Delete lines based on Rules

Shell Programming and Scripting
1

Hi My requirement is very simple .

I juts need to delte some lines from a file.

here comes theactual scenario

I have some data in file like say

 
srinivasa prabhu kumar antony 
srinivas king prabhu antony 
srinivas prabhu king yar venkata 
venkata kingson srinivas prabhu 
prabhu srinivas prabhu kingson

I want to delete the lines which contain both the words srinivas and prabhu.

please help

2

You can do this using sed with two nested ranges:

sed '/srinivas/ {
                  /prabhu/d
                }' /path/to/input

I hope this helps.

bakunin

3

Hi i tried this and is not workig ...means the lines are not getting deleted.

4

Are you getting the output correctly or is that the file remains unchanged? To edit the file inline, you need to use sed -i ...

--ahamed

5

In your above example, every single line should be deleted, meaning you would have no output at all.

Try also

$ grep -Ev "srinivas.*prabhu|prabhu.*srinivas" file
6 
awk '!(/srinivas/ && /prabhu/)' file

I am not sure if line #1 has a typo srinivasa prabhu kumar antony
If you only like to remove line containing exact words like srinivas and not srinivasa , use this

awk '!(/\<srinivas\>/ && /\<prabhu\>/)' file
7

Note that the output of sed always goes to <stdout>, which in most cases is the terminal. Try it, watch the output and if you are satisfied redirect it to some file and move this file over the original file:

sed '<commands>' /path/to/input > path/to/output
mv path/to/output /path/to/input

Yes and no. You should not use this feature, even if it is there, which is solely(!) with the GNU-variants of sed .

I hope this helps.

bakunin