sed command in multicharacter

Shell Programming and Scripting
1

Hello,

I have some problem about sed command. I have data which showed in below :

M001    ndle1
M002|bfn|n|bfn|bf       ndle1
M003|bfn|n|bf|bf        ndle1
M004|bfn|n|bf   middle1
M005|bfn|n|bfn|bf       middle1
M006|bnf|n|bfn|bf       middle1
M007|fn|q|n|bf  middle1

The expected output :

M001        ndle1
M002        ndle1
M003        ndle1
M004        middle1
M005        middle1
M006        middle1
M007        middle1

It looks like easy, but I can't solve in completely because I have a-z with | symbol. Normally, I used this code :

sed -e "s/|bfn*//g" -e "s/|n//g" -e "s/|fn//g" -e "s/|q//g" -e "s/|bnf//g" inputfile

Could you please suggest me?
Thanks in advance.

2

If the first field length is fixed (4 char long alphanumeric with uppercase), one way could be

sed 's/^\([A-Z0-9]\{4\}\).*  *\(.*\)/\1 \2/' file

If you are not sure about the length, the awk version might come handy

awk -F "[ |]" '{print $1,$NF}' file
1 Like
3

Try using the inverse of the spacing characters:

sed 's/|[^[:blank:]]*//' file
2 Likes
4

Thank you very much. for your suggestion. The first field length is not fixed, they vary in other length. I tried to use this code, but it do not give me in expected output.

---------- Post updated at 03:26 PM ---------- Previous update was at 03:24 PM ----------

Thank you very much for your suggestion. I tried to use this code, I can get the expected output.