Python 每小时计数日志文件

Question

我有一个 python 脚本，它在日志文件中显示攻击的日期、时间和 IP 地址。我的问题是我需要能够计算每天每小时发生多少次攻击，但是当我实施计数时，它只计算总数而不是我想要的。

日志文件如下所示：

Feb  3 08:50:39 j4-be02 sshd[620]: Failed password for bin from 211.167.103.172 port 39701 ssh2
Feb  3 08:50:45 j4-be02 sshd[622]: Failed password for invalid user virus from 211.167.103.172 port 41354 ssh2
Feb  3 08:50:49 j4-be02 sshd[624]: Failed password for invalid user virus from 211.167.103.172 port 42994 ssh2
Feb  3 13:34:00 j4-be02 sshd[666]: Failed password for root from 85.17.188.70 port 45481 ssh2
Feb  3 13:34:01 j4-be02 sshd[670]: Failed password for root from 85.17.188.70 port 46802 ssh2
Feb  3 13:34:03 j4-be02 sshd[672]: Failed password for root from 85.17.188.70 port 47613 ssh2
Feb  3 13:34:05 j4-be02 sshd[676]: Failed password for root from 85.17.188.70 port 48495 ssh2
Feb  3 21:45:18 j4-be02 sshd[746]: Failed password for invalid user test from 62.45.87.113 port 50636 ssh2
Feb  4 08:39:46 j4-be02 sshd[1078]: Failed password for root from 1.234.51.243 port 60740 ssh2
Feb  4 08:39:55 j4-be02 sshd[1082]: Failed password for root from 1.234.51.243 port 34124 ssh2

我目前的代码是：

import re

myAuthlog=open('auth.log', 'r') #open the auth.log for reading
for line in myAuthlog: #go through each line of the file and return it to the variable line
ip_addresses = re.findall(r'([A-Z][a-z]{2}\s\s\d\s\d\d).+Failed password for .+? from (\S+)', line)

print ip_addresses

结果如图所示

[('Feb  5 08', '5.199.133.223')]
[]
[('Feb  5 08', '5.199.133.223')]
[]
[('Feb  5 08', '5.199.133.223')]
[]
[('Feb  5 08', '5.199.133.223')]
[]
[('Feb  5 08', '5.199.133.223')]

Answer 1

python 函数groupby() 将根据您指定的任何标准对您的项目进行分组。

此代码将打印每天每小时的攻击次数：

from itertools import groupby

with open('auth.log') as myAuthlog:
    for key, group in groupby(myAuthlog, key = lambda x: x[:9]):
        print "%d attacks in hour %s"%(len(list(group)), key)

或者，来自 cmets 的额外要求：

from itertools import groupby

with open('auth.log') as myAuthlog:
    myAuthlog = (line for line in myAuthlog if "Failed password for" in line)
    for key, group in groupby(myAuthlog, key = lambda x: x[:9]):
        print "%d attacks in hour %s"%(len(list(group)), key)

或者，使用不同的格式：

from itertools import groupby

with open('auth.log') as myAuthlog:
    myAuthlog = (line for line in myAuthlog if "Failed password for" in line)
    for key, group in groupby(myAuthlog, key = lambda x: x[:9]):
        month, day, hour = key[0:3], key[4:6], key[7:9]
        print "%s:00 %s-%s: %d"%(hour, day, month, len(list(group)))

Answer 2

import collections
from datetime import datetime as dt

answer = collections.defaultdict(int)
with open('path/to/logfile') as infile:
    for line in infile:
        stamp = line[:9]
        t = dt.strptime(stamp, "%b\t%d\t%H")
        answer[t] += 1