请教多文件按文件名关键字分组的

jacson007

场景

当前目录下多个文件

#ls -l *.txt|awk '{print$9}'
tmp_aaa_mo_201505010500_00001.txt
tmp_aaa_mo_201505010530_00001.txt
tmp_aaa_mo_201505020500_00001.txt
tmp_aaa_mo_201505020530_00001.txt
tmp_aaa_mo_201505020530_00002.txt
tmp_aaa_mo_201505030500_00001.txt
tmp_aaa_mo_201505030530_00001.txt
tmp_aaa_mo_201505030530_00002.txt
tmp_aaa_mo_201505030530_00003.txt
tmp_aabb_mo_201505010500_00001.txt
tmp_aabb_mo_201505010530_00001.txt
tmp_aabb_mo_201505010545_00001.txt
tmp_aabb_mo_201505010545_00002.txt
tmp_aabb_mo_201505020500_00001.txt
tmp_aabb_mo_201505020515_00001.txt
tmp_aabb_mo_201505020530_00001.txt
tmp_aabb_mo_201505020530_00002.txt
tmp_aabb_mo_201505030500_00001.txt
tmp_aabb_mo_201505030530_00001.txt
tmp_aabb_mo_201505030530_00002.txt
tmp_aabb_mo_201505030550_00001.txt
tmp_aabb_mo_201505030550_00002.txt

要求输出
关键字 日期  文件个数
aaa  201501 3
aaa  201502 3
aaa  201503 4
aabb 201501 4
aabb 201502 4
aabb 201503 5

即按照 关键字 aaa 和日期(天)分组 统计文件个数 如果能加上文件大小总和就更好了。
关键字 日期  文件个数  总大小MB

songyc_2015

回复 1# jacson007

1. ls -l *.txt|awk '{split($9,a,/_/);b[a[2]" "substr(a[4],1,8)]++}END{for(s in b)print s,b[s]}'

复制代码

jason680

回复 1# jacson007

$ ls -l *.txt | awk 'BEGIN{print "关键字 日期  文件个数 总大小MB"}{split($9,a,"_");k=a[2]" "substr(a[4],1,8);c[k]++;s[k]+=$5}END{t=asorti(c,b);for(n=1;n<=t;n++)printf("%s %s %.1f\n",b[n],c[b[n]],s[b[n]]/(1024*1024))}' | column -t
关键字  日期      文件个数  总大小MB
aaa     20150501  2         0.0
aaa     20150502  3         0.0
aaa     20150503  4         0.0
aabb    20150501  4         0.0
aabb    20150502  4         0.0
aabb    20150503  5         0.0

   

jacson007

两位高手，我明天试试
我可能漏了一点
关键字可以有多个，而且长度不定
比如还有
tmp_bcs_mo_201505030530_00002.txt
tmp_xyz_mo_201505030550_00001.txt
tmp_mn_mo_201505030550_00002.txt

jacson007

回复 2# songyc_2015


    结果是一个总数，没有分组

jacson007

回复 3# jason680


  完美实现，多谢了

songyc_2015

回复 5# jacson007

ls -l *.txt文件名在第几个列，就把split($9,a,/_/)中的$9改为几，试试看