I have a command to print out the top 5 most cpu intensive commands:
ps aux --sort=-%cpu --no-headers
I would like to make a printout containing only the parsed file name.
So, I want to turn this:
$ ps aux --sort=-%cpu --no-headers
user 24883 4.3 1.7 2010436 131440 ? Sl 17:07 0:06 /usr/lib/x86_64-linux-gnu/qt5/libexec/QtWebEnginePr
user 25475 1.9 2.4 2048140 182668 ? Sl 16:25 0:52 /usr/lib/x86_64-linux-gnu/qt5/libexec/QtWebEnginePr
user 20794 1.7 3.6 3393352 277388 ? Sl 16:22 0:49 falkon
user 541 0.3 0.5 433640 38748 tty1 Sl 10:45 1:18 /usr/lib/xorg/Xorg -nolisten tcp :0 vt1 -keeptty -a
user 18867 0.2 0.0 4396 1020 ? S 15:40 0:11 /bin/sh /home/user/scripts/./status.sh
root 8 0.1 0.0 0 0 ? I 10:45 0:23 [rcu_sched]
root 1 0.0 0.1 216728 8760 ? Ss 10:45 0:07 /sbin/init
root 2 0.0 0.0 0 0 ? S 10:45 0:00 [kthreadd]
root 4 0.0 0.0 0 0 ? I< 10:45 0:00 [kworker/0:0H]
root 6 0.0 0.0 0 0 ? I< 10:45 0:00 [mm_percpu_wq]
root 7 0.0 0.0 0 0 ? S 10:45 0:00 [ksoftirqd/0]
root 9 0.0 0.0 0 0 ? I 10:45 0:00 [rcu_bh]
root 10 0.0 0.0 0 0 ? S 10:45 0:00 [migration/0]
root 11 0.0 0.0 0 0 ? S 10:45 0:00 [watchdog/0]
root 12 0.0 0.0 0 0 ? S 10:45 0:00 [cpuhp/0]
...
Into this:
QtWebEnginePr
QtWebEnginePr
falkon
Xorg
sh
Note that '/usr/lib/xorg/Xorg -nolisten tcp :0 vt1 -keeptty -a' has been trimmed to say only 'Xorg'; I would like the file name itself with no command parameters at the end.
So far I have this:
$ ps aux --sort=-%cpu --no-headers | awk 'NR==1,NR==5{print $11}'
/usr/lib/x86_64-linux-gnu/qt5/libexec/QtWebEngineProcess
falkon
/usr/lib/x86_64-linux-gnu/qt5/libexec/QtWebEngineProcess
/usr/lib/xorg/Xorg
/bin/sh
I understand that I could just slap a cut command on the end of it and it's all fine:
ex.,
ps aux --sort=-%cpu --no-headers | awk 'NR==1,NR==5{print $11}' | cut -d "/" -f2)
but I don't want to use anymore commands to complete this task; this command will be repeated every 1 second or so and printed out in my status bar, so I want it to be efficient. And I know that awk is powerful enough to do this; but I just don't seem to be able to do this all in one. I could parse the beginning of the filename but the end out be out of control. And vise-versa. I have a feeling it's something to do with '-F/' but how to incorporate the two?
Any ideas?
I'm using sh shell and running Debian 9.
I'm guessing that this is far more efficient than using awk.