UNIX for Advanced & Expert Users

Thanks ripat and era.

ripat, your script will work fine but the problem is sometimes
I get only open or close within the begin and end script.

era,
Your script works beautifully. I just need one request if you dont mind.
begin and end is something I will add later. What if I dont want to add "end" and only "begin". Now I have to use the "begin" and the "begin" of the next section as an end.

Code:

begin
  open
    ......
  open
    ......
  close
    ......
begin
   .....
begin
   .....

When count=0 and begin is the next begin then display.
I want to check for the exact "begin" so
it will only look for "begin" and not "beginning" or "prebegin",
just like what you said.

Something like this but it does not work yet.

awk -F '|' '/"begin"/ { heading = $2 }
/open/ { count = 2; next }
count && /"begin"/ { if (! --count) { data = $0; print heading "|" data; heading = data = ""; count = 0 }' $1

era, I really appreciate your help. I wish you all the best!

Double quotes have no significance in regular expressions, you are looking for literal double quotes around the begin, which obviously won't be there.

However, you can check for equality: if field $1 is exactly equal to "begin" then it's a match. And if you have some data you have remembered from a previous section, print that and flush the state variables before starting over.

Code:

awk -F '|' '$1 == "begin" { if (data) { print heading "|" data; heading = data = ""; count = 0 } 
  heading = $2 }
/open/ { count = 2; next }
count{ if (! --count) data = $0; }
END { if (data) print heading "|" data }

The last line copes with the case when at end of file there is no new "begin" line but you have something you want to print from the previous "begin" section.

Just add a condition that awk needs to see two occurrences of "open" before it parses the record.

Code:

#!/usr/bin/awk -f

BEGIN {
        RS=""
        FS="open"
}

/open.+open/ {
        split($1, dte, "|")
        split($3, out, "\n")
        gsub(/\n| /, "", dte[2])
        sub(/ +/, "", out[3])
        print dte[2] "|" out[3]
}