Question

在下面我试图grep所有geoIs模式。我的问题是如何列出geoIs的不同值及其计数

例如：

GeoIs:"Paramount","sumthing else"
GeoIs:"undefined","sumthing else"
GeoIs:"undefined","sumthing else"
GeoIs:"178","sumthing else"
GeoIs:"178","sumthing else"
and many more
...
...

预期结果：

GeoIs:"Paramount" 1
GeoIs:"undefined" 2
GeoIs:"178" 2

命令

zcat file.gz | grep -P '"geoIs":".*?.undefined*?"' | sort -u -T.|wc -l

EDIT1：

在字符串

中找到GEOIS

  012-10-02 09:32:45{"e":{"ec":100001,"st":1349170352455,"bd":"Mozilla%2F5.0%20(Windows%20NT%206.1)%20AppleWebKit%2F537.4%20(KHTMf01f02008592~rt%2366.657~rv%2366.228~as%2317~st%231349170293955~cat%231349170352431~sp%23as~c%2334~pat%231349128562942","smplCookie":"undefined","geoIPAddress":"122.107.154.58","geoCountry":"australia","geoCity":"Vermont","geoRegion":"Victoria","geoPostalCode":"undefined","geoLatitude":"undefined","geoLongitude":"undefined","geoMetro":"0","geoArea":"0","geoIs"}}

Answer 1

要返回频率表，请使用

sort | uniq -c | sort -n

对于您提供的示例数据，我会使用

zcat file.gz | cut -f1 -d, | sort | uniq -c | sort -n

zcat file.gz | grep -o '"searchstring":"[^"]*"'| sort | uniq -c | sort -n

Answer 2

awk替代方案：

awk -F, '{a[$1]++;}END{for(x in a)if(x)print x,a[x]}' file


kent$  echo 'GeoIsp:"Paramount","sumthing else"
GeoIsp:"undefined","sumthing else"
GeoIsp:"undefined","sumthing else"
GeoIsp:"178","sumthing else"
GeoIsp:"178","sumthing else"
'|awk -F, '{a[$1]++;}END{for(x in a)if(x)print x,a[x]}'
GeoIsp:"Paramount" 1
GeoIsp:"undefined" 2
GeoIsp:"178" 2

linux搜索模式并打印其计数

2 个答案: