How to remove all matching pattern except the top one in sed?

Question

file.txt

1.1.1.1 abc 
1.1.1.1 ccc
1.2.2.2 ddd
1.2.2.2 qqq
1.2.2.2 zzz

Expected result:

1.1.1.1 abc
1.2.2.2 ddd

I tried:

sed -i '/1.1.1.1/d' file.txt 

sed -i '/2.2.2.2/d' file.txt

But however it remove all instead of leaving 1 result. How to delete all matching pattern but left 1?

score 1 · Answer 1 · answered Aug 31 '19 at 06:16

1

You can use sort with the --unique option to achieve what you want:

$ sort --unique --key=1,1 file.txt
1.1.1.1 abc
1.2.2.2 ddd

or in short

sort -uk1,1 file.txt

The --unique option suppresses lines with equal fields that were already processed and --key=1,1 defines the first field as sort key (start and end position).

answered Aug 31 '19 at 06:16

Freddy

25,172
1
21
60

Thank you! Im looking for this for ages! – Muzaffar Mohamed Aug 31 '19 at 10:22
Is there anyway in sed? – Muzaffar Mohamed Aug 31 '19 at 11:39
This might be possible with `sed`, but it I guess it would be quite complicated. Better use `awk` or a another method from linked [answer](https://unix.stackexchange.com/a/160015). – Freddy Aug 31 '19 at 11:51
Thank you... Ill check it out! – Muzaffar Mohamed Aug 31 '19 at 14:12

How to remove all matching pattern except the top one in sed?

1 Answers1