Shell: replace \n with space between two specific lines - shell

Question

This is my first post here, thanks in advance for your prompt support.
I did some greps on a big file and came with file like the below. I want to make all lines after that starting with XX# in the same line.
So, I want the flowing:
XX#DEV1>
Feb 23 07:00:03
Feb 23 07:00:05
Sent : 4608
Received : 4227
Feb 23 07:00:07
Feb 23 07:00:09
XX#DEV2>
Feb 23 07:00:32
Feb 23 07:00:34
Sent : 4608
Received : 4232
Feb 23 07:00:36
Feb 23 07:00:38
XX#DEV1>
Feb 23 08:00:03
Feb 23 08:00:06
Sent : 4608
Received : 4265
Feb 23 08:00:07
Feb 23 08:00:09
XX#DEV2>
...
To become:
XX#DEV1> Feb 23 07:00:03 Feb 23 07:00:05 Sent : 4608 Received : 4227 Feb 23 07:00:07 Feb 23 07:00:09
XX#DEV2> Feb 23 07:00:32 Feb 23 07:00:34 Sent : 4608 Received : 4232 Feb 23 07:00:36 Feb 23 07:00:38
XX#DEV1> Feb 23 08:00:03 Feb 23 08:00:06 Sent : 4608 Received : 4265 Feb 23 08:00:07 Feb 23 08:00:09
XX#DEV2> ...

Answer 1

You can do that with awk pretty easily:
awk '/^XX/{if(length(out))print out;out=$0;next}{out=out" "$0}END{print out}' yourfile
That says.. if the line starts with "XX", print whatever we have accumluated in variable out, then save current line in variable out. If the line is anything else, append it to variable out after adding a space. At the end, print whatever we have accumulated in variable out.
Output:
XX#DEV1> Feb 23 07:00:03 Feb 23 07:00:05 Sent : 4608 Received : 4227 Feb 23 07:00:07 Feb 23 07:00:09
XX#DEV2> Feb 23 07:00:32 Feb 23 07:00:34 Sent : 4608 Received : 4232 Feb 23 07:00:36 Feb 23 07:00:38
XX#DEV1> Feb 23 08:00:03 Feb 23 08:00:06 Sent : 4608 Received : 4265 Feb 23 08:00:07 Feb 23 08:00:09

Answer 2

Or you can do it bash in much the same way as the awk in my other answer:
#!/bin/bash
while read line
do
if [[ $line == XX* ]] ;then
echo $OUT # Output whatever we have accumulated in $OUT
OUT=$line # Zero out $OUT and restart accumulating
else
OUT="$OUT $line" # Append this line to $OUT
fi
done < yourfile
echo $OUT # Output any trailing stuff at the end

Answer 3

sed -n '
# use -n to prevent sed from printing each line
# if the line read starts with XX, go (b) to :printit
/^XX/ b printit
# if it is the last line ($), also go to :printit
$ b printit
# for all other lines, append the read line to the hold space (H)
# and then go to the :end (so we read the next line)
H; b end
:printit {
# swap the hold and the pattern space
x
# replace all (s/../../g) single or multiple newlines (\n\n*)
# with a space
s/\n\n*/ /g
# print the pattern space
p
}
:end
' your_file