SED Delete multiple sets of 2 Lines in stacked file

[Tim Wilk]

Looking for a way to delete 2 lines that start with 2 values (header and footers), for example, a merged/stacked file w/ 1000 lines, and randomly in the file there will be 2 lines, back to back, that start with 01 (header) and 99 (footer).

01 LINE (keep this header)
SOME LINE
SOME LINE
SOME LINE
99 LINE (delete this footer)
01 LINE (delete this header)
SOME LINE
SOME LINE
SOME LINE
SOME LINE
99 LINE (delete this footer)
01 LINE (delete this header)
SOME LINE
SOME LINE
99 LINE (delete this footer)
01 LINE (delete this header)
SOME LINE
99 LINE (keep this footer)

Results would have just 1 01 and 99 line.

[Steven E. Protter]

Shalom,

http://www.unix.com/shell-programming-scripting/24126-how-change-file-sed.html

You may need to take multiple passes on the file to do this.

SEP

[James R. Ferguson]

Hi Tim:

Using your actual, posted sample:

# perl -ne 'print if eof;print unless /^99 LINE/../^01 LINE/' file

Regards!

...JRF...

[Jdamian]

Try awk code (I assumed that the line "01 LINE" always is the next line a "09 LINE" but first and last lines) :

cat file | awk '
/^99 LINE/ {
OLDLINE=$0;
if (!getline) {
print OLDLINE
exit
}
}

{ print $0 }
'

[Dennis Handly]

Cleaning up Oscar's awk script and removing assumptions and cat:
awk '
BEGIN { getline; print $0 }
/01 LINE/ { next }
/99 LINE/ { save=$0; next }
{
print $0
}
END { print save } ' file