topic sort question in Operating System - HP-UX

sort question

machoq1 — Wed, 07 Sep 2005 10:37:46 GMT

I have a file with multiple node names in it. I need to know how many times does a node name appear in a file
I am looking for something to give result like
NodeA
NodeB

Re: sort question

Rick Garland — Wed, 07 Sep 2005 10:41:05 GMT

In addition to piping your results to the sort utility, also pipe to the uniq utility

... | sort | uniq -c

This list each item individually with a count of how many times it is listed.

Re: sort question

machoq1 — Wed, 07 Sep 2005 10:43:08 GMT

Thanks !

Re: sort question

Mel Burslan — Wed, 07 Sep 2005 10:45:47 GMT

If this file contains nothing but the host names

you can use this:

myinfile=/path/somefilename
outfile=/path/myoutfilename

#in case multiple host names on one line we'll separate them
for h in `cat $myinfile`
do
echo $h > /tmp/temphostfile
done

for host in `cat /tmp/temphostfile|sort|uniq`
do
ct=`grep $host /tmp/hostfile |wc -l`
echo $host" "$count >> $outfile
done

if you can provide a section of your input file, you can get more sensible answers

Re: sort question

Alan Meyer_4 — Wed, 07 Sep 2005 10:47:28 GMT

Given a data file containing

NodeA
NodeC
NodeB
NodeB
NodeC
NodeA
NodeA
NodeA
NodeA
NodeC
NodeC

running the script

sort node.dat|uniq |while read node;do
echo "$node = \c"
echo "`grep $node node.dat |wc -l`"
done

produces

NodeA = 5
NodeB = 2
NodeC = 4