Shell Programming and Scripting

hello,
I need help on making a script that does the following job:
read an input file, sort it, then when finds lines with the same firt part (20 chars) count them (in the future I'd like to also make other sums)

can you please help me?
thanks a lot to all of you


#!/bin/ksh93
clear

#nl=`wc -l /usr1/air/tmp/1.txt`
#arrei_nl=($nl)
#NL=${arrei_nl[0]}
#echo $NL
(this part i suppose to be useless)

sort /usr1/air/tmp/1.txt > /usr1/air/tmp/2.txt

while read line
do
if [[ $line[i]:0:20} == ${ $line[i+1]:0:20} ]]
then
countl+1
:
else
echo $countl
fi
done < /usr1/air/tmp/2.txt



thanks again

Elio

Code:

awk ' arr[$0]++
        END{for (i in arr  ){print i,arr[i] }} ' inputfilename

thank you very much for your help,
I hoped you to find a solution without awk command (I'm not very used at it)...'cause now I don't know how to handle it
the next step would have been to check only the firt 20char and sum a number in the line...something like this:

AAA 1000
BBB 399
AAA 500
CCC 2000

result:
AAA 1000
AAA 500
tot 1500

BBB 399
tot 399

CCC 2000
tot 2000


thank you again

awk will be more helpful, don't shy away from it

to see if there are duplicate records where first 20 characters match

cut -c1-20 filename | sort | uniq -d > tempfile

step 2:
for a in 'cat tempfile'
grep -c $a filename

thanks paresh...I'm just starting to script...I think i can handle a lot ot other commands...but I've found awk so difficult to understand...
just to make an example...I don't know even how to cut and paste what you give to me