UNIX for Dummies Questions & Answers

thx for the reply. but this is not working. I had tried this earlier and it doesn't give any output at all maybe due the volume of records ? not sure....when I size down the file to only two to three records the same sed command works.

anyone got an idea on the above ?

Try the following:
1) Copy the file to a temp file.
2) Keep dividing it in half until you can run the "sed" command successfuly.
3) From this successful file, keep adding X amount of records until it does not work.

Let us know the results:
a) What is the file size (bytes) where the command breaks.
b) How many lines it has.

Are you giving sufficient time for the "sed" command to work?

Does it give any error message at all?

The problem is that your file is made of 1 very long line (at least, that is the file i have downloaded)
maybe give a try to

Code:

awk 'gsub("<[/]*FEEDMessage>","\n&\n",$0)' File.xml | sed -n '/<FEEDMessage>/,/<\/FEEDMessage>/p'

Use nawk instead of awk if you are on SunOS / Solaris

If you want it then collapsed into one long line you can still pipe it into something like

Code:

| paste s -d "\0" -

so something like

Code:

awk 'gsub("<[/]*FEEDMessage>","\n&\n",$0)' File.xml | sed -n '/<FEEDMessage>/,/<\/FEEDMessage>/p' | paste -s -d "\0" - >new.xml

thanks ctsgnb! from this I got a clue and wrote below command which worked for me

nawk -F"<" ' { for(i=18;i<=NF-1;i++){ print "<"$i } }' File.xml > temp.xml
tr -d '\n' < temp.xml > File.xml