awk怎么把匹配到的行放到其他行尾

moperyblue

1. sed '/^add/{:a;N;/\nadd/!{s/\n/ /;ta};P;D}'
   
2. awk -vRS='add' 'NF{print gensub("\n"," ","g",RS $0)}'

复制代码

_R__

awk 'NF==2{getline $3}1' file

ll104567

1. root@lp:~/pwb/tmp/awk# cat file|awk '{if(NF==2&&NR%2==1){ORS=" "}else{ORS="\n"}}1'
   
2. add zha by-ser 005
   
3. add wei by-ser 006
   
4. add xia by-ser 007
   
5. add lin by-ser 008
   
6. add qin by-ser 009
   

复制代码

1. awk '{if(NF!=4&&NR%2==1){ORS=" "}else{ORS="\n"}}1' file

复制代码

jason680

$ awk '{ORS=/by-/?"\n":FS}1' FILE
add zha by-ser 005
add wei by-ser 006
add xia by-ser 007
add lin by-ser 008
add qin by-ser009

hz_oracle

cat file | xargs -n4

moperyblue

1. sed 's/\S*/&/3;t;N;s/\n/ /'

复制代码

tonyibm

awk -vRS=add 'NF{gsub("\n"," ");print RS $0}'

这条命令中的NF作为pattern是起什么作用，我试着将NF去掉，结果是这样的，

add
add zha by-ser 005
add wei by-ser 006
add xia by-ser 007
add lin by-ser 008
add qin by-ser 009

多出了第一行add, 没太明白有NF和没有NF的区别，

reyleon

1. sed ':1;N;/\nadd/!s/\n/ /;t1;P;D' file

复制代码