From: Joe Van Dyk <joevandyk@...>
Date: 2005-11-17T14:18:15+09:00
Subject: Re: Improving code...

On 11/16/05, Gregory Brown <gregory.t.brown@gmail.com> wrote:
> On 11/16/05, Joe Van Dyk <joevandyk@gmail.com> wrote:
>
> >   # See 'man proc' for details.
> >   def get_process_stats
> >     proc_stats = File.read("/proc/#{ @pid }/stat")
> >
> >     # EEWWWW -- gets the stuff from /proc/<pid>/stat and puts it into vars
> >     pid, comm, state, ppid, pgrp, session, tty_nr, tpgid, flags,
> >       minflt, cminflt, majflt, cmajflt, utime, stime, cutime,
> >       cstime, priority, nice, who_cares, itrealvade,
> >       starttime, vsize, rss, rlim, startcode, endcode, startstack,
> >       kstkesp, kstkeip, signal, blocked, sigignore, sigcatch,
> >       wchan, nswap, cnswap, exit_signal, processor =
> >     proc_stats.scanf("%d %s %c %d %d %d %d %d %d %d \
> > %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d \
> > %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d %d")
>
> proc_stats = File.read("/proc/#{ @pid }/stat").split(" ")
>
> you might need to change your split to match the appropriate pattern,
> matching a space above.  This makes proc_stats an array
>
> >     # Just put the ones I need into the hash
> >     stats = {}
> >     stats[:system_cpu_time] = stime
> >     stats[:user_cpu_time] = utime
> >     stats[:processor] = processor
> >     stats
> >   end
> > end
>
> stats = { :system_cpu_time => proc_stats[14],
>                :user_cpu_time     => proc_stats[13],
>                :processor            => proc_stats.last }
>
> the indicies might not be correct above, but hopefully the idea works ;)


That is *exactly* what I had before I decided to use scanf.  scanf has
the advantage of automatically converting everything to the correct
type.  But I'm not sure what's best in this case.

Later on, I might need to get more of the process information, so I
thought getting all the values, and then just putting the ones I need
into the hash would be better.

Which version would people rather work with?