Replace record in the previous line.

[shaokat]

I have a script that matches two fields (at position 7 length 20 and position 479 length 6) and inserts character '01' and '02' on subsequent matching records. Works fine so far.

Now, I want to replace (say position 6) with 'Y' in the first record of the matching condition.

ie. When I get to the '02' matching result, I want to replace position 6 with 'Y' of the previous line that matched '01'

How do I do this?

-KAT




awk '{
curr1 = substr($0, 7, 20);
curr2 = substr($0, 479, 6);
if (curr1 == last1 && curr2 == last2)
{
printf("02%s\n", $0)
}
else
{
printf("01%s\n", $0)
last1 = curr1
last2 = curr2
}
}'

 

[CaKiwi]

You will need to explain your problem a little better. Some sample input and output data would be helpful. CaKiwi

 

[marsd]

When you say "position 6" we have no clue as to what you 
are referring to..
If you are talking about substituting a record 
portion from a preceding NR you will need to load your 
file into an awk array, use numbered subscripts 
and create a method for reading and writing to file.




BEGIN {
   a = 1
   while ((getline array[a] < ARGV[1]) > 0) {
          a++;
   }
   close(ARGV[1])

   for (i=1 ; i <= a ; i++) {
        if (array[i] ~ matchingpattern) {
            do_change(array,(i - 1))
            changes++
        }
   }

   for (i=1 ; i <= a ; i++) {
       print array[i] >> ARGV[2]
   }
   close(ARGV[2])
   exit
}

 

[shaokat]

Here is the problem with samples. Forget the "position" mentioned earlier.

Consider this as input:

ABCD 123456789
DERS 234234234
ERED 345634563
RTYR 343434343
TYUT 343434343
TERE 368876655

This is the expected output:

ABCD 123456789
DERS 234234234
ERED 455634563
RTYR 343434343 Y
TYUT 343434343
TERE 368876655

Thanks for your response earlier.

 

[CaKiwi]

Untested.

{
if (NR > 0) {
printf s0
if (s2 == $2) printf " Y"
print ""
}
s0 = $0
s2 = $2
}
END {print s0}

Gotta go. CaKiwi