getting sed to hold lines and print them out again- help please

inonzi_prowler · 02-20-2007, 05:58 AM

hello all.

so i have this text file, a section of which looks like this:

Quote:

KH6:1:1
4691814
4691814
4691814
4691814
4691814
KT3:1:1
4691815
4691815
4691815
4691816
4691816
4691816

i would very very much like for it to look like this:

Quote:

KH6:1:1
KH6:1:1 4691814
KH6:1:1 4691814
KH6:1:1 4691814
KH6:1:1 4691814
KH6:1:1 4691814
KT3:1:1
KT3:1:1 4691815
KT3:1:1 4691815
KT3:1:1 4691815
KT3:1:1 4691816
KT3:1:1 4691816
KT3:1:1 4691816

reading around i guess i need to get it to hold the line in the buffer when it finds a ":" (the common bit of text in the left column) and print at the start of each line until it finds another ":".

is that right guys? i am very new to sed and am struggling.

many thanks in advance for any help.

colucix · 02-20-2007, 06:43 AM

Here is a simple awk solution:

Code:

$0 ~ /:/ { header = $0 ;
           print header
}
$0 !~ /:/ { print header, $0
}

I guess the first line of the input file contains a field like KH6:1:1, otherwise we have to parse the first lines of the input file in a different way.

Nick_Battle · 02-20-2007, 06:47 AM

You can do this with sed, but it is very unintuitive. I would recommend awk:

/^K/ {
prefix=$0
print prefix
}
/^[0-9]/ {
print prefix, $0
}

Put that in a file xxx and run with "awk -f xxx yourdata"

HTH

inonzi_prowler · 02-21-2007, 03:26 AM

thanks very much guys.