CODEWITHSUNDEEP
bash
Mindy
Mindy

Reputation: 21

Using 'tr' to print repeated new lines to single new line?

I have a list of words which I have sorted using the sort command and then printed all the duplicates using uniq -D. The output file foo.txt is as follows:

always  
always  
bat  
bat  
boot  
boot  
boot  
hi  
hi

I was then attempting to use the following command in a script to read the file with the tr command.

It's not quite working, and the output looks the same as above.

This is what I have so far:

 cat foo.txt | tr -s '\n'

The output, however, appears the same. My goal is for the output to be:

always always  
bat bat  
boot boot boot  
hi hi

Upvotes: 0

Views: 602

Answers (2)

ctac_
ctac_

Reputation: 2471

You can separate each group by a newline with uniq.

After that sed take this newline as field separator.

uniq --all-repeated=separate infile | sed ':A;$bB;/\n$/!{N;bA};:B;s/\n/ /g'

Upvotes: 0

thanasisp
thanasisp

Reputation: 5965

another awk:

awk '{c=(($0!=p)?"\n":" "); printf c $0} {p=$0}' file
  • c is the character to print before any word.
  • p is the previous word.
  • if previous word is different, set c as newline or else as space.

it prints newline at the beginning and no newline at end, so we can add a few more:

awk '{c=(($0!=p)?"\n":" "); p=$0} NR==1{printf $0;next} {printf c $0}
     END {printf "\n"}' file

another solution, you could run uniq -c file instead of -D which prints

  2 always
  2 bat
  3 boot
  2 hi

and then print the word for the times the first field says:

uniq -c file | awk '{for (i=1;i<$1;i++) printf $2 " "; print $2}'

Upvotes: 1

Related Questions