0

I am in need of a solution, preferably sed, that can match on an IP octet that contains any number combination and remove the line. For example:

My file:

 192.168.100.12
 192.168.200.12
 10.10.20.10
 10.10.30.20

I need to remove the line that contains any IP found on the 10.10.20.x subnet and remove it. So, if the file contains 10.10.20.10 , 20.11, 20.12, etc... They all get removed. I also need an inline replacement.

What I have tried:

sed -i '/10\.10\.20\.\([0-9]\{1,3\}\)/d' file

and this:

sed -i '/10.\.10\.20\.*\([0-9]\{1,3\}\)*/d' file​

And neither one works. The command completes with no errors, but the lines are still in the file. I am also open to other solutions such as awk.

1 Answer 1

Reset to default
2

With GNU sed:

sed -i '/10\.10\.20\.[[:digit:]]\{,3\}/d' file

With BSD sed, you need sed -i ''.

2
  • Perfect thanks! But why does [0-9] not work? I have seen it in other examples here and there?
    – user53029
    May 9, 2016 at 2:04
  • 1
    @user53029: sed -i '/10\.10\.20\.\([0-9]\{1,3\}\)/d' file should work, but [0-9] can match other than digit 1 to 9, base on your system and locale. [:digit:] guarantees to match anything considered digits in current locale.
    – cuonglm
    May 9, 2016 at 2:25

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .