output two commands on same line

Question

I have seen this question a few times, but the solutions I cannot get to work.

I have the following command:

printf '%s\n' "${fa[@]}" | xargs -n 3 bash -c 'cat *-$2.ss | sed -n 11,1p ; echo $0 $1 $2;'

where

printf '%s\n' "${fa[@]}"

O00238 115 03
O00238 126 04

and cat *-$2.ss gives:

   1 D C   0.999  0.000  0.000
   2 L C   0.940  0.034  0.012
   3 H C   0.971  0.005  0.015
   4 P C   0.977  0.005  0.009
   5 T C   0.970  0.009  0.018
   6 L C   0.977  0.006  0.011
   7 P C   0.864  0.027  0.014
   8 P C   0.966  0.018  0.011
   9 L C   0.920  0.038  0.039
  10 K C   0.924  0.043  0.039
  11 D C   0.935  0.036  0.035
  12 R C   0.934  0.023  0.053
  13 D C   0.932  0.022  0.046
  14 F C   0.878  0.041  0.088
  15 V C   0.805  0.031  0.198
  16 D C   0.834  0.039  0.108
  17 G C   0.882  0.019  0.071
  18 P C   0.800  0.031  0.132
  19 I C   0.893  0.039  0.070
  20 H C   0.823  0.024  0.179
  21 H C   0.920  0.026  0.070
  22 R C   0.996  0.001  0.002

running the command then produces

  11 D C   0.935  0.036  0.035
O00238 115 03
  11 K C   0.449  0.252  0.270
O00238 126 04

Even lines are the output of sed -n 11,1p, odd lines the output of echo $0 $1 $2

How do I pair the output on the same line i.e.

  11 D C   0.935  0.036  0.035 O00238 115 03
  11 K C   0.449  0.252  0.270 O00238 126 04

I have tried:

printf '%s\n' "${fa[@]}" | xargs -n 3 bash -c 'cat *-$2.ss | {sed -n 11,1p ; echo $0 $1 $2;} | tr "\n" " "'

as suggested here: Concatenate in bash the output of two commands without newline character

however I get

O00238: -c: line 0: syntax error near unexpected token `}'
O00238: -c: line 0: `cat *-$2.ss | {sed -n 11,1p ; echo $0 $1 $2;} | tr "\n" " "'

What is the problem?

What is in the original array? Perhaps it'd be best to start from there, rather than picking through the pipeline you already have. — Tom Fenech, Apr 21 '15 at 08:08
Instead of people trying to bolt something onto the end of your command,if you provide us with the original input and expected output then i think this can all be done in a single command. — , Apr 21 '15 at 08:09
thanks both - have added output of `printf '%s\n' "${fa[@]}"` — brucezepplin, Apr 21 '15 at 08:12
@brucezepplin we would also need to know what is in all the files in `*-$2.ss` (or an example of what would be in there) — , Apr 21 '15 at 08:13
@ Tom - each call to cat finds a different file. JID asked for an example - the one I give refers to the first line i.e. `11 D C 0.935 0.036 0.035 O00238 115 03` — brucezepplin, Apr 21 '15 at 08:23
this could all be done in one simple, clear awk command but I see you've already selected an answer that adds more complexity to your pipeline so if you'd like to see how to do it all at once post a followup question. — Ed Morton, Apr 21 '15 at 13:33
hi Ed - I'm open to any solution really. It's only that I was already trying to do it in bash only and I can only choose one answer to accept. It may be helpful for others as well to see an awk solution — brucezepplin, Apr 21 '15 at 13:44

score 2 · Answer 1 · answered Apr 21 '15 at 08:52

2

You could try using something like this:

i=0
for f in *-"$2".ss; do printf '%s %s\n' "$(sed -n '11p' "$f")" "${fa[$((i++))]}"; done

This loops through your files and prints the 11th line alongside a slice from the array fa, whose index i increases by 1 every iteration.

answered Apr 21 '15 at 08:52

Tom Fenech

72,334
12
107
141

@brucezepplin It's also shorter and more efficient! – Tom Fenech Apr 21 '15 at 09:28

martin · Accepted Answer · 2015-04-21T08:40:27.443

1

I could not reproduce your setup, but

printf "O00238 115 03\nO00238 126 04" | xargs -n 3 bash -c 'cat test.dat | sed -n 11,1p | tr -d "\n"; echo " $0 $1 $2"'

gives

11 D C   0.935  0.036  0.035 O00238 115 03
11 D C   0.935  0.036  0.035 O00238 126 04

which should work in your case. I just deleted the newline of the sed command.

edited Apr 21 '15 at 08:40

answered Apr 21 '15 at 08:31

martin

3,149
1
24
35

output two commands on same line

2 Answers2