result = cur.execute(sql,value)
        #conn.commit()
        print result
    except MySQLdb.Error,e:
        print "Mysql Error %d: %s" % (e.args[0], e.args[1])
conn.commit()
conn.close()