Question

I have a output which is not very well formatted and that causes the trouble in capturing variable value.

For example:

This is counter1 1000
this counter2 2000
this counter3 is higher value 3000

Now from above file I want to just capture 1000, 2000 and 3000. If I use awk with print $x where x = any value then it allocates the different values for each line.
If I use $4 then for first line I will be able to select 1000 but for 2nd line it will not provide any value and for third line it will print higher so I want to insert the field separator and read the variable. Any help is greatly appreciated!

Answer 1

If you only want last argument, and only if a number:

echo $'For example:\nThis is counter1 1000\nthis counter2 2000\n
     this counter3 is higher value 3000\ndone.\n' |
  sed -ne 's/^.* \([0-9]\{1,99\}\)/\1/p'

1000
2000
3000

You could even trap the counter’s number (and so read only line containing counterX):

echo $'For example:\nThis is counter1 1000\nthis counter2 2000\n
     this counter3 is higher value 3000\ndone.\n' |
  sed -ne 's/^.*\(counter[0-9]\{1,99\}\) \(.* \)\{0,1\}\([0-9]\{1,99\}\)$/\1 \3/p'

counter1 1000
counter2 2000
counter3 3000

Or even for just locating separator:

echo $'For example:\nThis is counter1 1000\nthis counter2 2000\n
     this counter3 is higher value 3000\ndone.\n' |
  sed -ne 's/^\(.*counter[0-9]\{1,99\}.*\) \([0-9]\{1,99\}\)$/\1 :: \2/p'
This is counter1 :: 1000
this counter2 :: 2000
this counter3 is higher value :: 3000

or, or…

echo $'For example:\nThis is counter1 1000\nthis counter2 2000\n
     this counter3 is higher value 3000\ndone.\n' |
  sed -e 's/^\(.*counter[0-9]\{1,99\}.*\) \([0-9]\{1,99\}\)$/\1 :: \2/'

For example:
This is counter1 :: 1000
this counter2 :: 2000
this counter3 is higher value :: 3000
done.