URL请求数-MRJob-Python数据分析(6)

更新：2022-05-23 13:17:29编辑：Story归类：数据库人气：170

1.1. 前言

我们计算URL请求数经过这四个步骤:

Mapper: 将以行数据解析成 key=请求的URL数 value=1的形式

Shuffle: 通过Shuffle后的结果会生成以 key 的值排序的 value迭代器

结果如: 请求的URL数 [1, 1, 1 ... 1, 1]

Reduce 1: 在这边我们计算出每个请求的URL 的访问量

输出如: None [sum([1, 1, 1 ... 1, 1]), key]

Reduce 2: 对sum([1, 1, 1 ... 1, 1]) 进行排序并输出 TOP 100

输入如: 246361 "/wp-admin/admin-ajax.php"

1.2. 代码

cat mr_url_req.py
# -*- coding: utf-8 -*-
 
from mrjob.job import MRJob
from mrjob.step import MRStep
from ng_line_parser import NgLineParser
 
import heapq
 
class MRUrlRef(MRJob):
 
    ng_line_parser = NgLineParser()
 
    def mapper(self, _, line):
        self.ng_line_parser.parse(line)
        yield self.ng_line_parser.reference_url, 1 # 外链域名
 
    def reducer_sum(self, key, values):
        """统计 VU"""
        yield None, [sum(values), key]
 
    def reducer_top100(self, _, values):
        """访问数降序"""
        for cnt, path in heapq.nlargest(100, values):
            yield cnt, path
 
    def steps(self):
        return [
            MRStep(mapper=self.mapper,
                   reducer=self.reducer_sum),
            MRStep(reducer=self.reducer_top100)
        ]
 
def main():
    MRUrlRef.run()
 
if __name__ == '__main__':
    main()

cat mr_url_req . py

# -*- coding: utf-8 -*-

from mrjob . job import MRJob

from mrjob . step import MRStep

from ng_line_parser import NgLineParser

import heapq

class MRUrlRef ( MRJob ) :

ng_line_parser = NgLineParser ( )

def mapper ( self , _ , line ) :

self . ng_line_parser . parse ( line )

yield self . ng_line_parser . reference_url , 1 # 外链域名

def reducer_sum ( self , key , values ) :

"" "统计 VU" ""

yield None , [ sum ( values ) , key ]

def reducer_top100 ( self , _ , values ) :

"" "访问数降序" ""

for cnt , path in heapq . nlargest ( 100 , values ) :

yield cnt , path

def steps ( self ) :

return [

MRStep ( mapper = self . mapper ,

reducer = self . reducer_sum ) ,

MRStep ( reducer = self . reducer_top100 )

]

def main ( ) :

MRUrlRef . run ( )

if __name__ == '__main__' :

main ( )

运行统计和输出结果

<a href="http://www.ttlsa.com/python/" title="Python"target="_blank">python</a> mr_url_req.py < www.ttmark.com.access.log
 
No configs found; falling back on auto-configuration
Creating temp directory /tmp/mr_url_req.root.20160924.133027.483660
Running step 1 of 2...
reading from STDIN
Streaming final output from /tmp/mr_url_req.root.20160924.133027.483660/output...
246361  "/wp-admin/admin-ajax.php"
126012  "/tag/"
57325   "/"
......
778     "/meirong/2016/03/15/8442.html"
776     "/jiaju/2015/05/30/6058.html"
773     "/jiaju/2015/05/15/5747.html"
Removing temp directory /tmp/mr_url_req.root.20160924.133027.483660...

python mr_url_req . py < www . ttmark . com . access . log

No configs found ; falling back on auto - configuration

Creating temp directory / tmp / mr_url_req . root . 20160924.133027.483660

Running step 1 of 2...

reading from STDIN

Streaming final output from / tmp / mr_url_req . root . 20160924.133027.483660 / output . . .

246361 "/wp-admin/admin-ajax.php"

126012 "/tag/"

57325 "/"

. . . . . .

778 "/meirong/2016/03/15/8442.html"

776 "/jiaju/2015/05/30/6058.html"

773 "/jiaju/2015/05/15/5747.html"

Removing temp directory / tmp / mr_url_req . root . 20160924.133027.483660...

值得一提的是我们在这边使用了 heapq.nlargest 函数来计算 TOP 100 的URL请求数，或许多的程序员估计会使用 sorted([1, 1, 1 ... 1, 1])[:100] 这中形式来做。个人建议在数据量大的时候千万别这么高这样内存会被吃尽并报OOM错误。主要是因为 sorted 先生成了一个list 再对list进行切片。如果list数据量大就有问题。

昵称: HH

QQ: 275258836

ttlsa群交流沟通(QQ群②: 6690706 QQ群③: 168085569 QQ群④: 415230207（新）微信公众号: ttlsacom)

感觉本文内容不错，读后有收获？

逛逛衣服店，鼓励作者写出更好文章。

URL请求数-Pandas-Python数据分析(7) 时刻PV-Pandas-Python数据分析(5)