topic storing columns in different variables in Operating System - HP-UX

storing columns in different variables

Anand_30 — Fri, 20 Feb 2004 12:43:46 GMT

Hi,

Please find attached a file which has 2 columns. I need to store the column values in 2 different variables and use them in a SHELL script. I have an excutable which takes the coulmn values as parameters.

Thanks,
Andy

Re: storing columns in different variables

A. Clay Stephenson — Fri, 20 Feb 2004 12:48:58 GMT

Trivially simple:

MYFILE=myfile.txt

cat ${MYFILE} | while read A B
do
echo "A = \"${A}\" B = \"${B}\""
done

Re: storing columns in different variables

Chris Vail — Fri, 20 Feb 2004 16:50:01 GMT

Clay's script is, as he said, trivially simple. Its is limited to parsing the table from the beginning to the end. But what if you wanted to start at the end, and work up to the beginning? Here's a (probably needlessly) complex way to do this:

#!/usr/bin/ksh
LENGTH=`wc -l $FILE|awk '{ print $1 }'`
while test "$LENGTH" -ge "1"
do
LINE=`head -"$LENGTH" $FILE|tail -1`
A=`echo $LINE|awk '{ print $1 }'`
B=`echo $LINE|awk '{ print $2 }'`
(( LENGTH = "$LENGTH" - "1" ))
done

Another advantage (if there is one) is that not every version of Unix has the read command. But every one that I've ever seen has head, tail, and awk. Some versions (System V derivations, like old SCO) don't have wc either. If so, use
LENGTH=`pr -n -t $FILE|tail -1|awk '{ print $1 }'`
instead of the previous LENGTH statement. This uses up a lot more clock cycles than Clays script, but is portable across every version of Unix that I've ever seen. Its handy for log files.

Chris

Re: storing columns in different variables

john korterman — Sat, 21 Feb 2004 03:49:44 GMT

Hi,
this will probably also work:

# A=$(cut -f1 infile)
# B=$(cut -f2 infile1)

regards,
John K.