Shell Programming and Scripting

Hi All,
I have a file having content,
$ cat data1.txt
20060620 142 62310 959400 A 5.00
20060620 142 62310 959400 B 3.00
20060620 143 62310 959401 A 7.00
20060620 143 62310 959401 B 4.00
20060620 144 62310 959402 A 8.00
20060620 144 62310 959402 B 9.00
20060620 145 62310 959403 A 1.00
20060620 146 62310 959404 B 2.00

I need a awk file or one liner command, that will give me the output as below

20060620 142 62310 959400 A 8.00
20060620 142 62310 959400 B
20060620 143 62310 959401 A 11.00
20060620 143 62310 959401 B
20060620 144 62310 959402 A 17.00
20060620 144 62310 959402 B
20060620 145 62310 959403 A 1.00
20060620 146 62310 959404 B 2.00

Here the field separator is " ". For any two consecutive records, where $2 are same, $6 will be updated by adding rec1.$6+rec2.$6 for $5 = 'A' (first record) where as rec2.$6 will be substited by null( here $5 will be 'B' only, i.e second record). The records will remain unchange if $2 has one record only.
NOTE: For any value of $2, there can be maximum two records.

Thanks in advance,

Regards,
rin..

This should work

#! /bin/sh
while read line
do
key=`echo $line | awk '{print $2}'`
var=`echo $line | awk '{print $6}'`

awk -v key=${key} -v var=${var} 'key==$2 && var!=$6 {var2=var+$6;print $1,$2,$3,$4,$5,var2}' $1
done<$1>tempfile


sort -k 2,5 tempfile | awk '{if(NR%2!=0) print $0; else print $1,$2,$3,$4,$5}' > outputfile
rm -f tempfile

Thanks Buddy
This gave me the output as
20060620 142 62310 959400 A 8
20060620 142 62310 959400 B
20060620 143 62310 959401 A 11
20060620 143 62310 959401 B
20060620 144 62310 959402 A 17
20060620 144 62310 959402 B

But the last two lines are not here ( the lines without a pair). Can you please help me out in getting my toatal output?

This logic currently only adds only lines with multiple entries in the file. Will need extra logic to add remaining lines. A bit short of time, so may not be able to help you soon.

nawk -f rinku.awk data1.txt

rinku.awk:

Code:

prev == "" { prev = $0; f2=$2; flast=$NF; next }
$2 != f2 { print prev; prev = $0 ; f2=$2; flast=$NF; next}
{
   match(prev, "[^ ][^ ]*$")
   printf("%s%.2f\n", substr(prev, 1, RSTART-1), $NF + flast)
   NF--
   $1=$1
   print; prev = ""
}
END {
  print prev
}

#!/bin/sh

while read line
do
key=`echo $line | awk '{print $2}'`
var=`echo $line | awk '{print $6}'`
if [ `grep -c "$key" $1` -eq 1 ]; then
echo $line >> ofile
fi
awk -v key=${key} -v var=${var} 'key==$2 && var!=$6 {var2=var+$6;print $1,$2,$3,$4,$5,var2}' $1
done<$1>tempfile


sort -k 2,5 tempfile | awk '{if(NR%2!=0) print $0; else print $1,$2,$3,$4,$5}' >> ofile
rm -f tempfile

This one doesn't sorts data currently but shows all the rows. You can use
sort -k2 ofile to get the desired result.

Thanks Dude.
It worked..