[SOLVED] bash to print words/lines horizontally

nitya · 02-20-2013, 04:45 AM

Hi All,
I have a test file with below entries

Code:

$ cat /tmp/test 
RedHat
Cent
Fedora
Suse
Ubuntu

Now how can I print these entries horizontally i.e I am expecting the output to be as

Code:

RedHat Cent Fedora Suse Ubuntu

Thanks in advance for your kind help.

druuna · 02-20-2013, 04:54 AM

Using tr:

Code:

 cat infile | tr "\n" " "

Using awk:

Code:

awk 'BEGIN{ ORS=" " }{ print }' infile

There are probably other ways as well.

millgates · 02-20-2013, 05:03 AM

pure bash:

Code:

foo="$(<infile)"
echo "${foo//$'\n'/ }"

or

Code:

<infile mapfile -t array
echo "${array[@]}"

Of course, there's also Perl, sed and many others.

nitya · 02-20-2013, 05:37 AM

Thank you both, worked as expected. But druuna both of your commands output are in the same line where will be "user@hostname:~$", it is okay I can cut and paste the output.
Once again thanks a lot druuna and millgates.

millgates · 02-20-2013, 06:36 AM

Quote:

Originally Posted by nitya

both of your commands output are in the same line where will be "user@hostname:~$"

That's because the first example just replaces all newlines with spaces, including the trailing one. So there's no newline at the end of output and the bash prompt ends up on the same line. The awk example reads the file line by line and outputs them delimited by spaces. Again, no newline at the end of output. My examples work in a similar way, but I use echo to print the result, which automatically appends a newline to whatever it prints (this can be overriden by the -n switch). If you want those trailing \ns to be there, you can print them yourself:

Code:

cat infile | tr "\n" " " && echo

and

Code:

awk 'BEGIN{ ORS=" " }{ print }END{ print "\n"}' infile

or

Code:

awk 'BEGIN{ ORS=" " }{ print }' infile && echo

But it depends on what you want to do with the output. Do you want to use it in a script? Put it in a file?

RaviTezu · 02-20-2013, 07:33 AM

Adding to druuna post.

Quote:

cat infile | tr "\n" " " | grep -v @

Will work. i have ignored "user@hostname:~$" using grep command. :P

grail · 02-20-2013, 09:58 AM

Code:

paste -sd' ' file

colucix · 02-20-2013, 10:10 AM

Code:

echo $(<file)

nitya · 02-21-2013, 02:58 AM

Thank you all for giving lot of ways to achieve this.

Code:

@millgates
foo="$(<infile)"
echo "${foo//$'\n'/ }"

<infile mapfile -t array
echo "${array[@]}"

cat infile | tr "\n" " " && echo

awk 'BEGIN{ ORS=" " }{ print }' /tmp/test && echo

@RaviTezu
cat infile | tr "\n" " " | grep -v @

@grail
paste -sd' ' file

@colucix
echo $(<file)

All above commands worked great. This thread helped me a lot to know about different commands.
Once again thank you all.

shivaa · 02-21-2013, 04:51 AM

One more way:

Code:

~$ for word in $(cat /tmp/test); do echo -n "$word "; done; echo

millgates · 02-21-2013, 05:11 AM

Quote:

Originally Posted by shivaa

One more way:

Code:

~$ for word in $(cat /tmp/test); do echo -n "$word "; done; echo

In this case it will work, but keep in mind that reading files with for like this is generally a bad idea.

Code:

~$ while read word; do echo -n "$word"; done < /tmp/test; echo

might be a bit better.

shivaa · 02-21-2013, 05:19 AM

Quote:

Originally Posted by millgates;

In this case it will work, but keep in mind that reading files with for like this is generally a bad idea.

Oh yes, you're right.. how can I forget David the H's lession on use while instead of for.

@Everyone:
How can I read a file (data stream, variable) line-by-line (and/or field-by-field)? (Link)
And
Why you don't read lines with "for" (Link)

whizje · 02-21-2013, 05:32 AM

Code:

xargs echo < file

whizje · 02-21-2013, 05:32 AM

Code:

xargs echo < file