How to use sed to delete all lines before the first match of a pattern?

C_Blade · 04-14-2010, 07:26 PM

I have to use sed with this problem, and sed is extremely complicated, they could probably teach a whole class on this entire function.

Here's an example to show what I need to do:

Code:

aaaaaaaaaaaaacoolbbbbbbbbb
ccccccccoolddddcool
sweetcoolawesome
sweetgreat

how would I say that every line before I reach the first match of "sweet" should be deleted so that the output will look like:

Code:

sweetcoolawesome
sweetgreat

Tinkster · 04-14-2010, 08:02 PM

man sed
/print

geoff_f · 04-15-2010, 01:29 AM

This might be more useful than the man pages:

http://www.grymoire.com/Unix/Sed.html

Look under the section 'Ranges by patterns'.

r-t · 04-15-2010, 02:40 AM

sed -n '/sweet/,$p' file

http://www.computing.net/answers/uni...-one/3572.html

bakdong · 04-15-2010, 02:48 AM

If you put a little bit of explanation with your answer, assuming it's not just 'RTFM', which, let's face it, could answer 90% of posts here, we can all learn or refresh/reinforce our knowledge from it.

sed -n '/sweet/,$p' filename

-n don't print lines by default

/sweet/,$ is a restriction for the following command p, meaning 'only look from the first occurence of 'sweet' to '$' (the end of the file)

p print

pixellany · 04-15-2010, 07:58 AM

Guys;
With homework problems, we normally avoid simply handing over the answer.

Quote:

I have to use sed with this problem, and sed is extremely complicated

Au contraire!! SED is one of the simplest utilities in the toolbelt, with very limited action options. What you may be referring to is the fact that SED can be used to generate totally incomprehensible statements/scripts---many of which could be done better with some other tool........Watch for the periodic "SED vs. AWK" skirmishes here at LQ.

lbutler · 04-15-2010, 08:20 AM

Try,

sed -r -n -e '/PATTERN/,${p}' files

The flag -n means that lines that do not match are discarded;
-r means use extended regexps.

The lines from the first match of PATTERN to the end of file ($)
while be printed by the p command, otherwise the line is discarded
by default.

Leo

grail · 04-15-2010, 09:40 AM

Or we could delete as requested:

Code:

sed '/sweet/,$!d' file

Or just cause awk was mentioned

Code:

awk '/sweet/{i++}i' file

rkski · 05-01-2010, 01:06 AM

Hi grail

Could you please explain what's in bold. I'm trying to strengthen my awk skills.

Code:

awk '/sweet/{i++}i' file

TIA

grail · 05-01-2010, 04:18 AM

All variables in awk start with a value of 0, so until we first encounter sweet the 'i' at the end will be 0 so false therefore nothing to be done.
Once we find sweet and increase 'i' with '++', now 'i' is not zero (therefore true) and we do the default action which is to print the line