怎样删除某列中重复的字符串？ 谢谢

ly5066113

呵呵，考虑还是不周全。加个单词边界吧。

sed -r ':a;s/(\<[0-9]+(\.[0-9]+){3}\>)(.*),\1/\1\3/;ta'

BlueBlack

echo "1 2 3 4 4 3 1 1" | sed -e "s: :\n:g" | sort | uniq | xargs echo -n

ly5066113

体会一下，不难理解的。

ywlscpl

1. #awk -F "," '{for (i=1;i<=NF;i++) a[$i]=$i;for (j in a) if (a[j]!="") printf a[j]",";printf "\n";for (i=1;i<=NF;i++) a[$i]=""}' urfile
   
2. 10.1.2.3,172.23.2.34,10.2.3.4,
   
3. 172.23.12.35,10.1.2.4,10.2.3.78,
   
4. 10.2.1.65,10.7.8.79,
   

复制代码

顺序乱了，而且最后多了个,

BlueBlack

172.16.19.46,172.16.19.47
172.16.19.21,172.16.19.21,172.16.19.21,172.26.19.21,172.26.19.21,172.26.19.21
172.16.19.69,172.16.19.69,172.16.19.69,172.26.19.69,172.26.19.69,172.26.19.69
172.16.19.70,172.16.19.70,172.16.19.70,172.26.19.70,172.26.19.70,172.26.19.70

这个还是有点问题，还是有重复的

172.16.19.46,172.16.19.47
172.16.19.21,172.26.19.21
172.16.19.69,172.26.19.69
172.16.19.70,172.26.19.70

jackymail

sed 's/[,\ ]/\n/g' ufile | uniq

ynchnluiti

原帖由 BlueBlack 于 2009-1-6 16:02 发表
172.16.19.46,172.16.19.47
172.16.19.21,172.26.19.21
172.16.19.69,172.26.19.69
172.16.19.70,172.26.19.70

哪个是重复的？

我是DBA

我也没看到重复的。

ywlscpl

少打了个f

1. 208# cat o1
   
2. 172.16.19.46
   
3. 172.16.19.21,172.16.19.21,172.16.19.21,172.26.19.21,172.26.19.21,172.26.19.21
   
4. 172.16.19.69,172.16.19.69,172.16.19.69,172.26.19.69,172.26.19.69,172.26.19.69
   
5. 172.16.19.70,172.16.19.70,172.16.19.70,172.26.19.70,172.26.19.70,172.26.19.70
   
6. awk -F "," '{printf $1",";for (i=2;i<=NF;i++) {a=i-1;if ($i!=$a) printf $i","}printf "\n"}' o1
   
7. 172.16.19.46,
   
8. 172.16.19.21,172.26.19.21,
   
9. 172.16.19.69,172.26.19.69,
   
10. 172.16.19.70,172.26.19.70,
    

复制代码

基本满足要求，只要去掉最后的,

[ 本帖最后由 ywlscpl 于 2009-1-6 16:19 编辑 ]