SED or AWK

Mallardle · 08-30-2010, 06:12 AM

Hi,

I have a file that looks like this:

1
2 3 4 5 6 7 8 9
10 11 12 13 14 15
16 17 18 19 20 21
22 23 24
1
2 3 4 5 6 7 8 9
10 11 12 13 14 15
16 17 18 19 20 21
22 23 24
1
2 3 4 5 6 7 8 9
10 11 12 13 14 15
16 17 18 19 20 21
22 23 24
...

I would like to reformat it to look like this:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
...

Is there a nifty awk/sed one-liner to do this operation?

Cheers,
Nick

grail · 08-30-2010, 06:23 AM

Something like:

Code:

awk 'NR>1 && /^1$/{print ""}{printf $0" "}END{print ""}' file

Mallardle · 08-30-2010, 06:45 AM

Thanks very much Grail,

That worked perfectly in the example, but I have a follow up question. I have a file in a similar layout, but with different numbers and with irregular spaces between records and at the start of each line:

982.5760
982.5760 -982.5760 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
-999.2500 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
15.8611 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
-999.2500 -999.2500 -999.2500
982.6268
982.6268 -982.6268 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
-999.2500 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
15.6731 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
-999.2500 -999.2500 -999.2500
982.6776
982.6776 -982.6776 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
-999.2500 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
15.6087 -999.2500 -999.2500 -999.2500 -999.2500 -999.2500
-999.2500 -999.2500 -999.2500

Is there a way I can adapt your solution for this, and what does /^1$/ do in awk?

Many Thanks,
Nick

ghostdog74 · 08-30-2010, 06:53 AM

you can also do it with pure bash

Code:

#!/bin/bash
exec 4<"file"
while read -r a b <&4
do
  case "$b" in
   "") printf "$a ";;
   *) echo "$a $b";;
  esac
done
exec 4<&-

awk if you insist

Code:

awk 'NF==1{printf "%s ",$0;next}1' file

grail · 08-30-2010, 07:07 AM

Quote:

what does /^1$/ do in awk?

Based on original input this looked for the line that contains a '1' and nothing else.

You can use above bash or a combo of mine and awk above gets your results:

Code:

awk 'NR>1 && NF==1{print ""}{printf $0" "}END{print ""}' file

Mallardle · 08-30-2010, 07:21 AM

Thanks Grail and Ghostdog.

Both solutions worked a treat.

Nick

grail · 08-30-2010, 07:44 AM

Not sure if your requirements had changed but the bash doesn't quite work as is.
Slight change gets the same as my awk script:

Code:

#!/bin/bash

count=0
exec 4<"file"

while read -r a b 
do
  case "$b" in
   "") ((count++)) && out="\n$a" || out="$a";;
   *) out=" $a $b";;
  esac
	printf "$out"
done<&4
exec 4<&-

echo