scripting help with sed

n0gnus · 09-26-2008, 05:33 PM

Hi, I am not good with sed, and I'm having trouble trying to print only the characters between matching expressions. For example, if I have a file with many lines of this general format:

xxx.xxx.xxx.xxx User [joeschmoe] Group [somegroup] connected, Session Type

I only want to print the characters after User and before Group. Basically I want to get a user listing of all entries in the file. How would I go about this?

From what I've read this is supposed to work, but it fails terribly:

sed -n '/User/,/Group/p' users > temp

What would be even better would be to get the info after User and before the second bracket ']'.

Any help is greatly appreciated.

pixellany · 09-26-2008, 06:30 PM

welcome to LQ!!

The stucture you used is called an "addressing range". It says to print lines, beginning with one containing "User", and ending with one containing "Group".

What you need is something like this:

sed -n 's/.*User\(.*\)Group.*/\1/p'

Running to an appointment---can't do more now.

Excellent SED tutorial here:
http://www.grymoire.com/Unix/Sed.html

keefaz · 09-26-2008, 06:33 PM

I would use:

Code:

sed -e 's/.*User.*\[\(.*\)\].*Group.*/\1/' users > temp

Basically remove everything, just keep the value between \( and \)

Another alternative with cut:

Code:

cut -d[ -f2 users | cut -d] -f1 > temp

ghostdog74 · 09-26-2008, 07:27 PM

if you have Python,

Code:

for lines in open("file"):
    lines=lines.split() #get them into array
    #get the index where "User" is and add 1.
    # this will be position of the username
    print lines[lines.index("User") + 1]

n0gnus · 09-30-2008, 11:01 AM

Thanks for the help and the welcome! I really appreciate it; I'll attempt these today.