inplace deletion of a line before and after a pattern using sed command in linux

Question

forward-zone:
name: "."
forward-addr: 10.1.1.2

forward-zone:
name: "mydomain.com"
forward-addr: 10.1.1.1

I wanted to delete one line before the pattern 'name: "."' and one line after the pattern (including the pattern itself). I tried the following command but this is not modifying the file itself. I want to do inplace modification. Please help

sed -n '/name: "."/{N;s/.*//;x;d;};x;p;${x;p;}' file

Answer 1

Something like this might suit.

sed -rne '/^name: "\."/{n;n;h;d};1{h;d};x;p;${g;p}' -i file

The -i option lets Sed modify the file in place. The Sed commands h, g and x use Sed's hold space to hold the previous line until the current line has been examined. The addresses 1 and $ respectively treat the file's first and last lines, which probably want special handling.

Though this has been an educational exercise, admittedly, Awk or even Perl might be a better tool than Sed for such multiline processing. Nevertheless, with some extra effort, Sed can do the job, as you see.