怎么过滤重复的id呢?通过shell的grep

purelove_88

我又来了{:2_169:}

求帮忙.....

[2014-07-10 11:56:51] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1855","CHANNEL_CODE":"downjoy","ROLE_LEVEL":"3","USER_ID":"CU20140710115305930","ISRESIGN":"false","ARAM":"1","RECORD_TIME":"1404964611414"}}
[2014-07-10 11:56:53] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1809","CHANNEL_CODE":"pp","ROLE_LEVEL":"10","USER_ID":"CU20140710115028691","ISRESIGN":"false","ARAM":"1","RECORD_TIME":"1404964613716"}}
[2014-07-10 11:56:54] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1748","CHANNEL_CODE":"uc","ROLE_LEVEL":"12","USER_ID":"CU20140710114433186","ISRESIGN":"false","ARAM":"1","RECORD_TIME":"1404964614082"}}
[2014-07-10 11:57:20] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1850","CHANNEL_CODE":"uc","ROLE_LEVEL":"12","USER_ID":"CU20140604025033375","ISRESIGN":"false","ARAM":"1","RECORD_TIME":"1404964640320"}}
[2014-07-10 11:57:30] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1824","CHANNEL_CODE":"uc","ROLE_LEVEL":"5","USER_ID":"CU20140708114459219","ISRESIGN":"false","ARAM":"1","RECORD_TIME":"1404964650856"}}
[2014-07-10 11:57:31] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1892","CHANNEL_CODE":"downjoy","ROLE_LEVEL":"4","USER_ID":"CU20140129113214983","ISRESIGN":"false","ARAM":"1","RECORD_TIME":"1404964651880"}}
[2014-07-10 11:57:33] {"log_type":"SIGNIN","log_data":{"SERVER_ID":"11","ROLE_ID":"1904","CHANNEL_CODE":"uc",


我想在以上日志中  通过不同的role_id进行计算登陆数量, 如果相同, 则不累计

有方法么....


{:2_166:} 是真的不会,想跟各位多学习{:2_169:}  多谢多谢

reyleon

看到各种表情 ,心中为之躁动..

yestreenstars

请禁用表情！

yestreenstars

目测一下：

1. grep -oP '(?<="ROLE_ID":)[^,]+' i | sort -u | wc -l

复制代码

purelove_88

回复 4# yestreenstars


    亲....这个东西我应该怎么设置参数.....我不是很懂.....原来没接触过代码.  能详细的讲一下么{:3_187:}

zerostudy

1. awk -F "," '{a=gensub(/[^0-9]*([0-9]+)"/,"\\1","g",$3);b[a]++}END{print length(b)}' file

复制代码

练习下，用awk

reyleon

回复 5# purelove_88


    你发帖的时候,右侧上部有一个禁用表情的沟,你把这个沟勾上

purelove_88

回复 7# reyleon


    好的  谢谢大哥提醒

q1208c

1. sort -t \" log -k14,14 -u | wc -l

复制代码

yestreenstars

回复 5# purelove_88

正则如果看不懂的话搜索一下零宽断言~

grep
-o 只打印正则匹配的内容
-P 使用Perl正则

sort
-u 去重

wc
-l 打印行数