[SOLVED] grep filter within line?

qrange · 10-10-2011, 08:22 AM

after using grep, I have the following line:

Code:

:Temperature 49

how to filter it out some more, so that only '49' is left?
thanks.

colucix · 10-10-2011, 08:34 AM

You can either pipe the output to another grep to match exactly the number at the end of the line and using the -o option to print out only the matching part:

Code:

grep Temperature file | grep -E -o [0-9]+$

To avoid the double grep command, you can use awk (or even sed) to match the Temperature pattern and extract only the information you need:

Code:

awk '/:Temperature/{print $NF}' file

MTK358 · 10-10-2011, 11:44 AM

Or use sed:

Code:

sed 's/^:Temperature \([0-9]\+\)$/\1/'

This will eliminate the need for the first grep command, too.

rng · 10-10-2011, 12:14 PM

You can also use 'cut' command:

$ echo ":temperature 49" | cut -d" " -f2
49

qrange · 10-11-2011, 02:32 AM

thanks all.

er, there are actually two lines with :Temperature, and I'd like to merge both numbers into one line.
is there a way to combine these two awk into single awk:

Code:

 awk '/:Temperature/{print $2}' | awk '{ printf "%s ", $0 }'

(I plan to use it in cron script and wish to reduce resource usage as much as possible)
or some better way?

colucix · 10-11-2011, 02:34 AM

Quote:

Originally Posted by qrange

er, there are actually two lines with :Temperature, and I'd like to merge both numbers into one line.

Of course. Please show us the two lines and the desired result.

qrange · 10-11-2011, 02:43 AM

nevermind, I just found out how to do it:
awk '/:Temperature/{printf "%s ", $2}'

the two lines are these:
:Temperature 48
:Temperature 25

and output should be:
48 25

thanks.

colucix · 10-11-2011, 02:51 AM

That's it. You may want to add a terminating newline:

Code:

awk '/:Temperature/{printf "%d ", $NF} END{printf "\n"}' file