Need help using grep, need to grep something from a txt file.

[qsub]

Hi,

Basically I have a file which displays a users UID/GID/Username.

I want to basically take out the username and disgard the rest.

So lets pretend first line in a output.txt file is

uid=500(Bobby) gid=500(Bobby)
uid=501(Jim) gid=500(Jim)

I want to to basically read the first line in the text file and get the name Bobby to output. disgard the rest.

Can anyone help me with this? Preferably with the command grep in as short amount of characters as possible, just to keep it less complicated. This should be fairly simple id imagine.

Thanks,

[notageek]

awk ' {print $1 }' filename | sed 's/uid=[0-9][0-9][0-9](//'|sed 's/)//'

This probably can be cleaned up.

[weitjong]

The grep command is not suitable for this task as it outputs the matching lines. To cut the line and output its part I usually use the cut command.

$ head -1 output.txt | cut -f2 -d'(' | cut -f1 -d')'

[PabloTwo]

With the grep command, you have to know the pattern you are searching for. So, if you already know you are searching for the name Bobby in a file named grepfile.....

BASH:~/-> cat grepfile
uid=500(Bobby) gid=500(Bobby)
uid=501(Jim) gid=500(Jim)

BASH:~/-> grep -o Bobby grepfile
Bobby
Bobby

To grep for names Bobby and Jim

BASH:~/-> grep -o -e Bobby -e Jim grepfile
Bobby
Bobby
Jim
Jim

If you don't know the names you're searching for, then grep isn't the command you want to use.

[brunson]

I can't tell from the original question if you only want to print the username from the first line, or all lines.

If it's the former:

ebrunsonlx(~)$ sed -ne '1{s/.*(\([^)]*\)).*/\1/p}' grepfile 
Bobby

if the latter:

ebrunsonlx(~)$ sed -ne 's/.*(\([^)]*\)).*/\1/p' grepfile 
Bobby
Jim

[weitjong]

PabloTwo just gives me an idea of using grep to solve the problem. For me, grep is still not suitable for this task but it is not impossible.

$ grep -m 1 -o -P '(?<=\().*(?=\)\sgid)' output.txt

I am sure I will scratch my head to explain how it works myself after a while

Better alternative:

$ awk 'BEGIN {FS="[()]"}{print $2;exit}' output.txt

[PabloTwo]

BASH:~/-> grep -m 1 -o -P '(?<=\().*(?=\)\sgid)' grepfile
Bobby

Having read most of the Bash Programming book that I got for xmas from my wife, I actually understand most of that line

---------- Post added at 05:28 PM CST ---------- Previous post was at 10:44 AM CST ----------

I played around with this using only the printf external command, letting bash do all the rest.

BASH:~/-> read var < grepfile && var="${var##*(}" && printf "%s\n" ${var%%)*}
Bobby