CODEWITHSUNDEEP
bashsedgrep
user3254640
user3254640

Reputation: 39

getting a part of the output from a sed command

I have this Command :

cat -n file.log | grep "Start new test" | tail -1 |  cut -f 1 |  xargs -I % sed -n %',$s/is not alive/&/p' file.log

That gives the output of the whole line :

Jan 19 23:20:33 s_localhost@file platMgt.xbin[3260]: blade 10 is not alive
Jan 19 23:20:33 s_localhost@file platMgt.xbin[3260]: blade 11 is not alive

how can I modify it to get the last part only : blade 11 is not alive

can I modify that in a way to display :

Error:blade 11 is not alive ?

Thank you for your response

Upvotes: 0

Views: 84

Answers (3)

anubhava
anubhava

Reputation: 784888

To get the last part after colon awk is better tool:

s='Jan 19 23:20:33 s_localhost@file platMgt.xbin[3260]: blade 10 is not alive'
awk -F':' '{print "Error:" $NF}' <<< "$s"

OUTPUT:

 blade 10 is not alive

EDIT: WIth your piped commands you can combine it as:

grep "Start new test" file.log|tail -1|awk -F':' '{print "Error:" $NF}'

PS: Though this whole thing is possible in awk itself.

Upvotes: 1

Kevin
Kevin

Reputation: 2182

You can use cut to delimit it on the colons and then add the error message:

cat -n file.log | grep "Start new test" | tail -1 |  cut -f 1 |  xargs -I % sed -n %',$s/is not alive/&/p' file.log | cut -d: -f 4 | xargs -I % echo Error: %

Upvotes: 2

gskielian
gskielian

Reputation: 184

The following obtains the the last ":" separated field with a sed command,

cat text.txt | sed 's/^.*: \([^:]*$\)/\1/g'

Upvotes: 0

Related Questions