Need help

[suchitra]

I have a problem with comparing the columns of 2 files and write a status like add, modified or delete in the outfile along with the records of these 2 file.
For ex :

consider my 1st file has say
file1
==========
p1|y|500
p2|n|500

file2
======
p1| n| 500
p3 | y|501

now my output file should be like this

output file
=========
p1 | n| 500 |c b'cas it has been changed from y to n in the 1st file
p2| n |500 | d b'cas p2 is deleted in the 2nd file
p3 | y| 501 | a b'cas p3 is added in the 2nd file

could you please help me out. I want either a shell script or a awk command .

[Muthukumar_5]

Use this:

awk -F"\|" '{ var=$0;var1=$1;var2=$2;var3=$3;getline < "file2";split($0,a,"|"); if ( a[1] == var1 ) { if ( a[2] != var2 ) { print var"## c bcas "var2" is changed to "a[2];} if ( a[3] != var3 ) { print var"## c bcas "var3" is changed in 2nd file";} }
else { print var"## d bcas "var1" is deleted in 2nd file";print $0" ## a bcas "a[1]" is deleted in 2nd file";}}' file1

--
Muthu

[Steve Steel]

Hi

Look at the comm command

comm - select or reject lines common to two sorted files

And www.shelldorado.com


Steve Steel

[Peter Godron]

Suchitra,
my ititial solution:

#!/usr/bin/sh
echo "Changed"
echo "`join -j 1 -t '|' -o 1.1 2.2 1.3 file1 file2`|c"
echo "Deleted"
cut -f1 -d '|' file1 > file1.bck
cut -f1 -d '|' file2 > file2.bck
grep `comm -23 file1.bck file2.bck` file1 > file1.res
sed "1,$ s/$/|d/" file1.res
rm file1.res
echo "Added"
grep `comm -13 file1.bck file2.bck` file2 > file2.res
sed "1,$ s/$/|a/" file2.res
rm file2.res
rm file1.bck
rm file2.bck

[Peter Godron]

Muthukumar,
very smooth script!
The first "print var" could be replaced by:
print a[1]"|"a[2]"|"a[3]
to pick up the file2 values, rather than file1.

[Senthil Kumar .A_1]

Hi,

I have attached a script that uses "comm" command for your situation.

Regards,
Senthil Kumar .A

[Muthukumar_5]

Peter,

To avoid to print a whole line as a[1]"|"a[2]"|"a[3], I have stored in a separate variable. It is help ful to simply script ;)

--
Muthu

[Peter Godron]

Suchitra,
something else to keep in mind is that any script using comm would only work on sorted files.

Also, are there any duplicate keys like:
p1|y|500
p1|n|300
.
.

[Muthukumar_5]

Using perl:

#!/usr/bin/perl

open FD1,"file1" || die "Open Error: $!";
open FD2,"file2" || die "Open Error: $!";

@arr1=;
@arr2=;

for ($i=0;$i<@arr1;$i++)
{
@pat1=split (/\|/,$arr1[$i]);
@pat2=split (/\|/,$arr2[$i]);

$arr1[$i]=~chomp($arr1[$i]);
$arr2[$i]=~chomp($arr2[$i]);

if ( $pat1[0] eq $pat2[0] )
{
if ( $pat1[1] ne $pat2[1] )
{

print "$arr1[$i] | c b'cas it has been changed from $pat1[1] to $pat2[1] in field 2 in 2nd file\n";
}
if ( $pat1[2] ne $pat2[2] )
{
print "$arr1[$i] | c b'cas it has been changed from $pat1[2] to $pat2[2] in field 3 in 2nd file\n";
}

}
else
{
print "$arr1[$i] | d b'cas $pat1[0] is deleted in 2nd file\n";
print "$arr2[$i] | a b'cas $pat1[0] is added in 2nd file\n";
}

}

# END

--
Muthu

[Peter Godron]

Suchitra,
do these answers solve your problem?
Can you please have a look at:
http://forums1.itrc.hp.com/service/forums/helptips.do?#28
and then update the record.