-1

I have a really bad habit of abusing tr. I need to find another way, a different style. all I want to so is print the list horizontally instead of vertically - so I can cut and past it into an email. Check out the use of the TR command. Just terrible. 

$ cat /tmp/wig
update PTMM_ARCHIVE.FASTTRACK_USER set user_name = 'monohajoxx' where user_name = 'monohajo'
update PTMM_ARCHIVE.FASTTRACK_USER set user_name = 'wuemxx' where user_name = 'wuem'
update PTMM_ARCHIVE.FASTTRACK_USER set user_name = 'taraziemxx' where user_name = 'taraziem'
update PTMM_ARCHIVE.FASTTRACK_USER set user_name = 'mullankexx' where user_name = 'mullanke'
update PTMM_ARCHIVE.FASTTRACK_USER set user_name = 'fernanjaxx' where user_name = 'fernanja'
$ awk '{print $NF}' /tmp/wig | tr -d "'" | tr "\n" ", \s" ; echo  "\n"
monohajo,wuem,taraziem,mullanke,fernanja,\n
4ae1e1
  • 7,228
  • 8
  • 44
  • 77
capser
  • 2,442
  • 5
  • 42
  • 74
  • I don't see any questions here. If you're just looking for feedback, consider [Code Review](http://codereview.stackexchange.com/). – Jonathon Reinhart Dec 02 '15 at 03:02
  • Thank you - code review - I will check that out. - you must agree that I am abusing the tr command. I mean there has to be a better way to get the list printed out that way by using a tool besides tr. – capser Dec 02 '15 at 03:04
  • 1
    I honestly don't see anything wrong with it. That's a perfectly fine one-liner. – Jonathon Reinhart Dec 02 '15 at 03:05
  • Did you really want that last comma after `fernanja`? Typical comma-separated-value formats don't include that. – John1024 Dec 02 '15 at 03:49

1 Answers1

1

Using awk

Here is one way to do it entirely with awk:

$ awk '{gsub(/'\''/,"",$NF); printf "%s%s",(NR>1?",":""),$NF} END{print "\\n"}' wig
monohajo,wuem,taraziem,mullanke,fernanja\n

The gsub command removes the single-quotes from the last field. The printf command prints the last field preceded by a comma if this isn't the first line. The final print statement finishes the line.

And, here is another:

$ awk '{printf "%s%s",(NR>1?",":""),substr($NF,2,length($NF)-2)} END{print "\\n"}' wig
monohajo,wuem,taraziem,mullanke,fernanja\n

This uses a similar printf statement but uses substr to remove the first and last characters of the last field.

Using sed

$ sed -nE "s/.*'([^']*)'/\1/"'; H; 1h; ${x; s/\n/,/g; s/$/\\n/; p}' wig
monohajo,wuem,taraziem,mullanke,fernanja\n

How it works:

  • -n tells sed not to print anything unless we explicitly ask it to.

  • -E tells sed to use extended regular expressions so that we don't have to type as many backslashes.

  • s/.*'([^']*)'/\1/

    This removes everything from the line except for the last field single-quoted string (with the quotes are removed).

  • H; 1h;

    H adds a newline to the hold space followed by a copy of the current pattern space (which now contains the last field, minus the quotes).

    If this is the first line, however, the h command overwrites the hold space with just the current value of the pattern space (no newline).

  • ${x; s/\n/,/g; s/$/\\n/; p}

    On the last line, denoted by $, this does the following:

    - `x` exchanges the hold and pattern spaces.

- `s/\n/,/g` converts all those newlines to commas.

- `s/$/\\n/` puts a `\n` at the end.

- `p` causes this pattern space to be printed.
John1024
  • 109,961
  • 14
  • 137
  • 171