[SOLVED] splitting a coloum in row

hoi · 06-25-2013, 07:33 AM

I have data file like

1
2
3
4
5
6
7
8
9
10

I want to write like

1 2
3 4
5 6
7 8
9 10

How do I make it by 'awk command?

Madhu Desai · 06-25-2013, 08:03 AM

Code:

$ awk '{printf("%s",NR%2 ? $0" ":$0"\n")}' file

Or simply you can use

Code:

$ paste - - < file

## To put space instead of tab
$ paste -d' ' - - < file

grail · 06-25-2013, 09:33 AM

Code:

awk 'ORS = NR % 2?" ":"\n"' file

linuxCode · 06-25-2013, 04:06 PM

Code:

cat file | xargs -n2

Sorry, my mistake. you wanted awk

hoi · 06-26-2013, 12:23 AM

Thanks to all
really all answer are helpful for my work
again thanks to all

hoi · 06-26-2013, 12:29 AM

Quote:

Originally Posted by linuxCode

Code:

cat file | xargs -n2

Sorry, my mistake. you wanted awk

Thanks
xargs, it was not known to me
Now I am using xargs
thank you

hoi · 06-26-2013, 12:31 AM

Quote:

Originally Posted by mddesai

Code:

$ awk '{printf("%s",NR%2 ? $0" ":$0"\n")}' file

Or simply you can use

Code:

$ paste - - < file

## To put space instead of tab
$ paste -d' ' - - < file

Now I have understand that paste is better than awk in this case
Thank you

linuxCode · 06-26-2013, 12:47 AM

Quote:

Originally Posted by hoi

Thanks
xargs, it was not known to me
Now I am using xargs
thank you

You're welcome

BTW, the command I gave you doesn't modify your original file. Your original file will be the same. If you want the new changes redirect the output to another file. For example

Code:

cat file | xargs -n2 > newfile

Welcome to LQ!

David the H. · 06-28-2013, 05:22 AM

I must say I like the xargs solution. Clever. I'd skip cat though and use a shell redirection.

Code:

xargs -n2 <infile.txt

However, here's yet another option for you:

Code:

sed 'N;s/\n/ /' infile.txt

You can concatenate 3 or more lines together with additional N commands and the g flag to the substitution. And the -i option (assuming gnu sed) can be used to return the edits directly to the input file.

Code:

sed -i 'N;N;s/\n/ /g' infile.txt

Finally, here's a quick bash-only solution that will work as long as the input isn't too large and you're using v.4+. It should be very fast on small files.

Code:

mapfile -t nums <infile.txt
printf '%s %s\n' "${nums[@]}"