Sets of lines delimited by # to single lines

Hello gurus, wizards; magicians, beginners,

how to solve this with sed (preferred) or awk

SOURCE (ASCII file)
=================
Line01
Line02
Line03
#
Line03
Line01
#
Line01
Line04
Line02
Line03
#
Line02
#

RESULT (ASCII file)
===================
Line01 Line02 Line03
Line03 Line01
Line01 Line04 Line02 Line03
Line02

Thanks for any response, Paul

Try this out:

BEGIN { RS = "#\n" }

{
n = split($0, a, "\n")
for (i = 1; i < n; i++)
printf "%s ", a[i]
print a[i]
}

sed would be OT, so let's not go there.

BEGIN {RS="#"}
$1=$1

How simple, problem solved

thanks to all

Paul

"Kenny McCormack" <gazelle@xmission.xmission.com> schrieb im Newsbeitrag
news:ec6o40$e6s$1@news.xmission.com...

i wrote a sed one ,not good but it works

sed '/^$/d;:t;N;/#/!bt;s/#\|\n\n*/ /g'

regards

{
if ($1 != "#" )
{
printf ("%s ",$1);
}
else
{
printf ("\n");
}
}

hth

Error! Trailing space.

"data" is your datafile

/usr/bin/awk -f prog.awk data
-bash-3.00$ cat prog.awk
{
if ($1 != "#" )
{
printf ("%s ",$1);
}
else
{
printf ("\n");
}
}

hth

^^^

Your program prints a space at the end of each line.
Not desirable in most sets of data.

In fact, all proposed solutions so far have some gotcha. besides the
leading/trailing space problems, it's also problematic with the record
rebuilding when there are TABs or adjacent whitespaces in a record.
here are some other solutions:

awk -v RS="\n#" '{ gsub(/\n/," "); gsub(/^ | $/,"") }1'

or

{
if ($0 ~ /^#/) {
SP = "\n";
} else {
printf("%s%s", SP, $0);
SP = " ";
}
}

or use two s/// commands in sed/perl:

sed -n 'H;$ba;/^#/{:a;x;s/\n/ /g;s/^#\? \| #\?$//g;p;}'
perl -lp043e 's/\n/ /g;s/^ | $//g'

Regards,
Xicheng

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode test 33Switch to Hybrid Mode Switch to Threaded Mode

1 19th October 07:00 paul bullack External User Posts: 1	Sets of lines delimited by # to single lines Hello gurus, wizards; magicians, beginners, how to solve this with sed (preferred) or awk SOURCE (ASCII file) ================= Line01 Line02 Line03 # Line03 Line01 # Line01 Line04 Line02 Line03 # Line02 # RESULT (ASCII file) =================== Line01 Line02 Line03 Line03 Line01 Line01 Line04 Line02 Line03 Line02 Thanks for any response, Paul

2 19th October 07:00 vassilis External User Posts: 1	Sets of lines delimited by # to single lines Try this out: BEGIN { RS = "#\n" } { n = split($0, a, "\n") for (i = 1; i < n; i++) printf "%s ", a[i] print a[i] }

3 19th October 07:00 gazelle External User Posts: 1	Sets of lines delimited by # to single lines sed would be OT, so let's not go there. BEGIN {RS="#"} $1=$1

4 19th October 07:00 paul bullack External User Posts: 1	Sets of lines delimited by # to single lines How simple, problem solved thanks to all Paul "Kenny McCormack" <gazelle@xmission.xmission.com> schrieb im Newsbeitrag news:ec6o40$e6s$1@news.xmission.com...

5 19th October 07:00 joojle External User Posts: 1	Sets of lines delimited by # to single lines i wrote a sed one ,not good but it works sed '/^$/d;:t;N;/#/!bt;s/#\\|\n\n*/ /g' regards

6 19th October 07:00 mag gam External User Posts: 1	Sets of lines delimited by # to single lines { if ($1 != "#" ) { printf ("%s ",$1); } else { printf ("\n"); } } hth

7 19th October 07:00 gazelle External User Posts: 1	Sets of lines delimited by # to single lines Error! Trailing space.

8 19th October 07:00 mag gam External User Posts: 1	Sets of lines delimited by # to single lines "data" is your datafile /usr/bin/awk -f prog.awk data -bash-3.00$ cat prog.awk { if ($1 != "#" ) { printf ("%s ",$1); } else { printf ("\n"); } } hth

9 19th October 07:00 jon labadie External User Posts: 1	Sets of lines delimited by # to single lines ^^^ Your program prints a space at the end of each line. Not desirable in most sets of data.