請教一下，怎么將文件每行的時間進(jìn)行排序

_lpq

,12:13:57,a,b,c
12:14:11,12:14:18,a,e,a
,12:14:02,a,e,d
12:14:13,12:16:56,a,e,v
12:17:25,12:17:32,a,d,h
12:14:02,12:14:11,a,s,u
,12:14:14,a,z,u
12:14:29,12:14:36,a,g,d


求教各位大神！
第一個逗號前面為開始的時間，如果沒有則為空
第二個逗號為結(jié)束的時間，如果沒有則為空


現(xiàn)有兩個問題：
1）怎么先對開始的時間進(jìn)行排序，再對結(jié)束的時間進(jìn)行排序
2）假如格式是這樣的呢？怎么進(jìn)行排序
2016-01-14 12:14:39,2016-01-14 12:15:07,a,f,x
2016-01-14 12:13:52,2016-01-14 12:14:21,f,i,u
2016-01-14 12:13:22,2016-01-14 12:14:50,s,u,f
2016-01-14 12:14:52,2016-01-14 12:13:20,x,h,d
2016-01-14 12:13:12,2016-01-14 12:18:20,t,n,c
2016-01-14 12:14:11,2016-01-14 12:14:10,p,n,g

多謝各位大俠了~~

_lpq

然后就是修改完成的怎么輸出到原文件

cao627

1. sort -t',' file
   
2. sort -k2 -t',' file

復(fù)制代碼

_lpq

回復(fù) 3# cao627



謝謝~
能不能一行就搞定，然后重新輸出到原文件

   

cao627

1）怎么先對開始的時間進(jìn)行排序，再對結(jié)束的時間進(jìn)行排序

你這句話的意思是：
1：按開始時間這一列對行進(jìn)行排序得到一種結(jié)果，按結(jié)束時間這列對行進(jìn)行排序得到另一種結(jié)果，即用兩種不同的排序得到兩種結(jié)果？ 還是
2：先按開始時間排序，如果有開始時間相同的行，這些行再按慮結(jié)束時間排序。即用一種排序得到一種結(jié)果。

如果是1的話只能是兩條命令。
如果是2的話：

sort -k2 -k3  -t',' file

_lpq

回復(fù) 5# cao627


額。。。。。我寫了歧義了

我的意思是先對第一列經(jīng)行排序，然后相同的再第二列進(jìn)行排序（就是將相同的通過第二列，安插到第二列中）


試了你的貌似不成功

   

jason680

回復(fù) 6# _lpq

easy way by sort

$ cat FILE
2016-01-14 12:14:39,2016-01-14 12:15:07,a,f,x
2016-01-14 12:13:52,2016-01-14 12:14:21,f,i,u
2016-01-14 12:13:22,2016-01-14 12:14:50,s,u,f
2016-01-14 12:14:52,2016-01-14 12:13:20,x,h,d
2016-01-14 12:13:12,2016-01-14 12:18:20,t,n,c
2016-01-14 12:14:11,2016-01-14 12:14:10,p,n,g

$ sort FILE
2016-01-14 12:13:12,2016-01-14 12:18:20,t,n,c
2016-01-14 12:13:22,2016-01-14 12:14:50,s,u,f
2016-01-14 12:13:52,2016-01-14 12:14:21,f,i,u
2016-01-14 12:14:11,2016-01-14 12:14:10,p,n,g
2016-01-14 12:14:39,2016-01-14 12:15:07,a,f,x
2016-01-14 12:14:52,2016-01-14 12:13:20,x,h,d

   

_lpq

回復(fù) 7# jason680


    謝謝大神哈！，

_lpq

回復(fù) 7# jason680
我現(xiàn)在遇到的問題是：


,12:13:57,a,1,c
12:14:11,12:14:18,a,3,a
,12:14:02,a,6,d
12:14:13,12:16:56,a,4,v
12:15:31,,a,2,h
12:17:25,12:17:32,a,2,h
12:14:02,12:14:11,a,4,u
,12:14:14,a,9,u
12:14:13,,i,8,s
12:14:29,12:14:36,a,3,d

就是先對第二列排 用  sort -k2 -t',' file
運行結(jié)果：

12:15:31,,a,2,h
12:14:13,,i,8,s
,12:13:57,a,1,c
,12:14:02,a,6,d
12:14:02,12:14:11,a,4,u
,12:14:14,a,9,u
12:14:11,12:14:18,a,3,a
12:14:29,12:14:36,a,3,d
12:14:13,12:16:56,a,4,v
12:17:25,12:17:32,a,2,h

我希望得到的結(jié)果是：

,12:13:57,a,1,c
,12:14:02,a,6,d
12:14:02,12:14:11,a,4,u
12:14:13,,i,8,s
,12:14:14,a,9,u
12:14:11,12:14:18,a,3,a
12:14:29,12:14:36,a,3,d
12:15:31,,a,2,h
12:14:13,12:16:56,a,4,v
12:17:25,12:17:32,a,2,h

就是先對第二列排序，然后有第二列為空值是會運行到前面，
12:15:31,,a,2,h
12:14:13,,i,8,s

我希望的是將這兩列值與第一列值比較
若第一列有比   第一列值 <12:15:31 < 第一列值，則插入其中
像這樣
12:14:29,12:14:36,a,3,d
12:15:31,,a,2,h
12:14:13,12:16:56,a,4,v

謝謝哈~



   

jason680

回復(fù) 9# _lpq

here you are

$ awk -F, '{if($2=="")a[$1]=$0;else b[$2]=$0}END{bt=asorti(b,bs);at=asorti(a,as);for(an=bn=1;bn<=bt;++bn){while(as[an]<bs[bn]&&an<=at)print a[as[an++]];print b[bs[bn]]}}' FILE
,12:13:57,a,1,c
,12:14:02,a,6,d
12:14:02,12:14:11,a,4,u
12:14:13,,i,8,s
,12:14:14,a,9,u
12:14:11,12:14:18,a,3,a
12:14:29,12:14:36,a,3,d
12:15:31,,a,2,h
12:14:13,12:16:56,a,4,v
12:17:25,12:17:32,a,2,h