SED replace after ignoring

Question

There are many lines in a text file and I would like to change the value true to false using a one line script.

bash-3.2$ grep online test.txt
# online = false --> ONE SHOT import (delete all table contents before importing)
# online = true  --> DIFFERENTIAL import (import without clean the table contents)
online =  true

I tried this

sed -i.bak '/DIFFERENTIAL\|/ONE SHOT\/!s/online.*/online = false/g'

but it gave this error:

sed: -e expression #1, char 17: unknown command: `O'

I don't want the first 2 lines to change, only the third online = true to become online = false. It can be anywhere in the file but it will match online =. So I would want my command to ignore 1 and 2 and change 3 only.

Answer 1

If you are only looking to alter lines starting with online, it's perhaps easiest to do this:

sed 's/^online\ *=.*/online = false/g'

This way, the lines starting with # are completely ignored.

An even simpler (but not identical) alternative, suggested by @mikeserv:

sed '/^online/s/true/false/'

Answer 2

Via awk:

awk '/online =/{c++} c==3{gsub("true","false")}1' test.txt

This will match all occurrences of online = and increment the counter flag c++; When 3 occurrences found c==3, then replace the true with false.