Grep or Sed ? How can I use it ?

ArOnaXx · 04-10-2009, 10:50 AM

Hello.
First, my english is not very good, sorry, but try understand me please.

So...

I have one question, and I can't resolve it. I used google but I had no response for my question.

I created one file: "file.txt", and put it there:

status:active
color:blue
mode:normal

and now I want to get everything after the character ":"

example:

cat file.txt | grep -everything_after ":"

but I don't know how I do, if I use grep or if I use sed.

Please help me, thanks.

pixellany · 04-10-2009, 11:10 AM

sed 's/[^:]*:/:/1' filename

This won't do you any good unless you understand how it works.....Take a look here:
http://www.grymoire.com/Unix/Sed.html

And got to http://tldp.org and get the Bash Guide for Beginners

Robhogg · 04-10-2009, 11:14 AM

grep would not be the easiest tool to use for this - you should look at cut or awk instead:

Code:

cut -d: -f2 file.txt # output field number 2, fields separated by colon
awk -F: '{print $2}' file.txt

Here, I would choose cut (awk is more flexible, but more complex). You could do it in grep or sed, of course:

Code:

grep -o '[^:]*$' file.txt
sed 's/\(.*\)\(:\)\(.*\)/\3/'

... but these are not quite so readable

PTrenholme · 04-10-2009, 11:40 AM

Some other solutions:

gawk 'BEGIN {FS=":";} (print $2;}' file.txt
cut -d: -f2 file.txt

But grep only prints matched lines, not parts of them, so it's not appropriate for your stated problem.

However, if you wanted to print the string after the colon only when the line contains a colon, something like this could work:
grep ":" file.txt | cut -d: -f2

Note that all of these solutions assume that there is no more than one ":" in any line.

Robhogg · 04-10-2009, 11:55 AM

Quote:

Originally Posted by PTrenholme

But grep only prints matched lines, not parts of them, so it's not appropriate for your stated problem.

Depends on the version of grep - GNU grep (at least in recent versions) has an "-o" flag that only prints the part of the line matched by the expression, a feature lacking in the version on a Solaris machine that I work with.

custangro · 04-10-2009, 12:06 PM

Quote:

Originally Posted by ArOnaXx

Hello.
First, my english is not very good, sorry, but try understand me please.

So...

I have one question, and I can't resolve it. I used google but I had no response for my question.

I created one file: "file.txt", and put it there:

status:active
color:blue
mode:normal

and now I want to get everything after the character ":"

example:

cat file.txt | grep -everything_after ":"

but I don't know how I do, if I use grep or if I use sed.

Please help me, thanks.

If you only have those entries you can do...

Code:

awk -F':' '{print $2}'

ArOnaXx · 04-12-2009, 02:08 PM

GOT IT!
Thanks for all
I really appreciate