converting lines into one-CodePudding

could you please help to merge line in input file and produce output as below in descending order.

inputfile.txt

TerminalA/admin#
   51% used 
TerminalB/admin#
   62% used
TerminalC/admin#
   42% used

outputfile should contain as below in one line each terminal.

TerminalB/admin# 62% used
TerminalA/admin# 51% used
TerminalC/admin# 42% used

Below is code what I tried:

awk -f process1.awk | sort -t: -k1,2rn > output.txt

where awk script is:

cat process1.awk /^$/ { print " " } !/^$/ { printf("%s ",$0) }

CodePudding user response：

With your shown samples please try following awk code.

awk 'FNR%2==0{print val OFS $0 | "sort -nrk2";next} {val=$0}' Input_file

OR: In case you want to remove extra spaces before % line, which I think should be removed as per shown samples then try following:

awk '
FNR%2==0{
  sub(/^  /,"")
  print val OFS $0
  next
}
{
  val=$0
}
' Input_file | sort -nrk2

Explanation:

Checking condition FNR%2==0 to see if line number is even then do following.
Printing variable val value followed by OFS followed by $0(current line).
using | "sort -nrk2" to further is basically sorting output in reverse/increasing order of 2nd column.
next will skip all further statements from here.
Statement {val=$0} will store current line's value into variable val. This statement will Only be executed when its an odd line number.

CodePudding user response：

as a perl oneliner:

perl -0777 -pE 's/\n\s / /g' /tmp/xx2 | sort -rk2

First we tell Perl to read the complete file (slurp) instead of reading it line by line. It might not work with files which do not fit into RAM. Then we use regex to convert linebreaks and spaces into single space and reverse sorting by the second field with sort -rk2