remove lines matching more than one pattern using sed

Question

I'm trying to remove a few lines matching some regex.

curl <url> | sed '/\(foo\|bar\|baz\)/d'

i don't want any of those lines to show that match foo, bar or baz

it stops on foo

if this is easier with awk, i'm ok with that.

Do you mean "i don't want any of those lines to show that match foo, bar _and_ baz"? — potong, Feb 20 '14 at 09:50

score 1 · Accepted Answer · answered Feb 20 '14 at 08:17

1

Or with egrep:

curl <url> | egrep -v "foo|bar|baz"

answered Feb 20 '14 at 08:17

Mark Setchell

epeleg · Answer 2 · 2014-02-20T08:12:32.620

0

try

curl <url> | sed '/foo\|bar\|baz/d'

edited Feb 20 '14 at 08:12

answered Feb 20 '14 at 08:03

epeleg

it actually works as expect. the docs say that it stops at first matching regex. i need to make it greedy, but /g did not work. – chovy Feb 20 '14 at 08:04
you might find some of the things here http://stackoverflow.com/questions/5410757/sed-delete-a-line-containing-a-specific-string helpfull. – epeleg Feb 20 '14 at 08:09

score 0 · Answer 3 · answered Feb 20 '14 at 08:04

0

Using awk

curl <url> | awk '!/foo|bar|baz/'

answered Feb 20 '14 at 08:04

Jotne

score 0 · Answer 4 · answered Feb 20 '14 at 09:57

0

This might work for you (GNU sed):

sed '/foo/{/bar/{/baz/d}}' file

or:

sed '/foo/!b;/bar/!b;/baz/d' file

answered Feb 20 '14 at 09:57

potong

4 Answers4