How to convert rows to columns in unix

Question

I want my Unix file output, which has each value output on a new line, to be converted into grouped rows as shown below.

Say my output file in Unix looks like this:

jobname
userid
starttime
endtime
jobname2
userid
starttime
endtime

I want the output to be:

jobname1 userid starttime endtime
jobname2 userid starttime endtime

Are you trying to run a bash script? Or Python? Or C++? or ?? — fooOnYou, Aug 13 '12 at 20:55
What you describe in the question is that you have a column and you want to convert it to several rows, that is, "convert columns to rows" **not** "convert rows to columns" — Stephen P, Aug 14 '12 at 00:36

kev · Answer 1 · 2012-08-14T00:23:28.300

10

This is a minimal awk solution:

awk 'ORS=NR%4?" ":"\n"' input.txt

jobname userid starttime endtime
jobname2 userid starttime endtime

(If you want to align the fields, pipe to column -t)

edited Aug 14 '12 at 00:23

answered Aug 14 '12 at 00:14

kev

score 6 · Answer 2 · answered Sep 14 '15 at 08:18

6

This will paste each four consecutive lines as four tab-delimited fields:

cat source_file | paste - - - - > destination_file

answered Sep 14 '15 at 08:18

Julian TF

score 1 · Answer 3 · answered Aug 13 '12 at 20:56

If you are looking for a shell script, you can do this as the number of lines to be printed in the output seems to have to fixed length:

while read line1; do
  read line2
  read line3
  read line4
  echo $line1 $line2 $line3 $line4 >>output
done <file

score 1 · Answer 4 · answered Aug 13 '12 at 22:29

1

With GNU sed:

tr '\n' ' ' < file | sed -r 's/(\w+ +){4}/&\n/g'

answered Aug 13 '12 at 22:29

Thor

score 0 · Answer 5 · answered Aug 13 '12 at 21:07

0

Kind of ugly, but doesn't require a shell script:

cat outputfile | tr "\n" " " | sed -e 's/\W*\(\w*\) \(\w*\) \(\w*\) \(\w*\)\W*/\1 \2 \3 \4\n/g'

answered Aug 13 '12 at 21:07

Jon Lin

5 Answers5