[SOLVED] Trim operation using awk

luvshines · 10-10-2011, 03:37 PM

I wrote a couple of scripts some time back and remember using following code to trim the extra spaces

Code:

echo "single = word" | awk -F'=' '{gsub(/^[[:blank:]]+/,"",$2) gsub(/[[:blank:]]+$/,"",$2); print $2}'

I think it used to work fine but it's not working now.

Any suggestions, why [[:blank:]] thing is not working but [ \t] is ?

PS: In case it matters, I wrote this when I used to work on Fedora. Have been using Ubuntu for the past one year now

ntubski · 10-10-2011, 03:47 PM

You probably have mawk (minimal awk) installed on Ubuntu, and gawk on Fedora. mawk doesn't support [[:character class:]].

grail · 10-10-2011, 08:20 PM

Definitely the first piece of software to be changed on any Ubuntu installation (IMHO).

May I also add that the following would achieve your solution easier:

Code:

echo "single = word" | awk '{print $NF}'

By default gawk uses whitespace as the delimiter and so would remove all from each item on the line.

syg00 · 10-10-2011, 08:36 PM

The things you learn (mawk) - shows how little I use Ubuntu ...

luvshines · 10-11-2011, 04:35 AM

Quote:

Originally Posted by ntubski

You probably have mawk (minimal awk) installed on Ubuntu, and gawk on Fedora. mawk doesn't support [[:character class:]].

Yups !! mawk was playing naughty

All set now. Thanx

Just thinking loud - Had I not added that 'PS' note just before posting this, we could have gone on a totally different 'Pursuit of AWKwardness'

grail · 10-11-2011, 08:08 AM

Not really ... tends to be a question I ask as soon as I see Ubuntu and awk issues.

luvshines · 10-11-2011, 03:17 PM

Quote:

Originally Posted by grail

Not really ... tends to be a question I ask as soon as I see Ubuntu and awk issues.

And that is exactly what I was missing from my post when I was posting it. Just before pressing the 'submit' button, it struck that the script were originally written Fedora and I then added the PS note